Search is not available for this dataset
article
stringlengths 4.36k
149k
| summary
stringlengths 32
3.35k
| section_headings
sequencelengths 1
91
| keywords
sequencelengths 0
141
| year
stringclasses 13
values | title
stringlengths 20
281
|
|---|---|---|---|---|---|
The ability to respond to injury is a biological process shared by organisms of different kingdoms that can even result in complete regeneration of a part or structure that was lost . Due to their immobility , multicellular fungi are prey to various predators and are therefore constantly exposed to mechanical damage . Nevertheless , our current knowledge of how fungi respond to injury is scarce . Here we show that activation of injury responses and hyphal regeneration in the filamentous fungus Trichoderma atroviride relies on the detection of two danger or alarm signals . As an early response to injury , we detected a transient increase in cytosolic free calcium ( [Ca2+]c ) that was promoted by extracellular ATP , and which is likely regulated by a mechanism of calcium-induced calcium-release . In addition , we demonstrate that the mitogen activated protein kinase Tmk1 plays a key role in hyphal regeneration . Calcium- and Tmk1-mediated signaling cascades activated major transcriptional changes early following injury , including induction of a set of regeneration associated genes related to cell signaling , stress responses , transcription regulation , ribosome biogenesis/translation , replication and DNA repair . Interestingly , we uncovered the activation of a putative fungal innate immune response , including the involvement of HET domain genes , known to participate in programmed cell death . Our work shows that fungi and animals share danger-signals , signaling cascades , and the activation of the expression of genes related to immunity after injury , which are likely the result of convergent evolution .
The idea of regenerating lost body parts has fascinated humans since the beginning of history [1] . Human imagination was further captured upon witnessing the extraordinary capacity of species from almost all Phyla to , upon damage , regenerate a lost part or structure [1] . Nevertheless , our understanding of the biological significance and molecular mechanisms underpinning this remarkable phenomenon and its evolution is still poor . Multicellular organisms establish interactions with a great variety of other , potentially harmful , organisms throughout their life . Consequently , they require mechanisms to detect injury and to distinguish self- from non-self . Discrimination of self/non-self is a ubiquitous and essential function , which in animals relies on the immune system . Similarly , multicellular organisms require alarm signals known as Damage-Associated Molecular Patterns ( DAMPs ) to contend with a wound . In this regard , normal cellular components released into extracellular spaces , such as DNA , ATP and Ca2+ , represent reliable signals that indicate to other cells the disruption of tissue , and trigger a response [2] . Intracellular signaling after injury involves Mitogen Activated Protein Kinases ( MAPKs ) , as in the case of axon regeneration after spinal cord injury in vertebrates [3] and herbivory in plants [4] . Plant and animal cells have the ability to detect extracellular ATP ( eATP ) through recognition by specific , yet unrelated , receptors [2 , 5 , 6] . Activation of purinergic receptors or mechano-sensors triggers a transient increase in intracellular Ca2+ [7 , 8] . In plants , eATP promotes calcium influxes after wounding [9] and large increases in eATP serve as a key “danger” signals in the inflammatory processes of zebrafish and humans [8 , 10] . In animals , danger signals activate genes involved in cell signaling , stress responses , tissue patterning , cell matrix remodeling and growth [11–13] . Reactive Oxygen Species ( ROS ) are also considered as danger signals in plants and animals , and necessary to prevent infections during wound healing [2] . Sustained production of ROS is required for regeneration in Xenopus and zebrafish [14 , 15] . Furthermore , ROS is involved in the regulation of intracellular Ca2+ levels [16] . During tissue regeneration , a competent immune system is essential for effective wound healing [17] . Common features of innate immunity in vertebrates , invertebrate and plants include the basic chemical structure of signal molecules , signaling cascades , production of antimicrobial molecules , and transcriptional activation of defense genes [2 , 18] . Fungi like other organisms have natural predators , including fungivorous nematodes and arthropods , consequently they need effective mechanisms to contend with and survive injury . When fungal hyphae are damaged , the septal pore nearest to the point of injury is sealed to prevent excessive cytoplasmic leakage . Thereafter , new hyphal tips are generated from this position , resulting in regeneration and re-initiation of growth [18] . The common soil fungus Trichoderma atroviride responds to mycelial injury by rapidly regenerating its hyphae and , developing asexual reproductive structures ( conidia ) in a NADPH oxidase ( Nox ) dependent manner [19] . Interestingly , application of eATP also induces conidiation [20] . Like other multicellular eukaryotes , this fungus appears to perceive eATP , through a yet unidentified receptor , which triggers activation of the MAPKs Tmk1 and Tmk3 [20] . Tmk3 is activated in a Nox1-NoxR dependent fashion , whereas Tmk1 activation is independent of Nox [20] . Mutants in either tmk1 or tmk3 are affected in injury-induced conidiation . Interestingly , depletion of extracellular Ca2+ blocked injury-induced conidiation but allowed activation of both MAPKs [20] . In addition to its role in injury-induced conidiation , Tmk3 regulates tolerance to heat shock , osmotic and oxidative stress , and cell wall integrity [21] . Unexpectedly , tmk3 mutants are also impaired in light-induced conidiation [21] . In contrast Δtmk1 mutants , albeit of a different T . atroviride strain , hyperconidiate under standard cultivation conditions [22] . Tmk1 has also been suggested to regulate mycoparasitic activity and hyphal fusion [22 , 23] . Here we provide new mechanistic insights into the activation of a regeneration program , consisting of genes involved in cell signaling , stress responses , transcription , ribosome biogenesis/translation , DNA replication , growth , and defense through Ca2+/MAPK-dependent signaling pathways . Finally , we uncover the activation of genes of a putative fungal innate immune response involving genes previously known to participate in heterokaryon incompatibility [24] .
Considering the relevant role of calcium in regeneration in other organisms , we evaluated calcium signatures at wound sites in T . atroviride . Live-imaging analysis of wounded cells , expressing the Ca2+ sensor GCamP6 , revealed a transient spike of cytosolic free calcium ( [Ca2+]c ) immediately after injury ( Fig 1A , S1 and S2 Movies ) . To determine if extracellular calcium was involved in this response , we applied the calcium-chelating agent BAPTA . Injury promoted a transient elevation of [Ca2+]c , while treatment with BAPTA prior to injury completely suppressed this response ( Fig 1A ) . The injury-induced calcium signature showed a very rapid and strong peak of increase in [Ca2+]c that decreased over time ( Fig 1B ) . In contrast , in presence of BAPTA , the [Ca2+]c was not altered upon injury ( Fig 1B ) . The decrease in fluorescence observed in the graphs just before injury ( indicated by an arrow ) is an unavoidable artifact due to the introduction of the scalpel used to cause the injury , which blocks light . To test if the observed response involved uptake of extracellular calcium and/or release of calcium from intracellular pools , we applied the calcium release inhibitors verapamil , which blocks L-type calcium channels in the plasma membrane , and dantrolene that inhibits Ca2+-induced Ca2+release from the sarcoplasmic reticulum pool by targeting the ryanodine receptor [25] . Both compounds significantly reduced [Ca2+]c increases due to injury ( Fig 1C ) , consistent with the participation of a calcium-induced calcium release system . Next , we analyzed the role of calcium in the regeneration process . Application of BAPTA to the fungal mycelium prior to injury reduced regenerating hyphae to 20% , as compared to 64% observed in an untreated control ( Fig 1D and 1E ) . Addition of extracellular calcium to hyphae previously exposed to BAPTA partially restored regeneration upon damage , as evidenced by the formation of thin new hyphae as a result of tip growth re-initiation ( Fig 1D ) and an increase in regeneration ( 52%; Fig 1E ) . We then compared the transcriptional profile of the fungus in response to Injury in the presence ( IB ) and absence ( I ) of BAPTA . We identified a total of 421 Calcium-Dependent Injury Genes ( CDIGs ) , responsive only in the absence of BAPTA; of which 241 were up-regulated and 180 down-regulated . In addition , we identified 404 Calcium-Independent Injury-Genes ( CIIGs ) , responsive even in the presence of BAPTA; 201 of which were up-regulated and 203 down-regulated ( Fig 2A , S1–S3 Datasets ) . Upon injury , the fungus up-regulates genes associated to the following cellular components: chromosomes , lumen enclosed by membrane , ribonucleoprotein complexes , intracellular non-membrane-bound organelles ( associated with the cytoskeleton ) and macromolecular complexes ( Fig 2B , WT-I vs WT-C , S4 Dataset ) . BAPTA chelation of calcium blocked genes that were associated with chromosomes , and non-membrane-bounded organelles , but induced genes related to microbodies ( vacuoles ) , apparently in an injury independent fashion ( Fig 2B , WT-IB vs WT-C & WT-IB vs WT-I ) . Moreover , injury caused an increase in genes belonging to the biological processes of RNA processing , translation , DNA metabolic processes and replication ( Fig 2C; WT-I vs WT-C ) . Remarkably , upon injury , but in the absence of extracellular calcium , there was no significant enrichment in these processes ( Fig 2C , WT-IB vs WT-C ) . We had previously suggested that eATP could function as a DAMP [20] . To determine if indeed eATP serves as a DAMP that triggers Ca2+ influxes in T . atroviride after wounding , we evaluated Ca2+ dynamics and regeneration , upon addition of eATP or apyrase , an enzyme that hydrolyses ATP to AMP . Addition of eATP without injury provoked an increase of [Ca2+]c ( Fig 3A ) . However , when the mycelium was pre-treated with apyrase , the fluorescence signal after injury was abolished , as compared to an untreated ( no apyrase ) control ( Fig 3A ) . Furthermore , damaged hyphae treated with apyrase showed a strongly reduced regeneration capacity ( Fig 3B ) : only 26% of hyphae regenerated ( Fig 3C ) . We also evaluated the transcriptional changes that occur when adding eATP; remarkably , as upon injury , we found that gene expression associated with DNA replication , the cell cycle , RNA biosynthetic processes and organic acid transport was induced ( S1 Fig , S5 Dataset ) . Thus , ATP released from damaged cells promotes regeneration , likely by promoting calcium influxes . Similarly , we evaluated if ROS is necessary for triggering Ca2+ influx and hyphal regeneration . For this purpose , we evaluated changes in [Ca2+]c upon injury in presence of the antioxidant N-acetyl-Cysteine ( NAC ) and its analog N-acetyl-glycine ( NAG ) , as control . After injury , a spike of [Ca2+]c was observed even in the absence of ROS ( NAC treatment ) ( Fig 3D ) , suggesting that ROS are not required to trigger Ca2+ influxes and possibly regeneration . To determine if NOX-dependent ROS production plays a role in hyphal regeneration , we performed regeneration assays using the Δnox1 , Δnox2 and ΔnoxR mutants . In all cases emergence of new hypha from the cell adjacent to the broken one was observed ( Fig 3E ) . All mutants showed the same capacity to regenerate observed in the WT strain ( Fig 3F ) . To explore if ROS regardless of its source played a role in regeneration , we applied a NAC treatment prior to injury , observing no difference in the percentage of regeneration compared with the untreated control ( Fig 3F ) . These results are consistent with eATP acting as a DAMP ( a form of “danger signal” ) , that activates a Ca2+ influx which is necessary for hyphal regeneration . In contrast , ROS are not involved in causing the transient elevation in [Ca2+]c nor in the regeneration process . To determine if MAPK signal transduction pathways are involved in the regeneration process , we used gene replacements mutants of the MAPK encoding genes tmk1 and tmk3 . Hyphal regeneration in the WT , Δtmk1 and Δtmk3 strains was analyzed after damage with a scalpel . The regenerative capacity of the Δtmk1 mutant was drastically affected , since in many cases we did not observe the emergence of new hyphae ( Fig 4A ) . On average , only 20% of the hyphae showed regeneration in the absence of Tmk1 , compared with 68% in the wild type strain ( Fig 4B ) . In contrast , the Δtmk3 mutant exhibited only a slight decrease in regenerative capacity , with 56% of the hyphae regenerating ( Fig 4B ) . To identify injury-responsive genes linked to MAPKs , we performed a transcriptional analysis using RNA extracted from the Δtmk1 and Δtmk3 mutants upon injury and compared their transcriptional profiles with that of the WT strain . We identified a set of Injury-Responsive Tmk1 dependent ( IRK1 ) genes , whose expression changes upon injury in the WT and Δtmk3 strains ( 94 up-regulated and 127 down-regulated ) were no longer observed in the Δtmk1 mutant , which shows strongly reduced regeneration ( Fig 4C , S6 & S7 Datasets ) . These analyses clearly showed that most clusters of genes differentially expressed in both the WT and Δtmk3 strains remained nearly unresponsive in the Δtmk1 mutant ( Fig 4D ) . The IRK1 genes included key elements of cell signaling , DNA replication , and DNA metabolic processes ( S8 Dataset ) . The pattern of expression of the IRK1 genes in the Δtmk3 mutant was similar to that observed for the wild type strain at early stages of the response to injury . Nevertheless , we detected 13 up-regulated and 19 down-regulated genes that did not respond to injury in the Δtmk3 mutant but responded to the stimulus both in the WT and Δtmk1 strains ( Fig 4C ) . Given that the latter strains can proceed into conidiation , this set of Injury-Responsive Tmk3 dependent ( IRK3 ) genes might represent a set involved in the onset of conidiation ( S8 Dataset ) . A Gene Ontology analysis of the transcriptional response to injury in each strain suggests that the functional response in the tmk3 mutant is quite similar to that of the WT ( Fig 4E ) . Although cellular component organization and RNA processing categories responded similarly in all three strains , the tmk1 mutant did not show a significant up-regulation of genes involved in DNA replication nor down-regulation of several metabolic processes , including that of reactive oxygen species ( Fig 4E ) . On the other hand , genes encoding proteins associated with actin filaments and proteolysis , likely required for regeneration , were only down-regulated in Δtmk1 ( Fig 4E ) . Until now , two of our experimental conditions led to an impaired regeneration capacity after injury: Δtmk1 and WT treated with the Ca2+ chelator , BAPTA . We hypothesized that these facts could be used to define genes required for regeneration by comparing gene expression profiles between regenerating and non-regenerating conditions/strains: ( WT injury , Δtmk3 injury ) versus ( Δtmk1 injury , BAPTA injury ) . To make sure that the resulting genes respond to injury in the WT , we then compared them with those differentially expressed in the comparison WT injury vs WT control ( Fig 5A ) . The intersection of both groups represents what we define as the “Regeneration Associated Gene Set” ( RAGS ) , which expression is associated with regeneration , constituted by 520 up-regulated and 466 down-regulated genes ( Fig 5A ) . Based on Gene Ontology , the induced component of the RAGS was enriched in genes involved in DNA and RNA metabolic processes , cellular responses to stress , the cell cycle , and ribosome biogenesis , among others ( S2 Fig ) . However , no significant functional enrichment was found for the down-regulated genes . Closer inspection and manual annotation allowed us to classify the induced set of genes into 7 functional categories ( Fig 5B , S9 Dataset ) . Six of the RAGS categories are involved in processes clearly associated with regeneration in many other organisms: cell proliferation ( 35 genes ) , replication ( 43 genes ) , cellular signaling ( 10 genes ) , response to stress/DNA repair ( 25 genes ) , transcription regulation ( 69 genes ) , and ribosome biogenesis/translation ( 36 genes ) . Interestingly , the seventh group ( 22 genes ) is mostly constituted by genes related to programmed cell death and genes with a fungus-specific HET domain . Upon manual inspection of the down-regulated RGS , we found a large number of genes involved in oxido-reduction processes ( S10 Dataset ) . As expected , when looking at the individual treatments/strains , we observed that the RAGS did not respond to injury or showed a strongly diminished response in the Δtmk1 mutant and upon BAPTA treatment . Instead they followed similar patterns of expression to those of the Δtmk3 and WT strains upon injury , and after addition of ATP ( Fig 5C ) . Additionally , we observed that application of eATP mostly mimicked the response of the genes provoked by injury in the WT ( Fig 5C ) . Interestingly , the genes of the ribosome biogenesis/translation and transcription regulation categories , which are up-regulated upon injury of the WT , are down-regulated upon treatment with eATP . These observations indicate that eATP is not sufficient to trigger a full regeneration response ( Fig 5B ) . To better understand the dynamics of expression of the RAGS we selected four genes belonging to three different categories , namely cell proliferation ( rad5 , Id . 172559 ) , cell signaling ( cmk1 , Id . 301592 ) , and programmed cell death ( Het domain , Id . 294334; Nacht domain , Id . 88516 ) for quantitative gene expression analyses . As expected , we observed a strong increase in the level of all four transcripts early after injury , reaching their maximum at 15 min , when the new regenerating hyphae become barely visible ( Fig 6 ) . The expression of the genes clearly decreases by 30 min , when the new hyphae have already emerged ( Fig 6 ) . The level of expression of all four genes then starts dropping slowly , and 5 h after injury , when most ( 90% ) regenerated hyphae were completely evident ( Fig 6A ) , they reached the level observed in the control .
We have previously shown that T . atroviride responds to mycelial injury by rapidly regenerating its hyphae and , developing conidia in a Nox-dependent manner [19] . We had also shown that eATP induces conidiation and triggers activation of Tmk1 and Tmk3 , and that the latter is activated in a Nox1-NoxR dependent fashion [20] . Further , mutants in either tmk1 or tmk3 were affected in injury-induced conidiation , which is the final outcome of the process [20] . Intriguingly , depletion of extracellular Ca2+ blocked injury induced conidiation but allowed activation of both MAPKs [20] . In this regard , Ca2+ released from a damaged cell may be detected by neighboring cells as a signal molecule or serve as a second messenger liberated from intracellular pools and/or be transported across the plasma membrane upon detection of DAMPs . However , it was unclear whether the regeneration and conidiation processes were mechanistically linked and how all these elements were interconnected to regulate the response to damage . In this regard , two of the earliest signaling events after wounding in animals and plants are the activation of MAPKs and Ca2+ influxes [2] . MAPKs are also involved in the early stages of regeneration in hydra [26] and planaria [27] . Likewise , here we show that the MAPK Tmk1 is involved in hyphal regeneration control . Furthermore , [Ca2+]c increases are necessary for sealing the disrupted plasma membrane [28] , and Ca2+ signaling is essential to activate defense responses in plants and the immune system in mammals [29 , 30] . Similarly , we detected a transient increase in [Ca2+]c seconds after damage , which was promoted by eATP , as one of the earliest events of the response to injury with evidence that the increase in [Ca2+]c is regulated by extracellular Ca2+ involving a mechanism known as Ca2+-induced Ca2+-release [31] . Interestingly , in filamentous fungi , deletion of Cch1 or Mid1 , which are components of a mechanosensing Ca2+-channel complex , results in diminished thigmotropic responses , and failure to establish cell polarity [32] . A comparable mechanism of mechanosensing is used by mammals during immune cell activation [33 , 34] . A direct link between the increase in [Ca2+]c and the control of transcription , was revealed by the up-regulation of the Ca2+/calmodulin dependent kinase CAMK1 , and the transcriptional factor CRZA in response to injury . In planaria and hepatic stem cells calmodulin encoding genes are induced by wounding [13 , 35] , and intimately linked with the signals activating the innate immune system [36] . Moreover , we found that increasing [Ca2+]c and eATP are necessary for the formation of new , regenerated hyphal tips and re-initiation of mycelial growth upon injury . In this regard , it has been established that actin cytoskeleton rearrangements are Ca2+-dependent and determine the site of tip growth and polarized cell movement [32 , 34] . In plants and animals , ROS production is required during healing and regeneration [14 , 15 , 37] . In contrast , it appears that ROS are not required for hyphal regeneration , although they could participate as signal molecules in the early response inducing genes involved in injury-induced conidiation [18 , 19] . This is consistent with the fact that Tmk3 is activated by ROS , both necessary for injury-induced conidiation but not regeneration [20] . Thus , as in plants and animals , signaling by eATP , Ca2+ and MAPKs appears to be essential for regeneration in filamentous fungi . We further show that the Ca2+ and Tmk1 signaling pathways , appear to control the expression of thousands of genes . However , we defined a Regeneration Associated Gene Set ( RAGS ) in which these two signaling pathways converge . The RAGS could be involved in either the metabolic changes required to produce a new hypha and reinitiate growth ( regeneration ) , respond to mechanical stress , and/or defense . Six different processes that could clearly be linked to regeneration are strongly represented within the RAGS . Some of the individual genes found within the RAGS encode proteins involved in cell cycle regulation and two components of the condensin complex , whose participation in regeneration processes has been documented [38 , 39 , 40] . It is noteworthy that we also found genes , such as ssu72 , required for replication initiation and previously shown to participate in the control of cell cycle progression in mice in response to liver damage [41] . We also found DNA replication licensing factors ( mcm genes ) , which are up-regulated during regeneration in planaria , mice and axolotl [42–44] , and have been shown to play a key role in cell proliferation [45] . We defined a seventh group within the RAGS , which contains eight genes encoding HET domain proteins . In this regard , fungal hyphae from different individuals can fuse , resulting in the coexistence of genetically different nuclei in a common cytoplasm ( heterokaryon ) . The fate of the fused cell is determined by HET domain proteins through allorecognition processes , in which heterokaryons resulting from the fusion undergo a type of programmed cell death [24 , 46] . In addition , we found two induced homeodomain transcription factors , annotated as potential mating type factors . In Neurospora crassa , some HET domain proteins interact with mating factors to carry out the heterokaryon incompatibility process . Allorecognition processes allow the distinction of self from non-self in cells and tissues , and participate in processes , ranging from tissue transplant fusion to immune defense , across the tree of life [47] . Remarkably , in addition to HET domain protein encoding genes , we found 14 genes which participate in either cell death or the innate immune system in animals . Among them a caspase , a putative phosphatidylserine-specific receptor , a PITSLRE protein kinase , and the activation of apoptosis signal-regulating kinase 1 , all of which play major roles in apoptosis [48–50] . Other interesting genes within this group were a 3–5 exoribonuclease csl4 , a probable GMP synthase , and a Ca2+-independent phospholipase A2 , which are key elements of the innate immune response in animals [51–53] . Fungi , like all organisms , are potential hosts for microbial pathogens and have developed defense systems against competitors and pathogens . Programmed cell death and non-self-recognition systems are considered an important strategy to contend with infections in fungi , plants and animals [16] . Furthermore , heterokaryon incompatibility has been shown to prevent various forms of somatic parasitism , and to reduce the risk of transmission of infectious cytoplasmic elements and mycoviruses [54–56] . This set of HET domain proteins together with two DEAD/H-box helicases present in our RAGS , and which have been implicated in cytosolic DNA sensing [57] , could play a major role in detecting damaged or invading DNA molecules . Furthermore , in plants and animals , the innate immune response relies on specific proteins , the pattern-recognition receptors ( PRRs ) , which detect conserved pathogen-associated molecular patterns and “danger” signals [2 , 58] . The domain architecture of HET proteins is similar to that of both plant and animal cytosolic PRRs , suggesting that similar modes of activation occur even if primary sequences and downstream functions are diverse [52 , 59] . These incompatibility genes are extremely polymorphic and show signatures of diversifying selection [52 , 60] . Moreover , recent biochemical evidence showed that fungal NLR-like proteins function similarly to NLR immune receptors in plants and animals , concluding that NLRs are major contributors to innate immunity in three kingdoms , including fungi [59] . Consequently , HET proteins may be involved in protecting the fungus from pathogens , invading DNA/RNA molecules , and serve as damaged self-recognition system , as components of an innate immune system . Based on these observations , we propose that a filamentous fungal innate immune response process promotes regeneration and we designate this set of 22 genes as the innate immunity group . The importance of the activation of the immune system in regeneration in organisms such as zebrafish and hydra has previously been documented [61 , 62] . Thus , we postulate that there is a cellular Boolean system that determines entry into cellular proliferation , mediated by different genes involved in perception of exogenous and/or damaged genetic material , which lead to cell death if the cell/tissue damage is too extensive or caused by pathogenic organisms , since initiating DNA replication would compromise genome integrity . However , if this were not the case , rapid communication with the DNA repair and replication systems would take place and regeneration would be promoted . According to Sanchez-Alvarado [1] , the molecular cascades associated with regeneration may have appeared first and foremost as a way to asexually propagate species , and that such cascades may have been co-opted by many organisms to cope with injury . Here we show that filamentous fungi , which have the capacity to reproduce by fragmentation , share many elements thought to be exclusively used by animals for regeneration . Furthermore , although speculative at this stage , our data suggest that in fungi an innate immune system is involved in hyphal regeneration , with the participation of HET domain proteins , previously thought to exclusively trigger programmed cell death [24] . In this regard , although the domain architecture and mechanistic functioning of NLR proteins are strikingly similar , their evolution in the different kingdoms of life is thought to be convergent [63 , 64] . Thus , our findings indicate that the signaling pathways involved in regeneration across kingdoms were likely co-opted to aid this process after multicellularity evolved .
Trichoderma atroviride IMI 206040 was used as the wild type strain ( WT ) . The Δtmk1 and Δtmk3 mutants have been described previously [20] , as have the Δnox1 , Δnox2 , and ΔnoxR mutants [19] . The T . atroviride strain carrying the Ca2+ sensor GCamP6 [65] ( Calmodulin::GFP ) , was obtained by transformation with the plasmid pEM12 ( see below ) . All strains were propagated on potato dextrose agar . To generate pEM12 , the sequence of GCamP6 ( Calmodulin::GFP ) was obtained from plasmid pSK379 [66] and amplified by PCR using the primers GCaMP6-EcoRI-FW and GCaMP6-SalI-RV . T . atroviride was then transformed with pEM12 , as previously described [67] , and subjected to five passes through monosporic culture . All oligonucleotides used are indicated in S1 Table . Colonies of T . atroviride expressing the Ca2+-sensor GCamP6::GFP ( pEM12 ) were grown on Vogel’s minimal medium ( VMM ) or Potato Dextrose Broth in 0 . 5% agar and incubated for 36 h on glass slides ( Corning ) . Hyphae were damaged approximately 80 μm behind the tips of leading hyphae , using a scalpel . They were visualized using a confocal laser scanning microscope ( CLSM ) Olympus FluoView FV1000 ( Olympus , Japan ) fitted with an argon/2 ion laser ( EGFP: excitation , 488 nm; emission , 510 nm ) . A 60 X Plan oil-immersion objective ( 1 . 42 N . A . ) was used for image acquisition . Image projections consisting of stacks of images were captured at 5 min intervals and converted into movies using confocal FluoView FV1000 ( Olympus Corp . ) software . To analyze Ca2+ fluxes during cell damage , we used the following Ca2+-modulators: 5 mM verapamil , 100 μM dantrolene , and 10 mM of cell impermeant 1 , 2-bis- ( o-aminophenoxy ) -ethane-N , N , N’ , N’-tetraacetic acid ) tetrapotassium Salt ( BAPTA ) ( Life Technologies ) . The mycelium was incubated with these modulators for 15 minutes . The colonies were then damaged with a scalpel and visualized with a CLSM . To analyze the role of eATP in the promotion of Ca2+ fluxes , we treated the mycelium for 15 min with 2 units of apyrase ( Sigma ) , an enzyme that hydrolyzes ATP , before damaging the cells . Confocal imaging was performed immediately following injury or addition of 100μM eATP . To analyze the role of ROS , we exposed the WT strain to 30 mM N-acetyl-cysteine ( NAC ) or 30 mM N-acetyl-glycine ( NAG ) for 15 min before damage . Untreated colonies without treatment were used as controls . The fluorescence intensity was quantified per hypha using Image J software . The mean change in fluorescence measurements were made in a 50 μm x 30 μm region of interest drawn over 10 hyphae approximately 80 μm back from their tips . R packages were used for statistical analysis . Colonies of T . atroviride WT , Δtmk1 , Δtmk3 , Δnox1 , Δnox2 , and ΔnoxR mutants were grown in half strength PDB supplemented with 1% agarose for visualization of isolated regenerating hyphae as described above and incubated for 48 h at 27°C . The wild type strain was exposed to the Ca2+ modulators or apyrase ( Sigma ) , as described above . Mycelia of the different strains/treatments were then damaged with a scalpel and incubated for 5 h . Finally , the mycelium was stained with lactophenol cotton blue for 10 min . Mycelia were observed on a Leica DM6000-B microscope fitted with a 40x objective HCX PL Fluotar ( 0 . 75 N . A . ) and photographed with a Leica DFC 420C camera . R packages were used for statistical analysis . Mycelia of the Δtmk1 , Δtmk3 and WT strains were collected 30 min following mycelial damage and frozen immediately . The WT strain was previously treated with 10 mM BAPTA or 100 μM ATP for 15 min , as indicated . In all cases , an injured control without chemical treatment and a control without injury were included , and three biological replicates were analyzed per strain and/or treatment . Total RNA was extracted with TRIzol ( Invitrogen ) . Libraries for RNAseq were prepared using the TruSeq RNA library preparation protocol ( Illumina ) . Each library was sequenced using a NextSeq500 sequencer in the 1x75 format . The 75-bp reads were pseudo-aligned to the T . atroviride V2 transcripts , using kallisto [68] . On average , 25 million reads per library were obtained with high quality ( S2 Table ) . The RNAseq data discussed in this publication have been deposited in NCBI’s Gene Expression Omnibus [69] and are accessible through GEO Series accession number GSE115811 ( https://www . ncbi . nlm . nih . gov/geo/query/acc . cgi ? acc=GSE115811 ) . For the differential gene expression analysis , only those genes that had at least three counts per million in at least ten libraries were considered as transcribed . All analyses were carried out using the edgeR package [70] . For determining differential expression between the comparisons , we used the generalized linear model ( GLM ) likelihood ratio test . False discovery rates ( FDR ) were calculated and genes with an FDR < 0 . 05 and absolute log2Fold-change ≥ 1 were considered differentially expressed . Venn diagrams and heatmaps were constructed to compare the universes of differentially expressed genes using the gplots package in R . To identify the set of genes related to regeneration , a highly restrictive differential expression analysis was performed with a value of FDR < 0 . 01 . A Fold-change cutoff was not considered in this case . Enrichment analyses for Cellular Component and Biological Process GO terms were performed using camera , from the edgeR package [71] . GO terms with FDR ≤ 0 . 05 were considered significantly enriched in each comparison . We have presented this data in a clustered heatmap that highlights the categories enriched with asterisks , where ** represents FDR < 0 . 01 and * FDR < 0 . 05 . The plotted values are the percentage of genes belonging to each category that are deemed differentially expressed ( see previous section ) . The GO terms were first filtered for redundancy , removing those that contained more than 1000 or less than 4 genes . To validate the differential expression of regeneration associated genes , primers for qRT-PCR were designed to produce amplicons around 150 bp ( S1 Table ) . cDNA was synthesized using as template RNA extracted from injured and control mycelia , and RT II SuperScript ( Invitrogen ) using four biological and three technical replicates . The reaction mixture for quantitative PCR was as follows: 10 μl of SYBR green master mix ( Applied Biosystems ) , 3 μl of cDNA template ( 3ng/μl ) and 1 μl of each ( 10 μM ) of the primers . The PCR program was as follows: One cycle at 95°C for 5 min , 40 cycles at 95°C each for 30 s , at 65°C for 30 s , 72°C for 40 s . Melting curves for each product , starting from 60°C to 95°C at 0 . 2°C/s , produced a single melting point . All qRT-PCR reactions were repeated three times . Anova and Tuckey tests were performed to determine the significance of changes in gene expression . | The idea of regenerating lost body parts has always fascinated humans due to its impact on human health . Recently , the study of the response to damage has gained importance not only in regenerative medicine , but also in organ transplantation . We have established a microbial model that given its relative simplicity and ease to work with , promises to accelerate the advance of our understanding of damage responses and regeneration . We show that activation of injury responses and hyphal regeneration in a filamentous fungus relies on the detection of danger ( also called alarm ) signals also used by plants and animals . Finally , we discovered a set of genes that become active in response to injury , including those putatively participating in the immune response . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] | [
"fungal",
"genetics",
"statistics",
"light",
"microscopy",
"fungal",
"structure",
"fungi",
"statistical",
"data",
"confocal",
"laser",
"microscopy",
"mathematics",
"dna",
"replication",
"microscopy",
"genome",
"analysis",
"confocal",
"microscopy",
"dna",
"research",
"and",
"analysis",
"methods",
"mycology",
"gene",
"expression",
"gene",
"ontologies",
"biochemistry",
"eukaryota",
"nucleic",
"acids",
"genetics",
"biology",
"and",
"life",
"sciences",
"physical",
"sciences",
"genomics",
"computational",
"biology",
"organisms"
] | 2018 | Danger signals activate a putative innate immune system during regeneration in a filamentous fungus |
Lynch syndrome ( hereditary nonpolypsis colorectal cancer or HNPCC ) is a common cancer predisposition syndrome . Predisposition to cancer in this syndrome results from increased accumulation of mutations due to defective mismatch repair ( MMR ) caused by a mutation in one of the mismatch repair genes MLH1 , MSH2 , MSH6 or PMS2/scPMS1 . To better understand the function of Mlh1-Pms1 in MMR , we used Saccharomyces cerevisiae to identify six pms1 mutations ( pms1-G683E , pms1-C817R , pms1-C848S , pms1-H850R , pms1-H703A and pms1-E707A ) that were weakly dominant in wild-type cells , which surprisingly caused a strong MMR defect when present on low copy plasmids in an exo1Δ mutant . Molecular modeling showed these mutations caused amino acid substitutions in the metal coordination pocket of the Pms1 endonuclease active site and biochemical studies showed that they inactivated the endonuclease activity . This model of Mlh1-Pms1 suggested that the Mlh1-FERC motif contributes to the endonuclease active site . Consistent with this , the mlh1-E767stp mutation caused both MMR and endonuclease defects similar to those caused by the dominant pms1 mutations whereas mutations affecting the predicted metal coordinating residue Mlh1-C769 had no effect . These studies establish that the Mlh1-Pms1 endonuclease is required for MMR in a previously uncharacterized Exo1-independent MMR pathway .
DNA mismatch repair ( MMR ) acts to repair the potentially mutagenic misincorporation errors that occur during normal DNA replication and the absence of MMR results in increased rates of accumulating mutations . Consequently , defects in human MMR genes cause the hereditary cancer susceptibility syndrome HNPCC ( hereditary nonpolypsis colorectal cancer , otherwise known as Lynch syndrome ) [1] , [2] , and loss of MMR function also appears to underlie the development of some sporadic cancers [3]–[7] . MMR also repairs mispaired bases that occur in recombination intermediates as well as prevents inappropriate recombination between DNAs with imperfect homology where recombination could result in genome rearrangements [8]–[10] . The mechanism of MMR has been extensively characterized in both E . coli and different eukaryotic systems , with E . coli MMR being the best characterized [11]–[14] . In E . coli MMR , mismatches are recognized by the MutS homodimer [15] , [16] . Mispair bound MutS then recruits the MutL homodimer [17] . This recruitment leads to activation of the MutH endonuclease , which introduces single strand breaks , called nicks , at unmethylated GATC sites in the newly replicated and hemimethylated DNA strand [18] . Next , a combination of the UvrD helicase and one of four single stranded DNA specific exonucleases excise the nicked strand past the mispair and the resulting singled-stranded gap is filled in by DNA polymerase III , single strand DNA binding protein and DNA ligase [14] , [19] . In eukaryotes mispairs are recognized by either Msh2-Msh6 or Msh2-Msh3 , two partially redundant heterodimers of MutS family member proteins [12] , [20] , [21] . Mispair bound Msh2-Msh6 and Msh2-Msh3 recruit the MutL related complex , called Mlh1-Pms1 in S . cerevisiae and Mlh1-Pms2 in human and mouse [11] , [12] , [22]–[24] . The Pms1/Pms2 subunit of the Mlh1-Pms1/Pms2 complex is known to contain an endonuclease active site , suggesting that Mlh1-Pms1/Pms2 may be analogous to a combination of both E . coli MutL and MutH [25] , [26] . Exo1 , a DNA exonuclease from the Rad2/XPG family , has been implicated in the excision step of eukaryotic MMR; however , mutations in S . cerevisiae and mouse EXO1 only result in partial MMR defects , suggesting the existence of additional excision mechanisms [27]–[29] . Genetic and biochemical studies have also implicated DNA polymerase δ , RPA , RFC and PCNA in MMR [12] , [30]–[37] and have suggested that several of these proteins including PCNA and RFC may function both prior to excision and in the resynthesis steps of MMR [25] , [26] , [33] , [38] . MMR is spatially and temporally coupled to replication in vivo [38] , [39] , providing a mechanism to bring MMR proteins into the proximity of newly formed mispairs . DNA replication generates nicks in the nascent DNA strands that may be involved in MMR [30] , [34] , [40] , [41] , consistent with the observation that discontinuous lagging strand MMR is more dependent on excision catalyzed by Exo1 than leading strand MMR [38] , [42] . Furthermore , preexisting nicks in DNA target the Mlh1-Pms1/Pms2 endonuclease to the nicked strand in vitro [43] . However , these results raise two unresolved questions: Why is it necessary to target additional nicks to an already nicked DNA strand , and if preexisting nicks can support MMR in vitro then why is Mlh1-Pms1 absolutely required for MMR in vivo ? Part of the answer to the apparent contradictions implied by these experimental results could be the presence of multiple MMR pathways in which the same MMR proteins have differing roles . Consistent with this , biochemical studies have identified two types of excision mechanisms that may function in MMR , excision by Exo1 [44] and strand displacement synthesis toward the mispair by DNA polymerase δ potentially coupled with flap cleavage [45] . Both mechanisms could act at either a pre-existing 5′ nick or a 5′ nick introduced by the Mlh1-Pms1/Pms2 endonuclease . Genetic studies have also identified Exo1-dependent and -independent MMR pathways [27] . The Exo1-independent pathway requires the PCNA-Msh2-Msh6 interaction and the Pol32 subunit of DNA polymerase δ and is inactivated by separation-of-function mutations affecting Mlh1 , Pms1 , Msh2 , Msh3 , and PCNA that do not affect Exo1-dependent MMR [27] , [38] . How these mutations specifically affect in the Exo1-independent MMR pathway and how this pathway excises the nascent DNA strand is unclear . To better understand the role of both Mlh1-Pms1 and Exo1 in MMR , we performed a genetic screen for dominant mutations in the PMS1 gene . We identified pms1 null missense mutations that caused weakly dominant MMR defects when present in a wild-type S . cerevisiae strain on a single-copy plasmid . Interestingly , these mutations caused much stronger MMR defects when present on a single-copy plasmid in an exo1Δ mutant . Analysis of these mutations using the structure of the C-terminal domains of Mlh1-Pms1 [46] predicted that three amino acids altered by these mutations were metal ligands in the Mlh1-Pms1 nuclease active site and the fourth was a residue adjacent to the metal binding site . Biochemical analysis of mutant proteins confirmed that both pms1 and mlh1 mutations affecting the predicted active site eliminated or significantly reduced the RFC-PCNA dependent nuclease activity of Mlh1-Pms1 and in vivo imaging showed that the same mutations resulted in accumulation of Mlh1-Pms1-4GFP foci consistent with failure to execute a downstream step in MMR . These results both define the nuclease active site of the Mlh1-Pms1 endonuclease and thoroughly characterize the role of this endonuclease activity in a previously uncharacterized Exo1-independent MMR sub-pathway .
To gain insight into the role of PMS1 in mismatch repair , we sought to generate novel pms1 mutations that cause a dominant MMR defect . First we mutagenized the PMS1 gene by PCR amplification and gap-repaired the resulting DNA fragments into the low copy number ARS CEN pRS316 plasmid by co-transformation into a S . cerevisiae strain with a wild-type PMS1 gene . A total of 38 , 000 transformants were screened for increased reversion of the lys2-10A frameshift mutation . This screen identified 211 transformants potentially containing dominant pms1 mutations . Rescreening these 211 transformants for increased reversion of the hom3-10 frameshift mutation identified 8 transformants with mutator phenotypes . The pms1 mutation-bearing plasmids were isolated from these 8 transformants , and the PMS1 gene from each plasmid was sequenced . Site directed mutagenesis and sub-cloning were used to construct PMS1 plasmids containing single point mutations . These mutant plasmids were retested in the three mutator assays confirming four dominant pms1 mutations resulting from amino acid substitutions in the C-terminal domain of Pms1: G683E , C817R , C848S , and H850R . The four amino acid substitutions resulting from the dominant mutations altered highly conserved amino acids ( Figure 1A ) . Initial analysis of these amino acid substitutions using a homology model based on the C-terminal domains of endonuclease-proficient and zinc-binding Neisseria gonorrhoeae and Bacillus subtilis MutL homologs [47] , [48] indicated that they affect the active site of the Pms1 endonuclease . Mapping of the amino acid substitutions onto the newly available structure of the C-terminal domains of Mlh1-Pms1 ( Figure 1B , C ) confirmed that the C817R , C848S , and H850R amino acid substitutions each eliminated one of the 5 metal ligands in the Pms1 endonuclease active site; all 5 ligands are conserved in eukaryotic ScPms1/HsPMS2 proteins and in the N . gonorrhoeae and B . subtilis MutL homologs [47] , [48] . The fourth amino acid substitution , G683E , mapped to a conserved position adjacent to sites of metal coordination and could sterically disrupt the site or locally perturb the structure . We also constructed mutations resulting in the amino acid substitutions H703A and E707A to eliminate the remaining two predicted metal ligands not identified in our screen . Fluctuation analysis was performed to evaluate the mutator effects of the dominant pms1 mutants . Mutation rates were measured using the CanR forward mutation assay and the hom3-10 and lys2-10A frameshift reversion assays [21] , [27] when the pms1 dominant mutations were present on low copy plasmids in a strain with a wild-type PMS1 gene ( Table 1 ) . None of the mutant plasmids caused more than a 2-fold increased mutation rate in the CanR assay , which has a relatively high background mutation rate in wild-type cells and a low sensitivity for detecting MMR defects . In contrast , the mutations on the low copy plasmid caused between a 2- and 102-fold increase in mutation rate in the hom3-10 and lys2-10A assays compared to introduction of a wild-type copy of PMS1 on a low copy plasmid . Introduction of these mutations onto a high copy 2-micron plasmid resulted in much higher rates ( Table S1 ) , but still did not cause mutation rates that were as high as caused by deletion of PMS1 ( Compare Table S1 to Table S2 ) . The pms1 mutant plasmids were also unable complement a pms1Δ strain above that seen for the vector control ( Table S2 ) . Taken together , the fluctuation analysis indicated that the dominant pms1 mutations were null PMS1 alleles that caused a copy-number dependent dominant mutator phenotype . This dominant mutator phenotype was stronger for mutations causing amino acid substitutions of metal ligands as compared to the pms1-G683E mutation . Previous studies have shown that S . cerevisiae Mlh1-Pms1 has a metal-dependent endonuclease activity that can be stimulated by RFC and PCNA [26] . To determine if the dominant pms1 mutations affecting metal ligating amino acids disrupt the endonuclease function of Mlh1-Pms1 , we expressed and purified the S . cerevisiae wild-type Mlh1-Pms1 complex and the mutant Mlh1-Pms1-G683E , Mlh1-Pms1-C817R , Mlh1-Pms1-C848S , and Mlh1-Pms1-H850R complexes and assayed the ability of these complexes to nick supercoiled pRS425 plasmid DNA with or without accessory factors RFC-Δ1N and PCNA ( Figure 2A ) . Wild-type Mlh1-Pms1 alone showed little endonuclease activity . However , addition of PCNA and RFC-Δ1N to reactions containing wild-type Mlh1-Pms1 resulted in a 20-fold increase in endonuclease activity resulting in cleavage of nearly half of the original substrate DNA . The newly identified Mlh1-Pms1 mutant proteins and the previously studied Mlh1-Pms1-E707K mutant protein did not exhibit any PCNA and RFC-Δ1N stimulated endonuclease activity , with the exception of the Mlh1-Pms1-H850R mutant protein ( Figure 2A ) . An explanation of the ability of the Mlh1-Pms1-H850R mutant protein to nick supercoiled DNA is provided in the “Discussion” . These results support the idea that loss of metal coordination by Pms1 inhibits the endonuclease activity of Mlh1-Pms1 . Exo1 is a 5′-3′ nuclease that functions in MMR in vitro by resecting DNA from a preexisting nick to a point past the mispair [28] , [44] , [49] . Loss of Exo1 function in vivo by deletion of EXO1 or missense mutations inactivating the Exo1 active site only results in a weak mutator phenotype and hence only partial loss of MMR ( Table 1 ) [27]–[29] , [50] . To determine the consequences of eliminating the two known nucleases involved in S . cerevisiae MMR , we tested the effect of introducing plasmids containing the dominant pms1 mutations into an exo1Δ mutant strain containing a wild-type PMS1 gene . Relative to the effects in a wild-type strain , introduction of the dominant pms1 mutations on a low-copy plasmid into the exo1Δ mutant strain increased the mutation rate to a much greater extent ( Table 1 ) . The pms1-G683E mutation again caused the weakest mutator phenotype , whereas the other five pms1 mutations that affect the metal ligands caused relatively high mutation rates ranging from a 57- to a 210-fold increased mutation rate depending on the mutation and the assay . These results show that the dominant pms1 mutations cause a greater MMR defect in an exo1Δ mutant strain than in a wild-type strain , suggesting that Exo1 and the Mlh1-Pms1 nuclease have a redundant function in MMR . The four C-terminal residues of Mlh1 are almost completely conserved as the amino acids FERC in fungal and animal species ( Figure 1D ) . Furthermore , Mlh1 has no C-terminal extension beyond the FERC residues in almost all sequenced eukaryotic organisms except nematodes ( FERCG[T/S] ) whereas the ScPms1/HsPMS2 family of proteins have variable length C-terminal extensions ( Figure 1D ) . Consistent with a special role for the C-terminus of Mlh1 , C-terminal fusions to the S . cerevisiae Mlh1 are non-functional for MMR in vivo , whereas C-terminal fusions to S . cerevisiae Pms1 do not affect function [38] . A structure of this highly conserved Mlh1 C-terminus ( Figure 1B , C ) revealed that this region might be appropriately positioned to play a role at the endonuclease active site of Pms1 , with the C-terminal Mlh1 cysteine potentially acting as a metal ligand . We therefore generated the mlh1-E767stp , mlh1-C769stp , mlh1-C769A and mlh1-C769S mutations to probe the role of these highly conserved residues in endonuclease activity in vitro and MMR in vivo . Unlike the pms1 mutations affecting metal ligands , none of the mutations affecting the C of the conserved FERC motif of Mlh1 that is predicted to be a metal ligand including the mlh1-C769A , mlh1-C769S , and mlh1-C769stp mutations caused a dominant mutator phenotype when present on an ARS CEN plasmid in a wild-type strain or an exo1Δ strain ( Table 2 ) . The mlh1-C769A , mlh1-C769S , and mlh1-C769stp mutants fully complimented the MMR defect of an mlh1Δ strain ( Table S2 ) consistent with previously published results for the mlh1-C769A mutation but not the mlh1-C769S mutation [51] or the mlh1-C769stp mutation [46] . In this regard , it should be noted that our studies used a broader series of mutator assays , including more sensitive assays , than previous studies of mutations affecting Mlh1-C767 . In contrast , the mlh1-E767stp mutant plasmid failed to complement the MMR defect of an mlh1Δ strain and resulted in a null phenotype ( Table S2 ) . Furthermore , the mlh1-E767stp mutation on an ARS CEN plasmid caused a weak dominant mutator phenotype when present in a wild-type strain and a stronger dominant mutator phenotype when present in an exo1Δ strain ( Table 2 ) , although not to the extent as that caused by the PMS1 metal ligand mutations . We also tested the effect of the mlh1-C769stp and mlh1-E767stp mutations in the endonuclease assay ( Figure 2B ) . The mutant Mlh1-Pms1 protein lacking only the Mlh1 C-terminal cysteine that did not cause an MMR defect in vivo nicked supercoiled DNA to the same extent as the wild-type Mlh1-Pms1 protein . In contrast , the mutant Mlh1-Pms1 protein resulting from the mlh1-E767stp mutation was significantly defective for nicking supercoiled plasmid DNA , which parallels the effect of this mutation on MMR in vivo . These results support the idea that the C-terminus of Mlh1 functions in the endonuclease active site although if the terminal cysteine coordinates bound metal then this role is not required for MMR or endonuclease activity ( Figure 1C ) . Mutations affecting the active sites of leading and lagging strand DNA polymerases Pol ε , pol2-M644G , and Pol δ , pol3-L612M , have been identified that preferentially introduce misincorporation errors in their respective strand during DNA replication [52] , [53] . In a wild-type background , these lesions are then efficiently corrected by MMR . This strand-biased misincorporation can be used to determine strand preferences for MMR [52] , [54] , [55] . Here we probed mutants containing these polymerase active site mutations with ARS CEN plasmids encoding endonuclease defective pms1 and mlh1 mutations to investigate whether the Mlh1-Pms1 endonuclease preferentially functions in leading or lagging strand MMR . We found that ARS CEN PMS1 plasmids containing the pms1-C817R , pms1-C848S or pms1-H850R mutations caused a statistically similar synergistic increase in mutation rate when present in strains containing mutations affecting either DNA polymerase ( Table 3 ) . Of nine pairwise comparisons between pol2-M644G and pol3-L612M mutants containing the same pms1 mutation on a low copy plasmid using three different mutation rate assays , seven were not different ( p-value>0 . 05 , Mann-Whitney test ) . For the two comparisons that showed a difference , one showed a modestly higher rate in the pol2-M644G strain while the other showed a modestly higher rate in the pol3-L612M strain ( p-value<0 . 05 , Mann-Whitney test ) . This degree of similarity between the pol2-M644G and pol3-L612M mutants in these comparisons is in marked contrast to the effect of an exo1Δ mutation that caused a 9-fold higher increase in mutation rate in a pol3-L612M mutant compared to a pol2-M644G mutant [38] . Overall , these results suggest that Mlh1-Pms1 functions similarly on both the leading and lagging strands during MMR . We have previously demonstrated that Mlh1-Pms1 foci are an intermediate in MMR and that blocking MMR downstream of Mlh1-Pms1 recruitment results in increased levels of Mlh1-Pms1 foci [38] . To test the effect of the Mlh1-Pms1 active site mutations on the levels of Pms1 foci , different mlh1 and pms1 mutations were introduced into the relevant endogenous locus in a strain in which the single wild-type copy of PMS1 was functionally tagged with four tandem copies of GFP . Normally , Mlh1-Pms1-4GFP foci are present in approximately 10% of logarithmically growing wild-type cells whereas all of the mutations that affected endonuclease function in vitro that were tested caused increased Mlh1-Pms1-4GFP foci formation ( Figure 3 ) . These included the metal coordination pms1 mutations pms1-E707K , which was previously tested [38] , as well as pms1-C817R , pms1-C848S and pms1-H850R , the pms1-G683E mutation and the mlh1-E767stp endonuclease active site mutation , which increased the proportion of cells containing Mlh1-Pms1-4GFP foci to between 64 and 96% . In contrast , the endonuclease and MMR proficient mlh1-C769stp mutation did not alter the levels of Mlh1-Pms1-4GFP foci compared to wild-type strain . These results are consistent with the idea that mispairs are recognized normally in these Mlh1-Pms1 endonuclease active site mutants and that there is proper loading of the mutant Mlh1-Pms1 on DNA but instead there is a mispair processing defect resulting in decreased turnover of the mutant Mlh1-Pms1 from the DNA [38] .
In this study , we used the highly sensitive lys2-10A frameshift mutation reversion assay to screen mutagenized low copy PMS1 plasmids for pms1 mutations that caused a dominant mutator phenotype in the presence of a single-copy of the wild-type PMS1 gene . We identified four null mutations that caused a weak dominant phenotype under these conditions . These mutations all caused amino acid substitutions in or near the region predicted to contain the Pms1 endonuclease active site and three of the amino acid substitutions including Pms1-C817R , Pms1-C848S , and Pms1-H850R affected predicted metal binding motifs while Pms1-G683 was in close proximity to the metal coordination site . The weak effect of these mutations explains why a prior study of a predicted Pms1 active site mutation did not observe a dominant effect [56] . A model of the endonuclease active site predicted Pms1 amino acids H703 and E707 as well as Mlh1-C769 to also be a part of the Mlh1-Pms1 active site [46] . Consistent with this model , the pms1-H703A and pms1-H707A mutations were found to cause the same phenotypes as the other PMS1 metal ligand mutations . In contrast , no mutation affecting Mlh1-C769 caused either a dominant mutator phenotype or affected Mlh1 function whereas the mlh1-E767stp mutation caused phenotypes that were similar to those caused by the PMS1 metal ligand mutations . Remarkably , all of the pms1 mutations caused a stronger dominant mutator phenotype when present in an exo1Δ strain on a low copy plasmid . The mlh1-E767stp also caused an increased dominant mutator affect in the exo1Δ strain , although not to the extent seen with the pms1 mutations . The phenotype of these mutants is similar to previously described separation-of-function mutations in MSH2 , MSH3 , MSH6 , MLH1 , PMS1 , POL30 and POL32 that cause strong defects in Exo1-independent MMR but little if any defect in MMR when Exo1 is functional [27] . The asymmetry of the endonuclease active site is consistent with proposed roles of Mlh1-Pms1 in nicking double-stranded DNA during MMR; however , this asymmetry is not present in homodimeric bacterial MutL homologs with endonuclease function [47] , [48] . The similarity of the eukaryotic Mlh1-Pms1 and bacterial MutL homologs suggests that MutL-DNA complexes may be functionally asymmetric so that only one active site is positioned to cleave DNA , analogous to the functional asymmetry during mispair recognition by bacterial MutS homodimers [15] , [16] . The asymmetry of the eukaryotic MutL complexes , however , allows specialization of each subunit . The highly conserved C-terminus of Mlh1 is positioned in the Mlh1-Pms1 structure in a way that suggests that the Mlh1-ScPms1/HsPMS2 and Mlh1-Mlh3 complexes have a composite endonuclease active site . Potential roles for residues in the Mlh1 C-terminus include coordinating DNA phosphates , promoting nucleophilic attack by a water molecule or stabilization of the Pms1 active site . Consistent with this , the mlh1-E767stp mutant plasmid did not complement the mutator phenotype caused by a deletion of MLH1 , and the mlh1-E767stp mutation resulted in the accumulation of Mlh1-Pms1-4GFP foci and reduced endonuclease activity similar to mutations in PMS1 affecting the endonuclease active site . It was surprising that mutation or deletion of the highly conserved C-terminal cysteine did not cause a MMR defect in vivo or reduce endonuclease activity in vitro given the possibility that this residue might coordinate with metals in the endonuclease active site . It is possible that conservation of the C-terminal cysteine may reflect other roles for Mlh1 , potentially including crossover resolution during meiosis [57]–[59] . Analysis of the genetically identified and structure-based mutations in the MLH1 and PMS1 genes revealed that disruption of the metal binding sites leads to disruption of the Mlh1-Pms1 endonuclease activity and arrest of MMR repair at a step following recruitment of Mlh1-Pms1 into microscopically-observable foci . In a Mn2+- , RFC- , and PCNA-dependent endonuclease assay , amino acid substitutions affecting all of the metal ligands caused defects in the Mlh1-Pms1 endonuclease activity , which confirms and extends previous studies of the human Mlh1-Pms2 E705K amino acid substitution [25] , [26] . The pms1-H850R and the mlh1-E767stp mutations resulted in proteins with partial defects in the in vitro endonuclease assay , but caused complete MMR defects in vivo . The partial endonuclease defect caused by these mutations may reflect the fact that the in vivo metal ion in eukaryotic and bacterial homologs is Zn2+ , which has tetrahedral coordination geometry , whereas the metal that promotes the in vitro assay is Mn2+ , which prefers an octahedral coordination geometry [25] , [48] , [60]–[63] . All of the pms1 mutations that affect predicted metal binding ligands , as well as mlh1-E767stp , caused complete MMR defects in vivo , consistent with the observation that metal ligand defects in human Mlh1-Pms2 inactivate MMR in vitro [26] , [63] . These mutations also caused an accumulation of Mlh1-Pms1-4GFP foci indicating the step at which these mutations disrupt MMR is after loading of Mlh1-Pms1 by Msh2-Msh6 , suggesting that loss of endonuclease activity leads to a turnover defect of Mlh1-Pms1 during MMR . A striking property of the dominant pms1 and mlh1 mutations is that when present on a low copy plasmid they cause a much greater defect in Exo1-independent MMR compared to MMR when Exo1 is functional even though Mlh1-Pms1 appears to be absolutely required for all MMR . This phenotype is similar to the phenotype caused by the previously described separation-of-function mutations in genes like MSH2 , MSH3 , MSH6 , MLH1 , PMS1 , POL30 and POL32 that result in strong defects in Exo1-independent MMR but little if any defect in MMR when Exo1 is functional [27] . A hypothesis that could explain these observations is that Mlh1-Pms1 has two roles in MMR , one involving activation of the Mlh1-Pms1 endonuclease and one where Mlh1-Pms1 plays a role in the recruitment of downstream MMR factors . It was previously shown that Exo1 interacts with Mlh1 and that this interaction is required for Exo1 to function in MMR [50] . This suggests the possibility that mispair recognition by Msh2-Msh6 or Msh2-Msh3 recruits Mlh1-Pms1 which then targets Exo1 to DNA where it could promote excision at pre-existing nicks in the DNA , consistent with the observation that lagging strand MMR is more Exo1 dependent than leading strand MMR [38] , [42] . Such a reaction would be expected to be relatively insensitive to inhibition by competition with an endonuclease inactive but structurally normal form of Mlh1-Pms1 that would still bind Exo1 and target it to the site of MMR . In contrast , MMR in the absence of Exo1 might be completely dependent on the Mlh1-Pms1 endonuclease activity . This reaction would be expected to be competed for and interfered with by the presence of endonuclease inactive but structurally normal form of Mlh1-Pms1 . A model that summarizes these concepts is presented in Figure 4 .
S . cerevisiae cells were grown in YEPD ( 1% yeast extract , 2% Bacto peptone and 2% dextrose with or without 2% Bacto agar ) or SD ( 0 . 67% yeast nitrogen base and 2% dextrose with or without 2% Bacto agar ) medium . SD medium was supplemented with the appropriate dropout mix of amino acids ( USA Biological ) . The S . cerevisiae strains used in genetic experiments were derived from an S288c parental strain and the strains used for protein purification were derived from RDKY1293 or RDKY8053 ( listed in Table S3 ) . All strains were constructed using standard gene disruption and transformation procedures . E . coli strains were propagated in LB media ( 0 . 5% yeast extract , 1% tryptone , 0 . 5% NaCl , 50 µg/ml thymine with or without 2% Bacto agar ) containing 100 µg/ml ampicillin as required . All plasmids ( listed in Table S4 ) were maintained in E . coli TOP 10F′ . A pRS316 Ampr URA3 ARS-CEN PMS1 plasmid pRDK1667 was constructed by recombination in vivo . Briefly , PMS1 was amplified from S . cerevisiae S288c chromosomal DNA using the primers 5′ACGACGGCCAGTGAATTGTAATACGACTCACTATAGGGCGAATTGGAGCTattgccaaacaggcaaagac that contains 50 bp of homology to pRS316 upstream of the multiple cloning site followed by 20 bp of homology to the PMS1 genes 707 bp upstream of the promoter starting at chromosome XIV coordinate 472684 and 5′TTAACCCTCACTAAAGGGAACAAAAGCTGGGTACCGGGCCCCCCCTCGAGgcatacaagaaacaacgcga that contains 50 bp of homology to pRS316 encompassing the XhoI , DraII , ApaI and KpnI sites of the multiple cloning site followed by 20 bp at the 3′ end with homology to the PMS1 genes 302 bp downstream of the stop codon from chromosome XIV coordinate 476314 . The PCR product was mixed with an equimolar amount of pRS316 that had been linearized by digestion with SmaI and co-transformed into the wild-type S . cerevisiae strain RDKY3590 . The transformants were selected on SC-uracil drop out plates , DNA was isolated from individual transformants , rescued by transformation into E . coli and sequenced . The plasmid selected for further use has a silent C3055T mutation . A pRS426 Ampr URA3 2-micron PMS1 plasmid , pRDK1689 , was constructed by subcloning the XhoI to StuI ( StuI cuts in the URA3 gene ) PMS1 fragment from pRDK1667 into the XhoI to StuI backbone of pRS426 . The pRS316 Ampr URA3 ARS-CEN MLH1 plasmid pRDK1338 was from our laboratory collection and contains a SacI to XhoI MLH1 fragment inserted between the SacI and XhoI sites of pRS316 . The MLH1 fragment starts at the native SacI site 2281 bp upstream of the MLH1 ATG and ends at an XhoI site inserted by PCR 121 bp downstream of the MLH1 stop codon . The 2-micron Mlh1 and Pms1 over-expression plasmids pRDK573 TRP1 GAL10-MLH1 and pRDK1099 LEU2 GAL10-PMS1-FLAG have been described previously [64] . Mutations were made in these plasmids using standard site-directed mutagenesis methods or by subcloning from a mutant plasmid and the resulting plasmids were verified by DNA sequencing . The PMS1 mutant alleles E707K , H850R , C848S , G683E and C817R were introduced at the chromosomal locus using standard pop in/out techniques employing the integrative plasmids listed in Table S4 . These integration plasmids were generated by subcloning the XhoI-StuI fragment containing the pms1 mutant sequence from their respective pRS316-pms1 mutant series plasmids into the XhoI-StuI sites of the pRS306 backbone and were linearized with BlpI prior to transformation for integration into the strains of interest . The MLH1 mutant alleles E767stp and C769stp were introduced at the chromosomal locus using standard gene disruption employing an HPH disruption cassette generated by PCR such the upstream homology targeted the C-terminus of MLH1 and contained the mutations needed to introduce the E767stp and C769stp alleles . All of the chromosomal pms1 and mlh1 mutations were verified by sequencing the entire PMS1 or MLH1 gene as relevant , which also ensured that no additional mutations were introduced during strain construction . Mutagenesis of the PMS1 gene by PCR was performed essentially as previously described [65] with the following modifications . The primers used for PCR were those described above for amplification of PMS1 . Ten PCR reactions were performed using Klentaq DNA polymerase and a PMS1 gene containing plasmid pRDK433 from our laboratory collection as a template . The PCR reactions were pooled , aliquots of DNA were mixed with an equimolar amount of pRS316 that had been linearized by digestion with SmaI and co-transformed into the wild-type S . cerevisiae strain RDKY3590 . The transformants were plated on SC-uracil drop out plates to select for transformants , which were then replica plated onto SC-uracil-lysine drop out plates to screen for colonies that had increased rates of reversion of the lys2-A10 frameshift mutation . Candidate mutator mutants were retrieved from the uracil drop out plates , restreaked on SC-uracil drop out plates , patched in duplicate onto uracil drop out plates and replica plated onto threonine-uracil drop out plates to screen for patches that had increased rates of reversion of the hom3-10 frameshift mutation . Plasmid DNA was isolated from each mutator mutant , transformed into E . coli TOP10F′ and sequenced . Individual mutations identified were then transferred to a new pRS316 PMS1 plasmid pRDK1667 by either sub-cloning using appropriate restriction endonuclease cleavage sites or by site-directed mutagenesis and retested essentially as described for the initial screen above . S . cerevisiae Mlh1-Pms1 was purified from 2 . 2 L of culture of the overproduction strain RDKY7608 ( RDKY1293 containing the 2-micron plasmids pRDK573 TRP1 GAL10-MLH1 and pRDK1099 LEU2 GAL10pr-PMS1-FLAG ) ( Table S3 ) according to a previously published procedure [64] , except with the following 6 modifications: ( 1 ) Cell growth and induction of Mlh1-Pms1 expression utilized a published lactate to galactose shift protocol according to previously published methods [66]; ( 2 ) 2 mM β-mercaptoethanol was substituted for the 1 mM DTT in the buffers used to run the Heparin and FLAG antibody columns whereas all other buffers contained 1 mM DTT; ( 3 ) After washing the Heparin column with Buffer A containing 200 mM NaCl , the proteins were eluted using a single step of 1 M NaCl in Buffer A; ( 4 ) The pooled Heparin column fractions were diluted with Buffer A to obtain a final NaCl concentration of 500 mM prior to being subjected to 3 cycles of binding and elution from the FLAG antibody column; ( 5 ) The SP Sepharose column fractions were diluted with Buffer A to a final NaCl concentration of 200 mM prior to being loaded onto a 1 ml HiTrap Q column ( GE Healthcare ) followed by elution with a 100 mM to 1 M linear NaCl gradient run in Buffer A; and ( 6 ) The HiTrap Q column fractions containing the Mlh1-Pms1 were concentrated and desalted using a Centraprep ( Ultracel 30K ) spin column . The resulting Mlh1-Pms1 was contained in 0 . 5 ml of Buffer A +100 mM NaCl , and was frozen in liquid nitrogen and stored at −80 C . The Mlh1-Pms1-E707K , Mlh1-Pms1-C817R , Mlh1-Pms1-C848S and Mlh1-Pms1-H850R proteins were purified using the overproduction strains RDKY7696 , RDKY7756 , RDKY7759 and RDKY7793 ( Table S3 ) . The Mlh1-C769stp-Pms1 and Mlh1-E767stp-Pms1 proteins were purified using the overproduction strains RDKY8055 and RDKY8057 for which the RDKY8053 host strain ( Table S3 ) was a derivative of RDKY1293 containing a deletion of the MLH1 gene . Because mlh1-E767stp allele does not compliment the MLH1 deletion in the host , after the protein expression period , DNA was isolated from the culture , 20 independent MLH1 and PMS1 plasmids were rescued by transformation into E . coli and sequenced to ensure that no mutations had occurred in the expression plasmids . S . cerevisiae PCNA and RFC-Δ1N were purified exactly as described in published procedures [66]–[68] . All of the protein preparations used in these studies were greater than 98% pure as analyzed by SDS-PAGE . Mismatch-independent endonuclease assays were performed as a modification of one used previously [26] . 40 µL reactions containing 1 mM MnSO4 , 20 mM Tris pH 7 . 5 , 0 . 5 mM ATP 0 . 2 mg/mL bovine serum albumin ( BSA ) , 2 mM DTT and 100 ng pRS425 were incubated at 30°C for 30 minutes . Reactions were terminated by incubation at 55°C following introduction of SDS , EDTA , glycerol and proteinase K at concentrations of 0 . 1% , 14 mM , 8% and 0 . 5 ug/ml respectively . Mlh1-Pms1 , PCNA , or RFC-Δ1N were diluted to the appropriate working concentrations with a buffer comprised of 10% glycerol , 200 mM NaCl , 2 mM DTT and 20 mM Tris pH 7 . 5 . Following termination of the reaction the samples were electrophoresed on a 0 . 8% agarose gel , the gel was stained with ethidium bromide , extensively destained and then the bands were quantified using a BioRad ChemiDoc XP imaging system . Serial dilutions of XhoI linearized pRS425 ranging from 10–100 ng were used as a concentration standard for quantification . Mutation rates were determined by fluctuation analysis . A single colony was used to inoculate a culture that was then diluted and used for transformation with a selectable plasmid carrying the desired allele and transformed colonies were selected by growth for 3 days at 30°C on SC-uracil dropout plates . 7 independent colonies were used to inoculate individual overnight cultures containing 10 ml of SC-uracil dropout media . Following cell growth , appropriate dilutions of the cultures were plated onto SC –uracil , –uracil-lysine , -uracil-threonine , and -uracil-arginine+canavanine dropout plates . The resulting colonies counted after growth at 30°C for 3 days and the average mutation rate was calculated for each strain as described previously [21] , [27] . Each experiment was performed independently up to 4 times . Site-directed mutagenesis was guided by a molecular model of the C-terminal domains of Mlh1-Pms1 . The C-terminus of S . cerevisiae Mlh1 was modeled using Phyre , xfit , and CNS [69]–[71] starting from the crystal structure of the human Mlh1 C-terminal domain ( PDB id 3rbn ) . The C-terminus of S . cerevisiae Pms1 was similarly modeled using the crystal structures of N . gonorrhoeae [PDB id 3ncv; [47]] and B . subtilis [PDB ids 3gab , 3kdg , 3kdk; [48]] MutL homologs . Subsequently , the amino acid substitutions studied were mapped onto the newly available structure of the C-terminal domains of S . cerevisiae Mlh1-Pms1 [PDB id 4e4w; [46]] . For microscopy studies , the C-terminus of each PMS1 protein of interest was fluorescently tagged by targeting a 4GFP tag to the chromosomal locus so that the native promoter was intact and expression remained unaffected . Previous analysis of the tagged PMS1 gene demonstrated that the 4GFP tag did not affect the biological activity of Pms1 [38] . Exponentially growing cultures were washed and resuspended in water , placed on minimal media agar pads , covered with a coverslip , and imaged on a Deltavision ( Applied Precision ) microscope with an Olympus 100× 1 . 35NA objective . Fourteen 0 . 5 µm z sections were acquired and deconvolved with softWoRx software . Further image processing , including maximum intensity projections and intensity measurements were performed using ImageJ . | Lynch syndrome ( hereditary nonpolypsis colorectal cancer or HNPCC ) is a common cancer predisposition syndrome . Predisposition to cancer in this syndrome results from increased accumulation of mutations due to defective mismatch repair ( MMR ) caused by a mutation in one of the mismatch repair genes MLH1 , MSH2 , MSH6 or PMS2/scPMS1 . In addition to these genes , various DNA replication factors and the excision factor EXO1 function in the repair of damaged DNA by the MMR pathway . Although EXO1 is considered to be the major repair nuclease functioning in mismatch repair , the relatively low mutation rates caused by an exo1 deletion suggest otherwise . Here we used genetics , microscopy and protein biochemistry to analyze the model organism Saccharomyces cerevisiae to further characterize a poorly understood mismatch repair pathway that functions in the absence of EXO1 that is highly dependent on the Mlh1-Pms1 complex . Surprisingly , we found that the highly conserved metal binding site that is critical for the endonuclease activity of the Mlh1-Pms1 heterodimer is required for MMR in the absence of Exo1 to a much greater extent than in the presence of Exo1 . Thus , this work establishes that there are at least two different polynucleotide excision pathways that function in MMR . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [] | 2013 | Dominant Mutations in S. cerevisiae PMS1 Identify the Mlh1-Pms1 Endonuclease Active Site and an Exonuclease 1-Independent Mismatch Repair Pathway |
The immunization with genetically attenuated Leishmania cell lines has been associated to the induction of memory and effector T cell responses against Leishmania able to control subsequent challenges . A Leishmania infantum null mutant for the HSP70-II genes has been described , possessing a non-virulent phenotype . The L . infantum attenuated parasites ( LiΔHSP70-II ) were inoculated in BALB/c ( intravenously and subcutaneously ) and C57BL/6 ( subcutaneously ) mice . An asymptomatic infection was generated and parasites diminished progressively to become undetectable in most of the analyzed organs . However , inoculation resulted in the long-term induction of parasite specific IFN-γ responses able to control the disease caused by a challenge of L . major infective promastigotes . BALB/c susceptible mice showed very low lesion development and a drastic decrease in parasite burdens in the lymph nodes draining the site of infection and internal organs . C57BL/6 mice did not show clinical manifestation of disease , correlated to the rapid migration of Leishmania specific IFN-γ producing T cells to the site of infection . Inoculation of the LiΔHSP70-II attenuated line activates mammalian immune system for inducing moderate pro-inflammatory responses . These responses are able to confer long-term protection in mice against the infection of L . major virulent parasites .
Leishmaniases are a group of vector-borne diseases caused by the transmission of the protozoan parasite Leishmania in different mammalian hosts , during the blood meal of the invertebrate vectors ( phlebotomine sandflies ) . Depending on the species of the parasite and the immune response of the host , the disease outcome varies from asymptomatic infections to clinical forms of the disease . The cutaneous forms of the disease ( cutaneous leishmaniasis; CL ) are characterized by the generation of disfiguring skin ulcers . In the Old World , is caused , among others , by the infection of Leishmania major species and together with the other forms of leishmaniasis is included in the list of neglected tropical diseases , affecting various developing countries [1] . Many efforts have been made in terms of prevention of leishmaniasis in the last decades , since it is believed that a vaccine against leishmaniasis is feasible given that patients recovered from the disease become resistant to new infections . Mice infected with L . major have been widely used as experimental models for screening of vaccines . When BALB/c mice are experimentally challenged with L . major they suffer a progressive form of the disease , developing cutaneous lesions correlated to parasite multiplication at the site of infection as well as parasite dispersion to internal organs [2 , 3] . Parasite specific IL-4 driven production of antibodies as well as the development of Leishmania related IL-10 deactivating responses are correlated with susceptibility [4] . On the other hand , C57BL/6 mice experimentally infected with L . major promote T-cell dependent IFN-γ production that results in the activation of infected macrophages to produce nitric oxide and to destroy the intracellular parasites [4] . Control of parasite mediated inflammation by regulatory T cells results in parasite persistent infection and resistance to reinfection [5] . Despite the fact that memory T cells can persist after parasite control [6] , healed C57BL/6 mice lose their immunity to reinfection if they are manipulated to clear completely the parasites [7] . This has been taken as an evidence that persistence of live parasites is inevitably necessary for the maintenance of long-term protection [8] . In this context , the use of live attenuated parasites as vaccine candidates is a promising field of research . Live vaccines can induce adaptive immune responses relevant to protection by mimicking natural infection , without the adverse effects of leishmanization with virulent parasites . Heat Shock Protein 70 ( HSP70 ) plays a central role in both prokaryotic and eukaryotic cells because of its involvement in different aspects of protein metabolism ( folding , assembly , activation , subcellular location , and so on ) influencing many aspects of the cell biology , like cell growth and differentiation [9] . In Leishmania , the HSP70 also plays important roles in particular aspects affecting host-parasite interaction like virulence , drug resistance as well as in the induction of host immune responses ( reviewed in [10] ) . There are two types of genes encoding HSP70 in Leishmania infantum ( = L . chagasi ) , the causative agent of canine and human VL in the Mediterranean countries and in South America . Differences in their 3’ untranslated region ( UTR ) sequences have a great importance in the regulation of Lihsp70 gene expression [11] . Since mRNAs having the 3’UTR-II are preferentially translated at 37°C [12] , the expression of Lihsp70-II gene has been related to the response against the thermal stress caused by the parasite entry in the vertebrate host . Genetic elimination of the Lihsp70-II alleles resulted in a knock-out parasite line ( LiΔHSP70-II ) presenting a pleiotropic effect , influencing cell morphology , replication and , of special interest , virulence [13] . Hence , promastigotes ( the form found in the insect vector ) of the mutant line present some growth deficiencies in culture , and amastigotes show a limited capacity of multiplication inside macrophages , although Lihsp70-II gene deletion did not alter parasites uptake by these host cells . The inoculation of the mutant line did not produce any pathology in either hamster ( highly susceptible for L . infantum infection ) or in immune-deficient SCID mice , even though specific cellular responses were observed [13 , 14] . In addition , the mutant LiΔHSP70-II was able to induce a short-term protection against L . major infection in BALB/c mice [14] . In this work , we have extended the study of these protective capacities analyzing short- and long-time protection after intravenous ( i . v . ) or subcutaneous ( s . c . ) infection with the LiΔHSP70-II line in the BALB/c-L . major model of progressive leishmaniasis . The immune correlates of protection have been also analyzed . The studies regarding the prophylactic properties of the LiΔHSP70-II line administration have been extended to the L . major infection resistant C57BL/6 mouse model . The vaccine-mediated robust protection shown in this line has been associated to the rapid recruitment of pre-existing CD4+ and CD8+ IFN-γ producing T cells to the site of L . major challenge .
In a previous work , it was described that i . v . infection with the LiΔHSP70-II attenuated parasites ( vaccination ) resulted in short-term protection against L . major infection when challenged four weeks after vaccination [14] . This protection was correlated to the presence of the attenuated parasites in the liver and spleen of the protected animals [14] . Here , we firstly analyzed the evolution of the attenuated parasites in the internal organs of the i . v . vaccinated mice for a longer period of time . Parasites detected at week 4 after vaccination ( S1 Fig panel A ) [14] were undetectable in the spleen or liver at week 12 after vaccination ( Fig 1A ) . However , in the bone marrow ( BM ) , although at week 12 post vaccination parasite burden significantly decreased compared to the 4th week ( P < 0 . 05; unpaired T-test ) , we still detected parasites in 4 out of 8 mice ( Fig 1A ) . Vaccination induced a Leishmania-specific cellular response that was revealed after stimulation with soluble leishmanial antigen ( SLA ) of spleen cells . We detected SLA-specific secretion of IFN-γ and IL-10 to a lesser extent , both at the 4th ( S1 Fig panel B ) or at the 12th ( Fig 1B ) week post-vaccination . Alternatively , we administered s . c . the LiΔHSP70-II based vaccine to analyze its immunogenicity in a different vaccination setting . No parasites were observed in liver , spleen and BM of these animals at any time . Parasites were detected in the right popliteal which corresponds to the draining lymph node ( DLN ) close to the inoculation site ( Fig 1C ) . Although parasite numbers decreased over time ( P = 0 . 003 ) ( Fig 1C ) , all animals presented live parasites in the DLN at week 12 . The analysis of the parasite burden in the site of vaccination ( right footpad ) revealed the presence of live parasites in all the animals at week 4 after vaccination , that decreased significantly at week 12 ( P = 0 . 0013 ) . At this time , parasites were only detected in two out of eight vaccinated animals ( Fig 1C ) . Interestingly , although only a local infection occurred , the immune response detected in the spleen was similar in profile and magnitude to that found in the i . v . vaccinated animals , with a predominant SLA-specific production of IFN-γ , which was more prominent at 12 weeks after vaccination ( Fig 1D ) . The analysis of the parasite-specific production of cytokines by cells derived from the DLN ( right popliteal ) showed also an IFN-γ predominant response , higher in magnitude at the short-term ( Fig 1E ) , coinciding with the presence of high numbers of the attenuated parasites ( Fig 1C ) . In addition , at week 4 after vaccination , IL-10 and IL-4 were detected in SLA-stimulated cultures ( Fig 1E ) . Regarding the humoral response elicited by vaccination , i . v . inoculated mice showed a mixed IgG response at week 4 ( S1 Fig panel C and [14] ) and at week 12 after vaccination ( Fig 1F ) , with titers that decreased over time and were predominantly of the IgG1 isotype rather than IgG2a . Very low levels of anti-SLA IgG1 and IgG2a levels were detected in the sera of s . c . vaccinated animals , especially at long-term ( Fig 1F ) . Further , we explored the percentages of T cell populations in the spleen from control and vaccinated animals by flow cytometry . Two subsets of helper T cells ( CD3+CD4+ ) were characterized according to the presence of CD44 and CD62L molecules . All vaccinated groups showed an increase in the percentage of antigen-experienced CD4+ cells ( CD44high ) compared with the control ( Saline ) ( Fig 2A ) . Comparison between inoculation routes indicates that similar levels of CD4+ central memory T cells ( Tcm; CD44highCD62Lhigh ) were found in both groups . On the other hand , i . v . immunization elicited further expansion of CD4+ effector memory ( Tem ) or effector ( Teff ) T cells ( CD44highCD62Llow ) compared with s . c . route ( Fig 2A; S2 Fig panel A ) . Also , we determined that i . v . and s . c . vaccinated mice exhibited a higher frequency of both CD4+ and CD8+ IFN-γ producing splenic T cells compared to the unvaccinated group after in vitro stimulation with anti-CD3/anti-CD28 antibodies ( Fig 2B; S2 Fig panel C ) . This increment was also observed in the DLN ( right popliteal ) of the s . c . group ( Fig 2C; S2 Fig panel D ) . Altogether , these data allowed concluding that inoculation of the LiΔHSP70-II attenuated line , independently of the inoculation route , caused a persistent regressive infection that resulted in the induction of Tcm and Tem/Teff cell responses . Vaccinated animals showed a parasite dependent production of IFN-γ in which CD4+ and CD8+ T cells seem to be involved . To analyze the effect of the live vaccine on the development of a progressive leishmaniasis , BALB/c mice vaccinated with the LiΔHSP70-II line administered i . v . were challenged with L . major parasites ( 5 × 104 stationary phase promastigotes ) s . c . in the left footpad . As a control , mice inoculated with PBS at the time of the vaccination were also infected with L . major . Infective challenge was performed short- ( 4 weeks ) or long-term ( 12 weeks ) after vaccination . Fig 3A shows that i . v . vaccination also induced long-term protection , since very low footpad swelling was observed in the vaccinated groups , as it was reported for the short-term [14] and confirmed in this work ( S1 Fig panel D ) . The lack of lesions correlated to a decrease in L . major parasite burdens relative to saline controls in all analyzed organs ( Fig 3B ) . Regarding the presence of the parasite in visceral organs , a significant decrease was observed in both vaccinated groups with respect unvaccinated controls ( P < 0 . 05 ) . In the 4 weeks group , 50% ( 4/8 ) and 37 . 5% ( 3/8 ) of the mice had undetectable parasites in the spleen or liver , respectively ( S1 Fig panel E ) . In the 12 weeks group the percentage of negative mice reached values of 87 . 5% ( 7/8 ) in both organs ( Fig 3B ) . Data comparison among control and vaccinated mice groups revealed a higher decrease in parasite loads at the DLN for long-term infected mice ( 12 weeks; 2 . 4-log reduction; P < 0 . 001 ) ( Fig 3B ) than in the short-term group ( 4 weeks 1 . 2-log reduction; P < 0 . 01 ) ( S1 Fig panel E ) . A decrease in the evolution of footpad swelling was also observed in mice s . c . vaccinated compared to control mice ( saline group ) ( Fig 3C ) . Although no significant differences were found in the cutaneous lesions developed between mice of the 4 weeks and 12 weeks groups , short-term infected mice showed a progressive evolution of the footpad swelling evident from week 7 to week 8 . At that time , mice from all groups were euthanized because of the appearance of necrotic lesions in some mice of the control group . The clinical lesions evolution can be taken as an indication of a partial short-term protection driven by the s . c . inoculation of the LiΔHSP70-II line that was improved long-term . Determination of L . major parasite burdens in the spleen and liver ( Fig 3D ) also demonstrated a significant decrease compared to control animals in both s . c . vaccinated groups ( P < 0 . 05 and P < 0 . 001 , for 4 weeks and 12 weeks , respectively ) . In this case , most of the mice of the 4 weeks group were positive for L . major parasites in the spleen ( 75% , 6/8 ) or liver ( 87 . 5% , 7/8 ) , supporting the partial protection concluded from the clinical data . On the other hand , only two mice from the long-term protected groups were positive for live parasites in both internal organs ( Fig 3D ) . Regarding parasite loads in the left popliteal DLNs , a significant decrease ( P < 0 . 05 and P < 0 . 01 , for 4 weeks and 12 weeks respectively ) was obtained when vaccinated mice were compared to control mice . Similar decreased values with respect saline controls were found in the parasite numbers for both vaccinated groups in the liver and in the spleen ( 1-Log and 1 . 3-Log for 4 weeks and 12 weeks groups , respectively ) ( Fig 3D ) . On the other hand , we evaluated the presence of the LiΔHSP70-II attenuated parasites in different organs and tissues of the vaccinated mice after L . major challenge ( S3 Fig panel A ) . These analyses indicated that challenge with infective parasites did not reactivate the infection of the attenuated line . For i . v . vaccinated mice ( S3 Fig panel B ) no LiΔHSP70-II parasites were found in the internal organs , except parasites detected in the BM at short-term that were equivalent to those determined in the 12 wk vaccinated group ( Fig 1A ) . Interestingly , BM became negative for LiΔHSP70-II parasites at week 20 ( S3 Fig panel B ) . In addition , no attenuated parasites were found in the left popliteal LNs , in spite of the presence of L . major . Regarding the s . c . vaccinated mice , we only observed the persistent presence of the LiΔHSP70-II parasites in the LN draining the site of vaccination at week 20 ( S3 Fig panel C ) . Once we determined that vaccination with LiΔHSP70-II parasites induced protection for both , clinical manifestations and parasitemia , we analyzed the immune correlates of protection . For that , we next determined humoral and cellular responses specific for the parasite using SLA in ELISA assays and for cell stimulation in all vaccinated groups and their corresponding saline controls , 8 weeks after L . major challenge . Protection correlated with an IgG subclass redirection to Th1-related IgG2a subclass of SLA-specific antibodies in the vaccinated mice that were mainly of the IgG1 subclass in saline controls ( Fig 4A and 4B , for i . v . and s . c . , respectively ) . The magnitude of the IgG2a response was higher in long-term protected mice than in short-term groups ( P < 0 . 0018 and P < 0 . 023 for i . v . and s . c . , respectively ) . Cellular responses against SLA in the L . major infected mice were determined by stimulating spleen cells from mice receiving saline or the attenuated line ( 4 weeks and 12 weeks groups ) i . v . ( Fig 4C–4E ) or s . c . ( Fig 4F–4H ) . In agreement with the Th1-like profile of the humoral response , a SLA-dependent IFN-γ predominant response was found in all protected groups reaching higher P values with respect to saline control in short-term protected mice ( P = 0 . 0003 and P = 0 . 001 for i . v and s . c . groups , respectively ) than long-term groups ( P = 0 . 026 and P = 0 . 021 for i . v and s . c . groups , respectively ) . Interestingly , long-term protected mice showed a concomitant significant decrease in the IL-10 levels secreted after stimulation with parasite proteins when compared to saline controls ( P = 0 . 0006 and P = 0 . 0063 for i . v and s . c . groups , respectively ) ( Fig 4D and 4G ) . This decrease was absent in short-term protected mice . On the contrary , short-term protected group secreted higher amounts of IL-10 than control mice although only significant differences were observed in the i . v . vaccinated group ( P = 0 . 0361 ) ( Fig 4D ) . When IL-4 production was analyzed ( Fig 4E and 4H ) , a decrease in the levels of SLA-specific IL-4 in the culture supernatant was only found when saline controls were compared to long-term i . v . vaccinated mice ( P = 0 . 003 ) ( Fig 4E ) . The cytokine production specific for SLA was higher in short- than in long-term protected mice: P = 0 . 0003 , P = 0 . 0002 and P = 0 . 0062 for IFN-γ , IL-10 and IL-4 , respectively in the i . v . group ( Fig 4C , 4D and 4E ) and P = 0 . 022 and P = 0 . 0072 for IFN-γ and IL-10 , respectively in the s . c . group ( Fig 4F and 4G ) . Since s . c . inoculation of the attenuated parasites was able to long-term protect BALB/c mice against a L . major infective challenge , we decided to analyze the prophylactic properties of the s . c . administered vaccine in C57BL/6 mice . Inoculation of 1 × 107 LiΔHSP70-II promastigotes in the right footpad of mice produced a chronic infection in the DLN ( right popliteal ) as revealed by the analysis of parasite burdens at week 4 and week 12 post-vaccination , whereas parasites were found in the site of vaccination ( right footpad ) at week 4 after vaccination but disappeared at week 12 ( in 87 . 5% of the mice; 7/8 mice ) ( Fig 5A ) . Attenuated parasites were absent of the internal organs ( Fig 5A ) . The presence of a persistent number of LiΔHSP70-II parasites in the popliteal lymph node draining the site of the attenuated line inoculation was maintained after L . major challenge up to 25 weeks ( S4 Fig ) . Short-term and long-term vaccinated mice showed an IgG2c predominant antibody response against the parasite ( Fig 5B ) and their spleen cells secreted IFN-γ after in vitro stimulation with L . infantum SLA ( Fig 5C ) . Contrary to the long-term vaccinated group , short-term vaccinated mice secreted detectable levels of IL-10 in response to SLA ( Fig 5C ) . A SLA-dependent production of IFN-γ was detected in the popliteal LN culture supernatants from both vaccinated groups , higher in magnitude at week 4 after infection along with IL-10 production ( Fig 5D ) . Most importantly , neither short-term nor long-term vaccinated mice showed any inflammatory lesion when challenged with 1 × 103 L . major metacyclic promastigotes in the ear dermis ( Fig 6A ) . At week five after challenge , L . major burdens were similar in short- and long-term vaccinated mice , showing a 1 . 5-Log ( Fig 6B ) and 2-Log ( Fig 6C ) reduction in the ears and DLNs , respectively , when compared to the saline controls . No L . major parasites were found in visceral organs ( liver or spleen ) . Retromandibular LNs cells from mice of the saline group were able to secrete higher amounts of cytokines than vaccinated mice when analyzed at week 5 after L . major challenge ( Fig 6D ) . A significant increment in IFN-γ ( P = 0 . 023 relative to week 4 and P = 0 . 041 relative to week 12 ) and IL-10 ( P = 0 . 022 relative to week 4 and P = 0 . 012 relative to week 12 ) was observed in LN culture supernatants from cells obtained from saline controls when compared with both vaccinated samples after in vitro stimulation with SLA . As it was expected because of the presence of inflammatory lesions , control mice DLN cells secreted IFN-γ in the absence of SLA stimulation whereas this cytokine was absent in unstimulated cultures stablished from vaccinated mice ( Fig 6D ) . On the other hand , similar amounts of IFN-γ were observed among the three groups when the stimulation assay was performed in spleen cell cultures ( Fig 6E ) . These data , besides the presence of IgG2c anti-SLA antibodies in all groups ( Fig 6F ) allowed the conclusion that all mice groups have a systemic Th1 response against the parasite . In controls , the lymph node inflammatory response was related to the presence of high numbers of L . major parasites ( i . e . showing inflammatory lesions ) . In the vaccinated mice the limited infection in the DLNs was correlated to a lower IFN-γ local response . Thus , systemic response mounted by the asymptomatic infection of the attenuated line , resulted in a protective response against L . major challenge in the absence of pathological lesions . Next , we tested whether the Th1 response induced by the inoculation of the attenuated line is able to anticipate the response against L . major parasites in the site of infection , resulting in a non-pathological protection . For that purpose , C57BL/6 mice were inoculated with the attenuated line 4 weeks or 12 weeks before , and then challenged in the ears with 1 × 103 metacyclic forms of L . major . A progressive increment in the number of parasites found in the ear an in the DLNs was observed in the saline groups up to day 28 post-challenge ( Fig 7A–7D ) . On the contrary , the number of parasites were stabilized in vaccinated mice 28 days after challenge in the short-term protected mice ( Ear Fig 7A; DLNs Fig 7B ) , and from day 14 in the ear or day 21 in the DLNs in the long-term group ( Fig 7C and 7D , respectively ) . Parasite replication control was correlated to the early presence of circulating anti-SLA IgG2c antibodies in the sera from vaccinated mice after L . major challenge ( Fig 7E ) . A higher reactivity that was not statistically different was observed in mice of the long-term group when compared to the short-term vaccinated animals . In addition , short-term and especially long-term vaccinated mice were able to mount earlier cellular responses against the parasite than control mice , as demonstrated by the levels of IFN-γ secreted to the culture supernatants in the three groups , after in vitro stimulation with SLA of the cells obtained from L . major infected DLNs ( Fig 7F ) . Finally , we analyzed IFN-γ synthesis at the site of L . major challenge upon long-term vaccination ( Fig 8 and S5 Fig ) . For that purpose , mice inoculated with the attenuated parasites and their corresponding saline controls were challenged , 12 weeks after vaccination in the ear dermis with L . major ( 1 × 105 ) . As an additional control , a group of vaccinated mice was i . d . injected with PBS in the ears at the time of L . major challenge . Three days after inoculation , the presence of IFN-γ secreting cells in the ears and the DLNs was analyzed by flow cytometry . Both CD4+ ( Fig 8A ) and CD8+ ( Fig 8B ) IFN-γ secreting T cells were detected in mice vaccinated with the attenuated line shortly after L . major challenge . Such cells were absent from the site of infection of unchallenged vaccinated mice , or in non-vaccinated and infected animals .
The fact that patients recovered from CL disease are usually resistant to reinfection has been taken as an indication that a vaccine against this form of the disease is feasible . Historically , leishmanization ( inoculation of live virulent L . major parasites ) was employed to induce immunity against CL , and although it is currently in disuse , the practice is coming back in regions of high incidence because of its effectiveness [15–17] . The use of murine models of CL demonstrated that leishmanization protects C57BL/6 resistant mice from vector transmitted L . major infection contrary to vaccines based on parasite extracts [18] which failed to protect against natural infection . Also , some limitations were observed in the prophylactic properties of the most evolved recombinant molecules based vaccines showing different protection degree in distinct murine models of CL due to sand fly transmitted infection [19 , 20] . In addition , to maintain immunity , protein-based vaccines require boosting doses , since transient effector T cell responses preclude the induction of long-term immunity [21] . On the contrary , the balanced effector/memory T response induced by the infection with virulent parasites can be maintained by parasite persistence , resulting in long-term immunity [16 , 17] . A possible limitation of leishmanization has emerged after analyzing the influence of L . major challenge on the evolution of leishmaniasis caused by other species of Leishmania in murine models . Whereas cutaneous infection with L . major provided heterologous protection against VL due to L . infantum infection in C57BL/6 mice [22] the IL-4 mediated humoral response elicited against the parasite by leishmanization in BALB/c mice caused an aggravation of VL disease when ‘leishmanized’ mice were challenged with L . infantum [23] . In recent years , vaccination with genetically attenuated parasites is being contemplated as a promising alternative to leishmanization , avoiding the problems derived from using non-attenuated parasites [17 , 24] . As reviewed in [25] genetically modified L . major attenuated lines have shown some limitations when tested as vaccines against CL . Thus , protection against L . major challenge induced in resistant or susceptible mice by the inoculation of the L . major conditional auxotroph due to targeted deletion of the dihydrofolate reductase-thymidylate synthase gene ( Lmdhfr-ts-/- ) [26] was not reproduced in a primate model [27] . In addition , vaccines based on the L . major line lacking phophoglycans ( LmLpg2- ) presented differences in the induced protective immunity depending on the murine model assayed [28 , 29] . Also , infection of a L . major genetically modified arginase deficient line resulted in a chronic disease in which lesions did not disappear in the resistant mouse strain [30] . The most efficient genetically modified vaccine for CL was constructed in L . major by the inclusion of two suicide genes ( Lmtkcd+/+ ) that render the parasite susceptible to ganciclovir and 5-flurocytosine [31] . This vaccine has been tested to be effective in the BALB/c [31] or in the C57BL/6 models [32] , but treatment needs to be administered to recipients after vaccination complicating the vaccine schedule . As an alternative , in this work we propose the use of single dose of a live vaccine based on a L . infantum genetically attenuated line [13] to induce protection against CL . Regarding the evolution of the attenuated parasite burdens in the vertebrate host , inoculation of LiΔHSP70-II in BALB/c mice using the i . v . route led to a systemic infection with a pattern of parasite clearance with time post-infection in all internal organs . Importantly , challenge with infective L . major , did not produce the reactivation of the attenuated line . Similarly , i . v . inoculation of the latest and more promising attenuated vaccines based on L . donovani ( LdCen-/- ) deficient in centrin , a calcium binding cytoskeletal protein [33] or Ldp27-/-; lacking a protein forming part of the active cytochrome c oxidase complex [34] ) produced a transient systemic infection , resulting in the impossibility to detect vaccine parasites in internal organs at long-term [34 , 35] . The absence of detectable parasites in the spleen is characteristic of the infection with different attenuated viscerotropic lines and differs from the chronic infection of the spleen observed after challenge with infective parasites [36 , 37] , reinforcing the attenuated nature of the LiΔHSP70-II line . Further , a previous report indicated that the LiΔHSP70-II line tend to be undetectable when inoculated i . v . in immuno-deficient SCID mice [14] . These data were taken as a suggestion that parasite clearance did not strictly depends on the induction of T cell dependent responses . Something similar occurred with the LdCen-/- attenuated line [35] , but not with other versions of genetically modified parasites , as for example the SIR2-deficient L . infantum ( LiSIR+/- single knock-out for the sirTuin encoding gene ) [38] . Additionally , it was reported that LiΔHSP70-II intra-cardiac ( i . c . ) inoculation of hamsters , a highly susceptible VL model by L . infantum challenge [39] , generates an asymptomatic infection . The absence of clinical signs of disease was correlated to the impossibility to detect attenuated parasites in the internal organs up to 9 months after inoculation [14] . All these data may be taken as an indication of the high biosafety degree of the LiΔHSP70-II line . Interestingly , s . c . inoculation induced a localized infection without dissemination to internal organs , resulting in the persistence of the parasites in the DLN of the infection site in BALB/c and C57BL/6 mice . On the contrary , footpad parasite burdens decreased after infection and became undetectable at longer times . The presence of parasites in the DLN accompanied by a parasite clearance with time in the site of challenge has been also described for infective L . donovani [40] or L . infantum [41] , but in these models spleen macrophages resulted chronically infected . Since mice from the s . c . and i . v . vaccinated groups showed similar long-term protection against L . major challenge , it can be hypothesized that in the i . v . vaccinated mice some parasites may persist dispersed in different internal organs , but remain undetectable perhaps due to their low number . In addition , the presence of parasites in other cell types maintaining latent infections can be also a source of parasite persistence [42 , 43] . This is an important issue , since maintenance of the parasite in the vertebrate host would be assuring the maintenance of immunity in the absence of recall doses against Leishmania [8] or other parasitic infections [44–47] . In this regard , the persistence of parasites may be indispensable to produce a concomitant immunity maintaining the number of Teff cells [48] . Nevertheless , other cells implicated in protection , namely Tcm or tissue resident memory T ( Trm ) cells can persist after parasite clearance [6 , 49–51] . In this context , immunization with the LiΔHSP70-II line elicits both Tcm cells and Teff or Tem responses . One limitation of this work is the lack of studies performed to analyze the implication of Trm cells in the observed protection , an interesting question that should be addressed in future research . On the other hand , the implication of Teff cells in the robust protection associated with vaccination was demonstrated by the data obtained in the C57BL/6 mice model . It has been described that C57BL/6 mice healed from a first infection with L . major develop concomitant immunity to re-challenge consisting in the rapid migration of IFN-γ producing Teff cells to the site of reinfection [22 , 48 , 52 , 53] . Here , we observed that 3 days after L . major challenge , a group of IFN-γ producing CD4+ and CD8+ T cells were specifically detected in the ears and lymph node cells of vaccinated mice . These cells may correspond to pre-existing Teff cells , since Tcm cells may need more time to elicit a protective response [48] . Another limitation of our work is related with the fact that L . major was administered using a needle . However , the rapid Teff cell response demonstrated in the ear and retromandibular DLNs of C57BL/6 mice after L . major challenge may be considered as a good predictor for protection against natural challenge as occur in ‘leishmanized’ C57BL/6 mice [18] . The humoral and cellular response elicited by the BALB/c mice inoculated with the LiΔHSP70-II was quantitatively similar to that observed in BALB/c mice when infected i . v . or s . c . with the L . infantum infective parasites [36 , 41] or to that generated after intraperitoneal inoculation with another L . infantum based attenuated line vaccine ( LiSIR2+/- ) [38] . A mixed IgG1/IgG2a humoral response , higher in titer at short-term , was observed concomitant with the systemic secretion of parasite-specific IFN-γ and IL-10 by splenocytes and a local production of SLA-dependent IFN-γ , IL-10 and IL-4 cytokines in the LN draining the site of s . c . vaccination , especially at short-term . The higher IFN-γ/IL-10 ratio showed at long-term can be correlated with clearance of the attenuated parasite , since IL-10 production has been largely related to parasite persistence of viscerotropic species [54–57] . LiΔHSP70-II s . c . administration to C57BL/6 mice resulted in the induction of both local and systemic Th1-like response in short- and long-term vaccinated groups , characterized by the induction of SLA-dependent IFN-γ and the presence of IgG2c anti-Leishmania antibodies . The existence of susceptible and resistant models of L . major infection is an advantage when testing experimental vaccines . In the case of C57BL/6 mice , infection with a low number of metacyclic promastigotes in the dermis of the ear generates a clinically silent phase in which parasite replicates . In a second phase the IFN-γ mediated local inflammatory response reduces parasitic load leading to skin lesions similar to those of CL human patients [58] . In contrast , challenge with large numbers of parasites in the footpad of BALB/c mice generates a progressive infection associated with parasite-specific responses mediated by IL-10 and IL-4 [4 , 59 , 60] . Many vaccine candidates have been tested in both models with different results . In the resistant model , the protection has been linked to an anticipation of the inflammatory response , with the induction of CD4+ and CD8+ T cells producing IFN-γ which results in early control of the parasite and , therefore , in the appearance of lower grade lesions [61–64] . In the BALB/c model , numerous evidences suggest that the control of the infection not only depends on the induction of IFN-γ-mediated responses , but also on the control of IL-10 and IL-4 cytokines that are associated with pathology [29 , 62 , 65] . This is the case of the LmLpg2- line that was able to control the pathology in BALB/c mice alleviating disease associated responses , but did not reach the same degree of protection in the resistant model when inoculated in the absence of a cellular inducing adjuvant [28 , 29] . As occurred for some subunit [62] or live vaccines [31 , 32] , LiΔHSP70-II line inoculation was able to induce a robust protection in both murine models of CL . For the susceptible BALB/c mice , we first used the i . v . route , since it is the classical route of administration for viscerotropic specie based vaccines . Given that more acceptable routes of vaccination are desirable for human use the s . c . administration was also tested . Independently of the administration route and compared to unvaccinated mice , BALB/c mice inoculated with the LiΔHSP70-II line showed significant control of the leishmaniasis disease . This is an interesting property of our vaccine , since the Lm dhfr-ts-/- line conferred protection against L . major infective challenge in BALB/c mice when it is i . v . but not s . c . administered [26] . It was also reported that protection conferred against L . mexicana by intraperitoneal inoculation of an attenuated line of the same species ( lacking guanosine diphosphate-mannose pyrophosphorylase; LmΔGDP-MP ) was not attained when it was s . c . administered [66] . Our data demonstrated that i . v . or s . c . BALB/c vaccinated mouse groups showed a Th1-like response against parasite antigens after L . major challenge . Anti-SLA humoral responses changed from the IgG1 subclass ( found in the non-vaccinated controls ) towards a IgG2a response . Higher IgG2a titers were observed long-term compared to short-term , correlating to a better protection degree . The parasite dependent IFN-γ response was higher in vaccinated than in control animals . The production of this cytokine was detected in short-term groups , but accompanied by the secretion of the highest levels of IL-10 among all groups . On the other hand , long-term protected mice showed a moderate SLA dependent IFN-γ production accompanied by very low parasite dependent IL-10 responses , similar to the protection conferred by the Lmlpg2- parasites that was associated with control of parasite mediated IL-10 responses in this susceptible model [29] . For C57BL/6 mice , whereas the infection of non-vaccinated mice evolved as described [5 , 48 , 58] , vaccinated mice showed no lesions at all . During the first two weeks after L . major challenge parasites similarly grew in the ear and the DLN in both vaccinated and non-vaccinated mice . Afterwards , control group continued in the silent phase incrementing their parasite burdens , while immunized mice do not allow parasites to expand further and showed earlier production of IFN-γ in the DLN . The anticipation of the effector response implies that the production of low levels of IFN-γ is sufficient to control parasite burdens without producing tissue damage . Then , LiΔHSP70-II parasites achieve the gold standard of protection against CL in C57BL/6 mice reaching a degree of protection comparable to that described for the more protective subunit based vaccines [61–63] . Although genetically attenuated vaccines may be an alternative to leishmanization to control human CL , concerns regarding biosafety remain , as it is mandatory that the attenuated phenotype is maintained even in cases of severe immunosuppression . In this regard , it is very important to target parasite genes whose function cannot be regained by compensatory mutations that can lead to recover the virulent phenotype to genetically modified parasites [67] . Nevertheless , the results shown in this work together with promising results observed using attenuated parasites to control malaria [68–71] and other pathologies [72–76] support the idea that live attenuated vaccines might be the basis for the development of vaccines against human CL in the next future .
Female BALB/c mice and C57BL/6 ( 6–8 weeks old ) were purchased from Harlan ( Barcelona , Spain ) . All procedures were performed according to the Directive 2010/63/UE from the European Union and RD53/2103 from the Spanish Government . Procedures were approved by the Animal Care and Use Committee at the Centro de Biología Molecular Severo Ochoa ( CEEA-CBMSO 21/138 ) , the Bioethical Committee of the CSIC ( under reference 100/2014 ) . The final approval was authorized by the Government of the Autonomous Community of Madrid under the reference PROEX121/14 . The following parasites cell lines were employed: L . major clone V1 ( MHOM/IL/80/Friedlin ) ; L . infantum ( MCAN/ES/96/BCN150 ) and the attenuated line ( L . infantum MCAN/ES/96/BCN150 [Δhsp70-II::NEO/Δhsp70-II::HYG] ) [12] . The attenuated line was created as described in [12] . Briefly , both alleles of the single hsp70-II gene located at chromosome 28 of the L . infantum genome were replaced sequentially with the ORF of the NEO and the HYG selectable marker genes by homologous recombination using plasmids constructions containing the marker genes flanked by specific regions located upstream and downstream of the ORF for hsp70-II gene [12] . The LiΔHSP70-II line showed a mild growth-rate defect in the logarithmic growth phase , concomitant with a longer duration of the G2/M phase of the cell cycle . In addition , promastigotes of the mutant line reached lower cell density than wild type parasites in culture , suffering a rapid decrease after reaching the stationary growth phase [12 , 13] . Lack of functional HSP70-II gene did not affect the rate of macrophage in vitro infection but the infected macrophages showed reduced number of internal amastigotes when compared to the wild type line [13] . Parasite persistence was demonstrated in experimental infections performed in the BALB/c mice strain , since four weeks after challenge viable parasites were recovered from different organs [13 , 14] . The promastigote forms of the parasites were grown at 26°C in Schneider medium ( Gibco , NY , U . S . A . ) supplemented with 10% Fetal Calf Serum ( FCS ) ( Sigma , MO , U . S . A . ) , 100 U/ml of penicillin and 100 μg/ml of streptomycin . For the attenuated line , medium was supplemented with 20 μg/ml of G418 and 50 μg/ml of hygromycin . Parasites were kept in a virulent state by passage in BALB/c mice . For vaccination , two administration routes were employed . BALB/c mice were immunized by the administration of 1 × 107 LiΔHSP70-II promastigotes suspended in 100 μl of phosphate saline buffer ( PBS ) in the vein tail ( intravenously; i . v . ) . Subcutaneously ( s . c . ) immunization of BALB/c and C57BL/6 mice were performed with 1 × 107 LiΔHSP70-II promastigotes suspended in 30 μl of PBS in the right footpad . As control , mice were inoculated with PBS . In all experiments performed with BALB/c mice and in those shown in Fig 6 for C57BL/6 a single control group was employed for long- and short-term protection analyses . In these cases , mice were inoculated twice with PBS ( week 12 and 4 ) i . v . ( BALB/c ) or s . c . ( both mice strains ) . To obtain data shown in Fig 7 employing C57BL/6 mice , two different control saline groups were employed for long- or short-term analyses , receiving only one PBS dose coinciding with vaccination . For challenge , BALB/c mice were infected with 5 × 104 stationary-phase promastigotes of L . major suspended in 30 μl of PBS into the left footpads . Infection follow-up was performed by measuring footpad swelling with a metric digital caliper . Lesion size was expressed as thickness of the L . major infected left footpad minus thickness of the right footpad . C57BL/6 mice were challenged with 1 × 103 ( or 1 × 105 when indicated ) L . major metacyclic promastigotes isolated by negative selection with peanut agglutinin , suspended in 10 μl of PBS into the dermis of both ears ( intradermal; i . d . ) . Ear lesions diameter was measured with a metric caliper . The number of LiΔHSP70-II parasites was determined in the liver , spleen , BM ( after i . v . or s . c . administration ) and also in the DLNs and footpads after s . c . administration . In addition , L . major parasite burdens were determined in the DLNs ( popliteal for BALB/c mice and retromandibular for C57BL/6 mice ) , ears ( C57BL/6 mice ) or liver and spleen ( both strains ) . The number of parasites was determined by a limiting dilution assay as described in [77] . For cell preparation , the complete spleens , lymph nodes and footpads , or a piece of approximately 20 mg of liver were stored in Schneider medium containing 20% heat-inactivated , 100 U/ml of penicillin and 100 μg/ml of streptomycin at 4°C . Tissues were homogenized and filtered through 70 μm cell strainers ( Corning Gmbh , Kaiserslautern , Germany ) to obtain a cell suspension . BM samples were obtained by perfusion of the mouse femur marrow cavities with Schneider medium before filtration . For ear processing , the ventral and dorsal sheets were separated and incubated in Dulbecco's modified Eagle medium ( DMEM; Thermo Fisher Scientific , MA , U . S . A . ) containing Liberase CI enzyme blend ( 50 μg/ml; Roche Diagnostics , Basel , Switzerland ) . After 2 h of incubation at 37°C , the tissues were cut into small pieces , and homogenized and filtered using a cell strainer as indicated above . Each homogenized tissue sample was serially diluted ( 1/3 ) in a 96-well flat-bottomed microtiter plate containing the same medium employed for homogenization ( in triplicates ) . For LiΔHSP70-II parasite number determination , medium was also supplemented by 20 μg/ml G418 and 50 μg/ml hygromycin . The number of viable parasites was determined from the highest dilution at which promastigotes could be grown up to 10 days of incubation at 26°C and is indicated per whole organ ( spleen , lymph nodes and footpads ) , per g ( liver ) or as number of parasites in 107 cells for the BM samples . Sera were obtained from blood samples taken before and after leishmanization with the attenuated line or after infective challenge . The reactivity against parasite proteins was determined by ELISA , using SLA prepared from L . major or L . infantum promastigotes . Briefly , SLA was prepared by three freezing and thawing cycles of stationary promastigotes suspended in PBS followed by centrifugation for 15 min at 12 , 000 × g using a microcentrifuge . After determining protein concentration by the Bio-Rad Protein Assay Dye Reagent ( Bio-Rad laboratories , München , Germany ) supernatants were collected and stored at -70°C . Sera reactivity was calculated as the reciprocal end-point titer calculated as the inverse value of the highest serum dilution factor giving an absorbance > 0 . 15 . Briefly , MaxiSorp plates ( Nunc , Roskilde , Denmark ) were coated with 100 μl of SLA diluted in PBS ( 12 μg/ml for 12 h at 4°C ) . After four washes with 200 μl of PBS-Tween20 0 . 5% ( washing buffer ) , wells free binding sites were blocked with the same volume of the blocking solution ( PBS-Tween 20 0 . 5%–5% non-fat milk ) for 1 h at room temperature ( RT ) and incubated with serial dilutions ( 1/2 dilution factor in blocking solution ) of mouse sera for 2 h at RT . After four washes with 200 μl of washing buffer , wells were incubated for 1 h at RT with secondary antibodies . Anti-IgG , anti-IgG1 , anti-IgG2a or anti-IgG2c horseradish peroxidase-conjugated anti-mouse immunoglobulins were used as secondary antibodies at 1/2 , 000 dilution in blocking buffer ( Nordic BioSite Täby , Sweden ) . After four washes performed as above , the reaction was developed through incubation with orto-phenylenediamine for 10 min in the dark . Color development was stopped by the addition of 2 N H2SO4 . Optical densities were read at 490 nm in an ELISA microplate spectrophotometer ( Model 680 , Bio-Rad Laboratories ) . For cytokine analysis , primary cultures were stablished from spleens and LNs as described above , but using RPMI complete medium ( RPMI medium ( Sigma ) supplemented with 10% heat-inactivated FCS , 20 mM L-glutamine , 200 U/ml penicillin , 100 μg/ml streptomycin and 50 μg/ml gentamicin instead of Schneider medium . Cells ( 5 × 106 ) were cultured during 72 h at 37°C in 5% CO2 in the absence or in the presence of SLA at 12 μg/ml of final concentration . The levels of IFN-γ , IL-10 or IL-4 in culture supernatants were determined by sandwich ELISA using commercial kits ( Pharmingen , San Diego , CA , USA ) . For the analysis of effector T cells ( Teff ) or effector memory T cells ( Tem ) ( CD44+ CD62Llow subset ) and central memory T cells ( Tcm ) ( CD44+ CD62Lhigh subset ) , single cell suspensions from the spleen on the BALB/c mice were processed as above , and the single splenocytes were harvested , washed in PBS with 1% heat-inactivated FCS and incubated with Rat Anti-Mouse CD16/CD32 ( FcBlock , BD , Franklin Lakes , NJ , USA ) followed by the staining with the surface markers: AlexaFluor 647 Rat Anti-Mouse CD3 Molecular Complex ( 17A2 Clone , BD ) , APC/Fire 750 Anti-Mouse CD44 ( IM7 Clone , BioLegend , San Diego , CA , USA ) , BV421 anti-mouse CD62L ( MEL-14 Clone , BioLegend ) and BV570 anti-mouse CD4 ( RM4-5 Clone , BioLegend ) for 20 min at 4°C . After washing , cells were fixed and permeabilized with Cytofix/Cytoperm ( BD ) . Finally , cells were washed and analyzed . For identification of cell producing cytokines in BALB/c mice , single cell suspensions from the spleens or the popliteal lymph nodes of the BALB/c mice were processed as above . Subsequently , cells ( 1 x 106 ) were stimulated for 2 h at 37°C in RPMI complete medium with anti-mouse CD28 ( eBioscience , San Diego , CA , USA ) in flat-bottom 96-well plates previously coated with anti-mouse CD3e antibody ( eBioscience ) 24 h before . Afterwards , 10 μg/ml Brefeldin A was added to stimulated and non-stimulated cells and incubation continued for 4 h more . Then , cells were harvested , washed in PBS with 1% heat-inactivated FCS and incubated with Fc block followed by the staining with the surface markers FITC anti-mouse CD8a ( 53–6 . 7 Clone , BioLegend ) , and BV570 anti-mouse CD4 for 20 min at 4°C . After washing , cells were fixed and permeabilized with Cytofix/Cytoperm ( BD ) . Next , PE/Cy7 anti-mouse IFN-γ ( XMG1 . 2 Clone , BioLegend ) antibody was added for 30 min at 4°C . Finally , cells were washed and analyzed . For the analysis of the frequency of T cell producing IFN-γ in the ears and retromandibular lymph nodes of C57BL/6 mice , single cell suspensions were processed 3 days after L . major challenge and 1 × 106 cells were stimulated for 2 h at 37°C with anti-mouse CD3/CD28 ( eBioscience ) as described above . Afterwards , 10 μg/ml Brefeldin A was added and cells were incubated for 4 h more . Then , cells were washed and incubated with FcBlock followed by the staining with the surface markers FITC anti-mouse CD8a , AlexaFluor 647 Rat Anti-Mouse CD3 Molecular Complex and BV570 anti-mouse CD4 for 20 min at 4°C . After washing , cells were fixed and permeabilized with Cytofix/Cytoperm . Next , PE Rat Anti-Mouse IFN-γ ( XMG1 . 2 Clone , BD ) antibody was added for 30 min at 4°C . Finally , cells were washed and analyzed . All cells were analyzed using a FACS Canto II flow cytometer and FACSDiva Software ( BD ) and processed and plotted with FlowJo Software ( FlowJo LLC , Ashland , Oregon , USA ) . Statistical analysis was performed using the Graph-Pad Prism 5 program . Data were first analyzed by the D'Agostino & Pearson normality test when sample was n ≥ 8 . Parametric data were analyzed by a two-tailed Student t-test when comparing two samples or one-way ANOVA followed by the Tukey test when comparing more than two groups . Non-parametric data ( or data with n < 8 ) were analyzed by a Mann Whitney test or a Kruskal-Wallis test ( Dunn's post-test ) when comparing two or more groups , respectively . Differences were considered significant when * P < 0 . 05 . | Despite numerous efforts made , a vaccine against leishmaniasis for humans is not available . Attempts based on parasite fractions or selected antigens failed to confer long lasting protection . On the other side , leishmanization , which consists in the inoculation of live virulent parasites in hidden parts of the body , is effective against cutaneous leishmaniasis in humans but objectionable in terms of biosafety . Some efforts have been made to design live vaccines to make leishmanization safer . A promising strategy is the development of genetically attenuated parasites , able to confer immunity without undesirable side effects . Here , we have employed an attenuated L . infantum line ( LiΔHSP70-II ) as a vaccine against heterologous challenge with L . major in two experimental models . Infection with LiΔHSP70-II parasites does not cause pathology and induces long-term protection based on the induction of IFN-γ producing T cells that are recruited rapidly and specifically to the site of challenge with the virulent parasites . These results support the idea of using attenuated parasites for vaccination . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] | [
"medicine",
"and",
"health",
"sciences",
"immune",
"physiology",
"spleen",
"ears",
"immunology",
"microbiology",
"parasitic",
"diseases",
"parasitic",
"protozoans",
"protozoan",
"life",
"cycles",
"animal",
"models",
"preventive",
"medicine",
"developmental",
"biology",
"model",
"organisms",
"protozoans",
"leishmania",
"experimental",
"organism",
"systems",
"vaccination",
"and",
"immunization",
"promastigotes",
"inoculation",
"research",
"and",
"analysis",
"methods",
"public",
"and",
"occupational",
"health",
"life",
"cycles",
"mouse",
"models",
"head",
"leishmania",
"major",
"anatomy",
"physiology",
"biology",
"and",
"life",
"sciences",
"protozoology",
"organisms"
] | 2017 | Vaccination with a Leishmania infantum HSP70-II null mutant confers long-term protective immunity against Leishmania major infection in two mice models |
Since the pioneering work by Julius Adler in the 1960's , bacterial chemotaxis has been predominantly studied as metabolism-independent . All available simulation models of bacterial chemotaxis endorse this assumption . Recent studies have shown , however , that many metabolism-dependent chemotactic patterns occur in bacteria . We hereby present the simplest artificial protocell model capable of performing metabolism-based chemotaxis . The model serves as a proof of concept to show how even the simplest metabolism can sustain chemotactic patterns of varying sophistication . It also reproduces a set of phenomena that have recently attracted attention on bacterial chemotaxis and provides insights about alternative mechanisms that could instantiate them . We conclude that relaxing the metabolism-independent assumption provides important theoretical advances , forces us to rethink some established pre-conceptions and may help us better understand unexplored and poorly understood aspects of bacterial chemotaxis .
Behavioral analysis of chemotaxis in E . coli has shown that up-gradient or down-gradient directional movement is achieved through the combination of two basic types of movements , tumbling and running ( see Figure 1 ) . These two behaviors are both achieved through the rotation of flagella . Rotating the flagella in one direction ( counter-clockwise ) results in a directed motion of the bacterium called ‘running’ while brief periods of rotation in the other direction cause ‘tumbling’ , the production of a more-or-less random new orientation . Many swimming bacteria make use of similar patterns , alternating between a straight motion mode and a random change of direction [25] . Bacteria combine these movements in such a way as to produce a stochastic chemotactic behavior . How do they accomplish this ? In some species this is relatively well understood . For instance , in E . coli , a two component signal transduction system [13] compares current and past concentrations of attractants and causes the bacterium to run if the concentration of an attractant is increasing and tumble otherwise ( see e . g . [14] ) . This general strategy is sometimes known as ‘adaptive gradient climbing’ because it is capable of adapting to a wide range of concentrations , climbing gradients whether the local concentration is very low or very high . This is not the only way that these basic-movements can be modulated to produce chemotaxis . What we have called the selective-stopping strategy ( also called ‘inverted response’ elsewhere [26] ) consists of a combination of running and tumbling to perform a random walk until the relevant concentrations are high at which point the tumbling motion dominates and the bacterium more or less tumbles in-place . Regardless of which chemotactic strategy is employed , it is important to identify its sensitivity—what is it responding to ? The commonly accepted view is that chemotaxis mechanisms are only responsive to the concentration of attractant chemicals in the local environment of the bacterium and are not influenced by the current state of the bacterium's metabolism . This view is known as metabolism-independent chemotaxis , for while the metabolism produces the mechanisms of sensitivity , transduction and response , the behavior of the bacterium is not influenced by the ongoing dynamics of the metabolism ( see Figure 2 top ) . In metabolism-independent chemotaxis , it is not the effect of the attractant upon the metabolism that causes it to move towards it , it is simply the way the attractant excites the sensors . Adler , in his seminal 1969 paper , showed seemingly compelling evidence for this view of bacterial chemotaxis , providing evidence in support of the following ( taken from [4] ) . “[R]esults show” Adler concluded “that extensive metabolism of the attractants is not required , or sufficient , for chemotaxis” [4 , p . 1596] . After this evidence was presented the metabolism-independent nature of E . coli chemotaxis became a generally accepted fact . As Alexandre and Zhulin note: “From that time on , research focused on the metabolism-independent information flow from membrane receptors to flagellar motors . ” [21 , p . 4681] . But this assumption has recently been challenged and the consequences for the study of chemotaxis are yet to be fully disclosed . Despite the predominance of metabolism-independent research on bacterial chemotaxis , there is evidence of different types of metabolism mediated chemotaxis . Metabolism-dependent chemotaxis involves an ongoing influence of the metabolism upon the chemotaxis mechanism ( see Figure 2 middle ) . This introduces the potential for a sensitivity to the effects of environmental phenomena upon the metabolism . In A . brasilense , metabolism-dependent chemotaxis has been systematically studied and is considered as the dominant behavioral strategy [19] , [21] . The following behavioral phenomena have been well-established [21]: Similar metabolism-dependent chemotactic phenomena have also been found in other bacteria species like Pseudomonas putida [27] and Rhodobacter sphaeroides [28] , [29] . Evidence of metabolism-dependent chemotaxis has also been found in the same species that was studied by Adler , E . coli [20] . Adler found that although glycerol is extensively metabolized by E . coli , it is not found to act as an attractant . As such , the case of glycerol seemed to provide supporting evidence in favor of metabolism-independent chemotaxis , and so argued Adler . Interestingly , in apparent contradiction to Adler's findings , Zhulin and collaborators observed chemotaxis towards similar levels of glycerol as studied by Adler [30 , p . 3199] . Evidence of metabolism-dependent chemotaxis in E . coli has also been found for proline and succinate metabolic substrates [21] . In addition , chemotaxis to oxygen ( aerotaxis ) in E . coli [20] has been shown to depend on metabolism ( i . e . , reduction of oxygen is required for aerotaxis ) together with redox gradient climbing [31] . These type of taxis has been termed “energy taxis” meaning that bacterial movement is sensitive to the energy production ( generally by the electron transport system having modulatory effects over CheA phosphorilation , but other mechanisms have also been proposed ) ( for recent reviews of energy taxis , see e . g . , [22] , [32] ) . The results described in this section all suggest that metabolism-dependent chemotaxis might be more widespread than previously thought . The biochemical organization of bacteria is astonishingly complex and difficult to model . One approach is to assume some sort of functional decomposition and to study sub-systems separately . Dennis Bray and his group have achieved some impressive results along these lines [15] , [16] , [33] . Their model of E . coli's chemotactic behavior includes molecular level details of the membranes , signal transducers , Brownian motion of molecules within the cytoplasm and a number of genetic details . Their approach has achieved unprecedented levels of predictability and empirical accuracy . However , despite the promising results , some aspects of E . coli chemotaxis remain elusive . For instance , how is it possible for a small number of types of sensor ( 5 ) to cause an appropriate response to a large number of attractants and repellents ( 50 ) ? It is possible that the elusiveness of these and other aspects of bacterial chemotaxis may not be a question of lack of mechanistic detail in the simulations themselves but may relate to the long-standing assumption that chemotaxis is fundamentally a metabolism-independent phenomenon . Using a different approach , Goldstein and Soyer artificially evolved metabolism-independent chemotactic pathways ( abstracting away the sensory and motor details ) [26] . Chemical pathways partially reproducing a gradient climbing response could only evolve under special conditions and for a limited range of basal levels of stimuli . Contrarily , they found that the selective-stopping strategy ( tumble in place if resources are high , otherwise perform a random walk ) was comparatively easy to evolve . The resulting selective-stopping strategy is simpler , yet robust and efficient . This strategy is also found in different types of bacteria particularly in those performing metabolism-dependent chemotaxis—like that of Rhodobacter sphaeroides [34]—and in E . coli when “normal” transduction pathways are knocked down [35] . Goldstein and Soyer found very simple chemical pathways capable of generating such response patterns . They speculate that “non-adaptive dynamics [i . e . the selective-stopping strategy] could even be achieved without any signaling proteins; a small molecule , that is a by-product of metabolism or is taken into the cell via a transporter , could directly regulate tumbling probability of the cell” , [26 , p . 5] . In order to further develop this hypothesis , we start from an idealized minimal metabolism and make it support and generate chemotactic behavior . Our model provides a proof of concept of how a minimal metabolism could already support experimentally observed metabolism-dependent chemotactic patterns . The model displays the response patterns shown by Goldstein and Soyer and , in addition , it reproduces some of the metabolism-dependent phenomena described by Alexander and Zhulin ( 2001 ) [21] . To summarize , metabolism-independent chemotaxis refers to types of chemotaxis where the chemotaxis generating mechanisms operate rather independently from metabolism . Flagellar rotation is only influenced by chemical pathways that are independent from the metabolic network and only modulated by transmembrane receptor activity ( see Figure 2 top ) . In metabolism-dependent chemotaxis , the chemical pathway that mediates transmembrane receptors and flagellar rotation is influenced by or coupled with metabolic pathways and processes ( like the electron transport system ) with the result of a metabolism-sensitive behavior ( see Figure 2 middle ) . A third possibility is metabolism-based chemotaxis , in which metabolism itself directly modulates behavior . In this case , there are neither specialized sensors ( like transmembrane proteins ) nor specialized and dynamically decoupled chemical pathways ( see Figure 2 bottom ) . Most of the available simulation models of chemotaxis are of metabolism-independent scenarios , there are no models of metabolism-dependent or metabolism-based chemotaxis . In the next section , we introduce a first model of metabolism-based chemotaxis to study its chemotactic potential and some theoretical implications ( the likelihood of the specific mechanisms in play and experimental support remain out of the scope of this paper ) .
The first step to create a minimal metabolism-based chemotactic agent is to distill and justify what counts as a minimal model of metabolism . We do not pretend to settle this issue here but we provide a general context that justifies the assumptions built into the model . It is generally accepted that life depends upon energetic and material resources in its ongoing self- ( re ) -production , self-maintenance and growth . Energy and matter flow through living systems , maintaining their biophysical and chemo-dynamic organization . These two flows are coupled: energy is used to transform materials into structures that harness energy to perform further material transformations [36]–[38] . In concrete terms , the energetic flow through the system depends upon the existence of catalysts which in turn depends upon the flow of energy through the system . This autocatalytic closure of chemical reactions has long been argued to constitute the core or essence of living organization , exemplified in metabolism [38]–[41]—see [42] , [43] for recent reviews and , more generally , see [44] , [45] for the central role played by metabolism in grounding biological agency and adaptivity . This metabolic organization stands far from thermodynamic equilibrium . Energy and matter are lost as heat and waste , requiring the continued acquisition of new resources . Abstracting away from the particularities of different metabolic networks , the following three key features remain essential to characterize metabolism: One of the simplest systems that has these features consists of the following autocatalytic reaction . where and represent material and energetic resources respectively , is a constituent or catalyst molecule and is low-energy waste . The above the arrow represents catalysis of this reaction by . This system is illustrated in Figure 3 , which also indicates the relative free energies of the reactants on the vertical axis . One can conceive of this reaction , according to the free-energies , as an exergonic reaction ( in which energy is released as ) that is coupled to ( and drives ) the endergonic reaction of . Real metabolisms , of course , have many intermediate steps in the production of enzymes which complicate the system . We assume that such intermediate steps can be justifiably abstracted away to illustrate a minimal instance of metabolism-based chemotaxis , taking the equation above to represent the higher order chemical dynamics of a whole metabolic network . The autocatalysis and modulation of flagellar rotation by the concentration of is indeed a simple mechanism . In theory , this simplicity could be taken even further . For instance , the energetic and material resources could come from a single molecule , . We chose to keep these resources separate so that we could investigate issues related to integration and to explore more complex environments ( see Experiments 2–6 ) . For those experiments where and are distributed in the same spatial distribution , the reaction and the simpler reaction play qualitatively equivalent roles and behavioral results are identical ( results not shown ) . We chose to maintain the same core reaction of throughout this paper for simplicity and parsimony . The waste particle also does not play a critical role in the dynamics that we have observed . And in fact , the autocatalysis is not strictly necessary to produce chemotactic behavior ( although it has dynamic consequences ) . A chemical reaction as simple as could have captured a large portion of the behavioral dynamics we have described here . However , we set out with the motivation of exploring metabolism-based chemotaxis . As we described earlier in this section , a metabolism requires the channeling of energy by enzymes into reactions that produce more of those enzymes . We have tried to capture that essential relationship in the minimal reaction . The following subsections describe the details of the modeling environment , how the above core reaction ( and some variations ) are implemented to model a minimal metabolism and how a simple coupling of this metabolism to an abstraction of the flagellar machinery generates chemotactic behavior . The model consists of a two-dimensional environment , containing resource gradients and simulated bacteria . Each bacterium has a position , orientation and velocity as well as a metabolism , which is represented by a set of chemical concentrations . The concentrations of these chemicals are updated each iteration through numerical integration of the differential equations that represent ongoing chemical reactions in the metabolism as well as degradation of metabolites and transport of ‘resource’ molecules from the immediate environment of the simulated bacteria into their interior . Each bacterium is always either ‘running’ ( moving in a straight line ) or ‘tumbling’ ( changing its orientation randomly ) . The probability of tumbling is directly proportional to the concentration of , the autocatalytic product of the metabolism . This metabolism-based behavioral mechanism causes the bacterium to remain still when the metabolic rate is sufficiently high and to run when its metabolism is not operating above a threshold rate . The simulated environment is 200 units square . Bacteria trying to move out of this area are prevented from doing so as if running into a wall . Details of the reactions , the behavioral mechanisms and the environment are given below . The autocatalytic reaction that constitutes metabolism is more explicitly described by the following reaction equations that include the intermediate stage where the catalyst , , is bound to one of its substrates , , forming . ( 1 ) ( 2 ) Two other processes influence the concentration of the metabolic reactants . The first is the degradation of reactants and into non-reactive products . These chemicals are removed from the simulation at rates specified in Table 1 . The second is the influence of the local environmental chemical concentrations upon the concentration of reactants within the simulated bacteria ( described below ) . The metabolism dynamics are simulated by numerical integration of the differential equations in Table 2 ( we used an Euler timestep of 0 . 01 and typical chemical concentrations ranged between 0 and 2 . 0 ) . These equations include some reactants that are only used in certain experimental scenarios and are explained later in the text . The rate constants ( and ) in the differential equations were determined by assigning free-energies to each reactant and activation-energies for each reaction such that the system adhered to the constraints given in our definition of a minimal metabolism . Given chemical free-energies and reaction activation-energies , reaction rates can be calculated by applying the following equations which indicate the reaction rate for a forward ( exergonic ) reactions and backward ( endergonic ) reactions respectively . Figure 4 indicates why the forward and backward equations are different . This method of determining reaction rates allows the exploration of abstract chemistry while remaining congruent with the law of thermodynamics . The environment of a bacterium can affect the concentration of certain chemicals within it , specifically: , , , , and . For simplicity , we assume that these resources are actively transported into the bacteria at a rate independent of the concentration of the chemicals inside the membrane . The internal resource levels are increased by continuous transport from the environment into the bacterium according to the following function:where is the concentration of the relevant chemical inside the bacterium , is the rate of transport across the membrane and is the concentration of relevant chemical in the environment of the bacterium . This influence of the environment is included in the differential equations in Table 2 as the last terms of those equations that update chemicals that are affected by the environment ( chemicals , , , , and ) . The above chemical reactions are simulated as enclosed within a membrane , comprising a simulated bacterium . Although minimal metabolisms have been the subject of simulation models , there have been very few attempts ( see e . g . , [46] ) to study the dynamics of metabolism coupled to some form of movement generation mechanism . In this model , inspired by the motion mechanism of E . coli and other species , the simulated bacteria are capable of moving in either a directed , ‘running’ motion or by randomly changing their orientation ( ‘tumbling’ ) . Bacteria are , by default , in a ‘running’ mode . Each iteration , however , a bacterium has a chance of tumbling that is proportional to the concentration of the product of the metabolism , : . Running bacteria move in a straight line in the direction of their orientation ( ) , , . Tumbling bacteria remain at the same location , with changed to a random value selected from a flat distribution between and . A full schematic diagram of the minimal metabolism and its coupling to behavioral mechanisms can be seen in Figure 5 .
In this first scenario , a source of and is centered at . Figure 6 shows the distribution of simulated bacteria at the start , middle and end of a iteration simulation . It can be seen how bacteria perform chemotaxis to the area of high-concentration of and ( which is indicated by the concentric circles ) . When a bacterium has access to plenty of resources , it produces significant quantities of and . The high concentration of causes the rate of tumbling to increase to the point where the bacterium is more or less standing in place since the tumbling frequency is so high that it never runs for a significant distance in any direction . If , on the other hand , the bacterium has insufficient available resources to maintain high levels of , the probability of tumbling will fall and the bacterium will perform a combination of running and tumbling that results in a random walk . This random walk will continue until it comes across a region where it can produce sufficient to push it “above threshold” . In this manner , the simulated bacteria perform a simple form of “selective stopping” chemotaxis whereby they move in a random manner until they are in a resource rich area , at which point they tend to remain where they are . Statistically , the simulated bacteria show a correlation between final location of the bacteria and high concentration of metabolizable substrates in the environment , i . e . , a “chemotactic” movement towards high-concentration of attractants . The individual behavior of the bacteria may not follow a direct chemotactic path , but the probabilistically directed ( i . e . , corrected or regulated ) behavior clearly results in an up-gradient movement tendency . Note that even in experimentally observed chemotaxis in E . coli , with their more sophisticated gradient-climbing adaptive strategy , the path followed by a single bacteria is difficult to characterize as chemotactic , it is rather the global effect of a population of bacteria that results in a clear chemotactic distribution . This experiment demonstrates that a metabolism-based control of flagellar rotation could potentially perform chemotaxis without dedicated signal transduction pathways , transmembrane receptor proteins nor sensory adaptation . Alexandre and Zhulin observed that “the presence of another metabolizable chemical ( exogenous or endogenous ) prevents chemotaxis to all attractants studied” [21 , p . 4682] . We tested our simulated metabolism-based chemotactic bacteria to confirm that they also undergo inhibition of chemotaxis due to the presence of alternative metabolizable resources . To perform this test we used the above chemotaxis scenario as a control . The new experimental scenario is identical to the first except that two new resources , and , are included , uniformly distributed throughout the environment at a concentration of 0 . 5 . For this experimental scenario ( and subsequent ones ) alternative resource molecules and reactions were required . For the sake of simplicity we created a duplicate metabolic pathway with identical stoichiometry , reaction rates , etc . , the only difference being the chemicals involved ( see Figure 7 ) : resource ( as analogous to ) and material resource ( analogous to ) . ( 3 ) ( 4 ) The results can be seen qualitatively by comparing Figure 6 ( the control ) and Figure 8 . It is clear that chemotaxis has been inhibited . The mean distance from the source ( ) for 10 runs of 100 agents each was ( std . ) for the control , and ( std . ) for the experimental abundance of alternative resource condition . The mechanism for this inhibition is simple . Resources and are ubiquitous and sufficient to maintain the concentration of at the high value necessary to keep the bacteria tumbling . The predominant tumbling keeps the agents stationary , preventing any chemotaxis to the resource . A second result was published by Alexandre and Zhulin in support of energy-taxis as the primary mechanism of chemotaxis in A . brasilense: “[The] inhibition of the metabolism of a chemical attractant completely abolishes chemotaxis to and only to this attractant” [21 , p . 4682] . To test this phenomenon in our simulation of metabolism-based chemotaxis , simulated bacteria are placed in an environment with two resource gradients . The first , consisting of equal parts of resources and is highest in concentration in the upper-right corner of the simulated environment . The second is equal parts of and and is highest in concentration in the lower-right corner . In the upper-right corner , resources and are sufficient for a healthy metabolism to continue to autocatalyze and maintain its concentration high enough for the bacteria to remain in this corner . The same is the case for and in the lower-right corner . This can be seen in the central bottom plot of Figure 9 . Halfway through this scenario , we add a uniform concentration of to the entire simulated environment . This chemical inhibits the / metabolic pathway by exothermically and rapidly bonding to metabolizable substrate , transforming it into a non-reactive chemical , ( see Figure 10 ) . This process is described by the following reaction equation: ( 5 ) After is added to the environment , the simulated bacteria cease to remain in the area high in concentration of and , but continue to be attracted to the high concentrations of and , as shown in the right-most plot of Figure 9 ( bottom ) , demonstrating inhibition of chemotaxis to a reactant by inhibition of metabolization of that reactant . Specific metabolic inhibitors such as oxidized quinones or specific electron transport inhibiting molecules such as myxothiazol have been shown to inhibit chemotaxis [47] . It has also been shown that such metabolic inhibitors can act as repellents [19] , [48] . We tested to see if our model could also display metabolic inhibitors ( or toxins ) acting as repellents . A repulsion due to a metabolic toxin can be clearly seen in Figure 11 which shows a scenario in which bacteria are evenly distributed in an environment of uniform distributions of . Halfway through the simulation , a gradient of metabolic inhibitor is added to environment ( with a peak concentration 5 . 0 , centered at ) and the bacteria move away from the higher concentrations of that toxin . Figure 12 indicates the reactions that occur in this scenario . Experiments 1–4 have reproduced empirical observations made by Alexandre and Zhulin . The following experiments explore additional phenomena that could lead to some empirical predictions . This scenario is inspired by Alexandre and Zhulin's observation that bacteria demonstrate a sensitivity to their history of exposure to different resources . Specifically , “[s]tronger chemotaxis responses are observed when cells are grown on the sugar under test as the growth substrate” [21 , p . 4682] . In this experiment , there is no in the environment except along a strip defined by . Agents are initialized with no . A gradient of is placed at the center of the right side of the environment . Bacteria have no , so the resource is insufficient to produce . Only once they have encountered the region with and incorporated into their metabolism does act as an attractant . This process can be seen through observation of Figure 13 where agents are drawn with s if they have concentrations of less than and as circles otherwise . Early in the simulation , agents tend to run as none have access to resources sufficient to produce and maintain significant quantities of . As time passes , the random motions of the bacteria cause some to encounter the on the left . More time passes , and these agents , now rich in can produce while in areas rich in . At this point , chemotaxis towards high concentrations of is observed . The experiment shows how what becomes an attractant for metabolism-based chemotactic bacteria is not an environmental compound per se , but rather what , at a given point on the history and internal state of the bacterium , is required for metabolism to occur . Bacteria performing metabolism-based chemotaxis operate according to their current metabolic needs . By basing chemotaxis in metabolism , the simulated bacteria respond not to specific environmental phenomena , but to the combined effects of all environmental features upon metabolism . In Experiment 6 , we demonstrate this ability to integrate environmental phenomena by placing bacteria in a more complicated environment than those of previous experiments . The environment for Experiment 6 consists of perpendicular linear gradients of resources ( ) and ( ) with a Gaussian distribution of toxin , with a peak concentration of centered at . Figure 14 indicates the final position of 5000 bacteria ( the results of 50 trials , each with 100 evenly distributed bacteria as in the other experiments ) . It can be clearly seen that the bacteria are neither maximizing concentrations of or , nor the combination of them , but are performing chemotaxis to the areas where the combined effects of the environmental resources , , and allow the metabolism to operate sufficiently well . The overall bacterial distribution appears correlated with the spatial distribution of the optimal combination of compounds . One interesting aspect of this plot is the asymmetric distribution of bacteria along the x and y axes , corresponding to the gradients of and respectively . It appears that for our model bacteria , it is more important to have a high concentration of than of . This difference may be caused by a high concentration of transforming ‘free’ into . , unlike , does not degrade , so a high concentration of makes the metabolism less likely to degrade than a high concentration of . An alternative possibility is that the bottleneck in the metabolism lies in the first reaction ( ) as this reaction has a slower rate constant than the second reaction ( ) . High has little influence on this bottleneck , but a significant concentration of can open up the bottleneck , allowing for a more rapid production of . We confirmed that a third possibility , the asymmetric influence of upon the reactions , is not responsible for the asymmetric distribution of bacteria ( results not shown ) . It is interesting to note that in aerotaxis experiments with e . g . E . coli and A . brasilense , the aerotactic bands can form an asymmetric profile as well ( see e . g . [49 , p . 2238 Figure B , top] ) . The environmental conditions at the different sides of these bands could be slightly different with respect to the metabolism of the bacteria , and perhaps a mechanism similar to that described here could explain the asymmetric distribution of bacteria in these experiments .
Research in bacterial chemotaxis has operated largely under the assumption that the behavior is supported by transmembrane receptors and dedicated signalling pathways and that such pathways are metabolism-independent . Despite the growing body of evidence that in many species this might not always be the case ( even for those largely thought to be so , such as E . coli ) available simulation models of bacterial chemotaxis assume metabolism-independence . Here we have presented the first minimal model of metabolism-based chemotaxis . It recreates phenomena observed in bacteria and allows us to explore some potential consequences of metabolism-based behavior . The behavioral strategy employed by the simulated bacteria in our model is the “selective stopping strategy” in which bacteria move around in a random walk until they reach a satisfactory area , at which point they tumble in place . Recent artificial evolution of simulated chemotaxis [26] has shown that this strategy ( also referred to as the “non-adaptive or inverted response” ) is , under certain ecological constraints , the most likely chemotactic strategy . It has also been observed in some cases of metabolism-dependent chemotaxis [34] . In order to address how metabolism could directly produce such behavioral patterns , we have developed a model of what we have called metabolism-based chemotaxis ( a simpler case than that of metabolism-dependent chemotaxis ) . We first identified a minimal metabolic organization as that of an autocatalytic reaction . We then assigned a probability of running or tumbling to the concentration of the auto-catalyst . The resulting system is very simple , yet capable of instantiating four chemotactic phenomena observed in bacteria . The observations of history dependence reported by Alexandre and Zhulin led to the exploration of Experiment 5 where a compound is incorporated into metabolism and results in a change in chemotaxis pattern according to past experience . Finally , Experiment 6 demonstrated the capacity of metabolism-based chemotaxis to respond appropriately to combinations of a variety of simultaneous environmental influences . This experiment showed the potential of metabolism as a mechanism for effective chemotactic integration . The present simulation is not a model of the specific mechanisms supporting metabolism-dependent chemotaxis in bacteria . Yet , it serves as a proof of concept of how a very simple abstraction of metabolism can support , without the addition of specific signaling pathways and even without the need of transmembrane receptors , a wide range of chemotactic phenomena . As a conceptual model it can be further used to explore some theoretical implications of relaxing the metabolism-independent assumption . Evidence has been found of metabolism-dependent chemotaxis where metabolic processes such as the electron transport system influence flagellar rotation indirectly , by way of the dedicated chemotactic two-component signaling system . There is also evidence for a mechanism through which metabolism directly influences flagellar rotation , i . e . , without an intermediate dedicated signaling system , in a manner more similar to the metabolism-based chemotaxis modeled here . Specifically , it has been shown that E . coli can perform chemotaxis even when stripped of most of the signaling pathway typically associated with chemotaxis [35] , suggesting that there might be at least two different mechanisms supporting chemotaxis in E . coli [2 , p . 575] . Interestingly , the concentration of fumarate , an intermediate in the citric acid cycle that is part of the “universal metabolism” [50] , has been shown to influence the direction of flagellar rotation . A high concentration of this metabolic product increases chance of clockwise , tumble inducing , rotation; the same relationship of metabolic influence upon flagellar rotation that is used in the selective-stopping strategy . Fumarate operates directly upon the flagellar motor switch [51] in a manner that is independent of the protein signaling pathway typically associated with chemotaxis [52] . It turns out that fumarate might be currently instantiating mechanisms of metabolism-based chemotaxis; something that still remains to be experimentally tested . This hypothesis was anticipated by [26 , p . 5] and we have shown how fumarate-like intermediate metabolites ( in our model ) could not only produce simple chemotaxis but could reproduce a wide spectrum of non-trivial chemotactic phenomena . What is needed to achieve these behavioral patterns is not a complex system of transmembrane receptors influenced by metabolism in subtle ways but simply a metabolite capable of influencing flagellar rotation in the right manner . The present model also suggests that it might be time to re-consider part of the terminology and the externalist approach to chemotactic studies . For instance , it is generally assumed that environmental compounds are invariably either attractants or repellents for bacteria , as if bacteria were simply stimulus-driven systems . The model of metabolism-based chemotaxis shows , however , that environmental compounds are not attractants or repellents purely on the basis of their binding properties and their stereotypically elicited responses . Environmental compounds must instead be categorized within the context of metabolism , which is influenced by the history of the cell and its internal organization ( metabolic rates , active and non-active metabolic pathways , etc . ) . In other words , the behavioral significance of chemical compounds becomes a relational property that depends on the metabolic dynamics of the cell ( which cannot be abstracted away in the study of behavior ) . As Experiment 3 shows , if a resource ceases to be metabolized , it ceases to act as an attractant for bacteria . Also , ( as shown in Experiment 4 ) chemical compounds that are toxic for the metabolism of the bacteria can act as repellents without the need of any specific binding of it , or even without the bacteria ever encountering that compound in its evolutionary past . This capacity to be behaviorally sensitive to the effects of environmental compounds on metabolism provides a powerful means of behavioral evaluation and increased adaptive response ( at the organismic level ) . It is not clear how the same adaptation could occur for a metabolism-independent mechanism that requires binding with specific compounds to elicit specific responses . While advances have been made on the understanding of how receptor complexes integrate sensory information [53] , [54] the potential integrative role of metabolism remains under-explored . The classic view is that integration in metabolism-independent chemotaxis is accomplished in the group dynamics of the transmembrane sensors that all modulate CheA activity [55] . This metabolism-independent mechanism of integration relies upon specific interactions between stimulus chemicals and transmembrane sensors . We can compare this to metabolism-based behavior , which responds not directly to environmental phenomena but to the combined effects of environmental phenomena upon the metabolism . This indirect sensitivity to the environment makes it a good candidate for integrating different stimuli and producing the appropriate response ( move toward or away ) . Goldstein and Soyer [26] acknowledge this issue but their model does not address any integrative phenomena—their simulation results correspond only to a single attractant gradient scenarios . Despite its simplicity , the model presented here is able to effectively integrate information from multiple gradients in a straightforward manner ( Experiment 6 ) . By being sensitive to the production of , it integrates the effect of all environmental features upon metabolic rate . As an example of the potential integrative power of metabolism-based chemotaxis , imagine two compounds , and , each of which acts as a metabolic toxin when encountered on its own . But , when encountered together , they act as excellent metabolic resources . Metabolism-based chemotaxis would respond appropriately ( move towards when encountered together and away from or when either is encountered on its own ) , while metabolism-independent chemotaxis would require the evolution of considerable specific machinery to accomplish the same appropriate behaviors . A further development of the model presented here ( see [56] ) has allowed us to explore the potential of metabolism-based mechanisms to produce gradient-climbing strategies , in particular we have shown how a single new reactant could turn a network of metabolic reactions that produces the selective-stopping behavior into one that produces the more intricate gradient-climbing behavior . We have also designed a scenario where a simulated protocell incorporates a new attractant from the environment into its metabolism and becomes chemotactic towards it . We have used the above experiments to explore theoretically the potential of metabolism-based chemotaxis , the feedback between metabolism and behavior , to bootstrap and accelerate early evolutionary processes [56] . Among the further extensions , a very interesting development would be to study which situations are conducive to which relationships between metabolism and chemotaxis and how transitions from one form to the other could occur . In particular , artificial co-evolution of metabolic networks and behavioral mechanisms could help address questions regarding a ) the likelihood of metabolism-dependent or metabolism-independent chemotaxis under various environmental conditions , b ) the possibility of metabolism-independent chemotaxis arising from metabolism-dependent precursors and c ) how both types of chemotaxis might co-exist with a varying degree of influence . Despite the considerable advances that the segregated one-compound-one-response approaches to chemotaxis have provided so far , it is perhaps time to start integrating not only metabolism into the picture but richer and varying environments where metabolic modulation might be playing a more relevant behavioral role . We hope to have shown that such an integrative move does not necessarily require the inclusion of an overwhelming level of detail , but might instead be effectively dealt with by metabolism-based forms of regulation . Many aspects of metabolism can potentially be abstracted away to reproduce the higher order dynamics of complex metabolic networks and then coupled to a behavioral mechanism . Moving in this direction opens the space for interactions between internal and environmental chemical dynamics that are not reducible to the influence of environmental compounds upon transmembrane receptors . From the reported experiments we can generalize that , despite its simplicity , metabolism-based chemotaxis allows for an ongoing evaluation of environmental conditions . This evaluation is indirect in that behavior is not in response to the environment , but rather to the influence of the environment upon the metabolism . The ongoing and indirect nature of metabolism-based chemotaxis makes possible an automatic and appropriate response to a variety of encounters with environmental conditions that have never been experienced by the bacterium , nor even by its evolutionary ancestors , for it is not necessary to evolve trans-membrane sensors that interact in specific ways with each environmental influence . The evaluation of the environment is accomplished by the influence of the metabolism . These generalizations should be further examined both by empirical studies and elaborations of the current model . We would like to stress that the current model plays the role of a proof of concept by allowing us to see the possibility of metabolism-based chemotaxis at work and unveil some implications . As variations of the model start to address more specific issues , they will have to incorporate more realistic assumptions such as energetic requirements for movement , biomechanics , differences in timescales between behavior and metabolism , and potential interactions between optimal behavioral control , metabolic dynamics and stochasticity . Also required is a study of the parametrical robustness of the phenomena reported here . | Traditionally , bacterial chemotaxis has been treated as metabolism-independent . Under this assumption , dedicated chemotaxis signalling pathways operate independently of metabolic processes . There is however , in various strains of bacteria , growing evidence of metabolism-dependent chemotaxis where metabolism modulates behavior . In this vein , we present the first model of metabolism-based chemotaxis that accomplishes chemotaxis without transmembrane receptors or signal transduction proteins , through the direct modulation of flagellar rotation by metabolite concentrations . The minimal model recreates chemotactic patterns found in bacteria , including: 1 ) chemotaxis towards metabolic resources and 2 ) away from metabolic inhibitors , 3 ) inhibition of chemotaxis in the presence of abundant resources , 4 ) cessation of chemotaxis to a resource due to inhibition of the metabolism of that resource , 5 ) sensitivity to metabolic and behavioral history and 6 ) integration of simultaneous complex environmental “stimuli” . The model demonstrates the substantial adaptability provided by the simple metabolism-based mechanism in the form of an ongoing , contextualized and integrative evaluation of the environment . Fumarate is identified as possibly playing a role in metabolism-based chemotaxis in bacteria , and some consequences of relaxing the metabolism-independent assumption are considered , causing us to reconsider the categorization of environmental compounds into “attractants” or “repellents” based solely on their binding properties . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [
"microbiology/microbial",
"physiology",
"and",
"metabolism",
"cell",
"biology/chemical",
"biology",
"of",
"the",
"cell",
"computational",
"biology/metabolic",
"networks",
"computational",
"biology/systems",
"biology"
] | 2010 | A Minimal Model of Metabolism-Based Chemotaxis |
Under natural conditions , animals encounter a barrage of sensory information from which they must select and interpret biologically relevant signals . Active sensing can facilitate this process by engaging motor systems in the sampling of sensory information . The echolocating bat serves as an excellent model to investigate the coupling between action and sensing because it adaptively controls both the acoustic signals used to probe the environment and movements to receive echoes at the auditory periphery . We report here that the echolocating bat controls the features of its sonar vocalizations in tandem with the positioning of the outer ears to maximize acoustic cues for target detection and localization . The bat’s adaptive control of sonar vocalizations and ear positioning occurs on a millisecond timescale to capture spatial information from arriving echoes , as well as on a longer timescale to track target movement . Our results demonstrate that purposeful control over sonar sound production and reception can serve to improve acoustic cues for localization tasks . This finding also highlights the general importance of movement to sensory processing across animal species . Finally , our discoveries point to important parallels between spatial perception by echolocation and vision .
Animals experience complex and noisy sensory information when navigating in the natural environment , and they must employ strategies to selectively sample and process biologically relevant signals in dynamic contexts [1 , 2] . One strategy for parsing complex sensory information is “active sensing , ” which engages motor systems in the acquisition of sensory information [2 , 3] . Behavioral active sensing , the focus of the research presented here , can take two forms: ( 1 ) the generation of signals which the animal uses to probe the environment and ( 2 ) the purposeful orienting of peripheral sensory organs towards a selected target [2–4] . In humans and other animals that rely heavily on vision , active sensing involves orienting of gaze toward an object of interest to gain higher resolution of the visual scene [2 , 3 , 5] . Humans also employ behavioral adjustments during auditory localization in the form of head movements [6 , 7] , demonstrating that active sensing is a generalized strategy for gathering sensory information . The echolocating bat presents a powerful model to investigate active sensing for auditory spatial localization and tracking , as this animal adaptively controls the generation and reception of the sensory signals it uses to represent the environment through adjustments to sonar call production and echo acquisition on a very rapid timescale [8 , 9] . It is well established that bats of different species adapt the features of echolocation calls as they detect , track , and intercept prey [9–11] , but comparatively little is known about the active orienting of the head and ears to improve sonar target detection and localization [12 , 13] . Early research in the horseshoe bat—a species that emits long duration ( ~50–200 ms ) , constant frequency ( CF ) calls , in combination with brief frequency modulated ( FM ) sweeps—demonstrated that they time the sweeping movements of the pinna in the anterior/posterior direction with the arrival of an echo from a stationary target [14] . By moving the pinna during echo reception , the bat induces amplitude modulations and Doppler shifts in opposing directions at the two ears , amplifying inter-aural differences in the features of echoes [15 , 16] . The big brown bat ( Eptesicus fuscus ) emits short duration ( 0 . 5–15 ms ) , FM down-sweeps and successfully forages in both open and cluttered spaces and in the presence of nearby conspecifics [11 , 17 , 18] . These behaviors require fine control over sonar signal production and echo reception to extract biologically relevant signals from streams of acoustic stimuli . Here , we employed a novel method to measure in precise detail the dynamic coordination of sonar signal timing with head and ear movements as the bat tracked and intercepted prey from a platform . Our data demonstrate that the bat temporally synchronizes sonar vocalizations with changes in the 3-D orientation of the head and external ears to enhance cues for target detection and localization . Specifically , we discovered that the bat “waggles” the head to alter the relative elevation of the ears , while also changing the separation between the tips of the ears as it engages in target tracking . We hypothesize that the tight temporal coupling between adaptive sonar vocalizations and head/ear positioning is integral to high-resolution acoustic imaging by sonar and , more generally , applicable to sensory capture through active sensing . Our findings also reveal parallels between active vision and echolocation that point to general principles of spatial localization across sensory systems .
Three big brown bats ( E . fuscus ) were trained to track moving prey items from a platform ( experimental setup in Fig 1A and S1 Movie ) . By using this experimental paradigm , the position and movement of the insect target could be precisely controlled with respect to the bat , and detailed measurements could be made of the bat’s ear and head positions . Example oscillograms of sonar pulse sequences produced by the bat as it tracked targets with four distinct motion trajectories ( see Table 1 ) are shown in Fig 1B ( one-target simple , two-target simple , two-target pass , and one-target complex motions ) . One-target simple involves forward motion of a single tethered insect towards the bat’s position on the platform , and two-target simple involves the simultaneous forward motion of two insects on adjacent , parallel strings ( Fig 1A ) . In the two-target pass condition , one slower moving target initially approaches the bat and a second target starts 0 . 5 s later , overtaking the first around 1 m in front of the animal . In the complex target motion condition , a single tethered insect approaches the bat and then recedes twice before finally moving forward to reach the bat . ( see Table 1 for specific motion parameters ) . Note that each sonar vocal sequence exhibits a decreasing pulse interval with decreasing target distance and the stereotyped “buzz phase” just before target interception [8 , 17 , 19 , 20] . As reported for bats foraging on the wing [11] , animals here exhibited adaptive modifications in sonar acoustic features with changing target distance ( Fig 1C ) . It is also noteworthy that bats decreased pulse interval when tracking two targets , as compared with a single target ( Fig 1C , top; S1 Fig ) . In addition to pulse interval , bats showed range-dependent adjustments in sonar pulse duration to avoid an overlap in time between sonar pulse emission and echo arrival ( Fig 1C , bottom ) [8 , 15] . Furthermore , the bats showed consistent modifications in pulse duration , which varied with target motion condition ( Fig 1C , bottom; S2 Fig ) . Specifically , bats increased call duration when tracking two targets , as compared to one ( S2B Fig ) . Longer call durations result in echo-echo overlap from the two adjacent targets , and previous research has shown that bats can resolve echo-echo overlap from spectro-temporal information contained in the composite sonar returns [21] . Bats also decreased call duration when tracking one-target complex motion , as compared to one-target simple motion ( S2A Fig ) , demonstrating how the specific target motion trajectories can influence sonar vocal production . In addition to adaptations of individual sonar pulses , it has been shown that bats modify the temporal dynamics of pulses into what has been termed “sonar sound groups” ( e . g . , reference [22] ) . Sonar sound groups are defined as clusters of two or more pulses separated by a short interval and flanked by pulses at longer intervals ( S3A Fig ) . The incidence of sound group production varies with sonar task demands , and prior work demonstrated that bats increase the production of sonar sound groups when localization demands are high [22–26] . In agreement with these earlier results , we found that bats in the current study increased sonar sound group production during the more difficult tracking tasks of two-target-pass and one-target complex motion , as compared to one- and two-target-simple motion ( S3B Fig , t-test , p < 0 . 0001 , df = 118 for one-target complex/one-target simple; t-test , p < 0 . 0001 , df = 161 for two-target pass/one-target simple ) . Additionally , as reported previously [22] , bats in the current study produced sonar sound groups while tracking a target from a stationary position . This behavior demonstrates a decoupling between the coordination of sonar sound group production and the wing beat cycle of an echolocating bat ( see , for example , references [27] and [28] ) .
Using a novel method to investigate adaptive motor behaviors of echolocating bats engaged in target tracking , we report here that animals rapidly and tightly coordinated sonar signal timing with ear and head movements in relation to target distance and motion trajectory . Bats also showed adjustments in sonar pulse duration and interval , as well as the production of sonar sound groups , with the distance and trajectory of the target . The bat’s adaptive echolocation behavior was accompanied by rapid changes in head orientation and external ear ( pinna ) position , which collectively influence the acoustic cues used to localize and track the moving target ( s ) [9 , 11] . These findings demonstrate that auditory feedback in the echolocating bat drives very rapid motor adjustments in both sonar call production and head/ear movements on a millisecond timescale . In the current study , bats increased sonar pulse rate and call duration when tracking two targets compared to a single target . Furthermore , bats increased the production of sonar sound groups while tracking the target in the complex motion condition , an adaptive echolocation behavior that is hypothesized to increase sonar localization accuracy [9 , 11 , 22–26] . Through these adjustments in sonar vocal production , the bat directly controls the sensory information available to the auditory system during target localization and tracking . Importantly , bats also produced head waggles after the production of sonar sound groups . It has been shown in humans that movements of the head and body , active or induced , result in significantly increased acuity of sound localization [30–32] . Head waggles in the bat change the relative elevation of the two external ears , providing changes in the binaural cues carried by echo streams [33] . Specifically , the head waggles serve to modify and enhance interaural level difference ( ILD ) changes , as well as inter-aural spectral differences , of successively returning echoes from a sonar sound group . Indeed , the angular offset of the pinnae during head waggles alters the vertical positions of the external ear by up to four degrees . Previous head-related transfer function ( HRTF ) measurements in the big brown bat report a prominent spectral notch that decreases in frequency with sound source elevation [34 , 35] . In these studies , the spectral notch shifts by approximately 5 kHz for each 5-degree step in sound source elevation around the horizon , providing an indication of large differences between power spectra of echoes received at the two ears during a head waggle . Previous research in humans report that “rhythmic sampling” of the environment increases acoustic resolution by allowing the brain to anticipate and prepare for future sensory updates [6 , 7] . The echolocating bat’s production of sonar sound groups ( sonar sounds with regular intervals ) results in cascades of echoes returning at regular intervals [22] , potentially allowing the bat to anticipate the temporal patterning of echo returns . Timing head waggles across a series of echo returns could provide the bat with contrasting echo information , improving the distance and direction localization of a moving target in a manner similar to motion parallax in visually guided animals [36] . For visual motion parallax , small movements of the eyes have been found to accentuate depth perception and the tuning curves of neurons in visual area 5 , also known as middle temporal ( MT ) [37] . Using head waggles , the bats can similarly induce a related perceptual phenomenon via acoustic parallax . Acoustic parallax has been hypothesized to play a role in distance perception of auditory objects: a near auditory object moving across the horizon has a higher rate of ILD changes than a far object [38 , 39] . In this way , auditory objects at different ranges are tagged with specific changes in ILD over time . It has been reported that head movements by human listeners similarly assist in the segregation of multiple auditory streams by tagging each stream with specific inter-aural differences [40] , and that these movements contribute to distance perception [41] . We propose that the coupling sonar sound groups with head waggles serves to increase auditory localization accuracy in bats , much in the same way as rhythmic sampling and motor-induced ILD change hone auditory processing in humans . Echolocating bats tracking targets in the current study also adjusted the pinna position with respect to the head by altering the separation between the tips of the outer ears . These adjustments varied with target distance , changes in target trajectory , and the number of targets being tracked . In general , as a target approached , the separation between the bat’s pinnae significantly increased . We hypothesize that control over pinna separation aids in both target detection and localization . At greater target distances , smaller inter-pinna separations may aid in the collection of weak echo returns . Previous work on the greater horseshoe bat , a species that uses long duration , constant frequency ( CF ) sonar signals , suggests that a more erect pinna position ( decreased inter-pinna separation ) , increases the bat’s sensitivity to echoes returning along the midline , whereas lowered pinna tips ( increased separation ) boost sensitivity at more peripheral locations [12] . Big brown bats that use frequency modulated ( FM ) signals may similarly enhance echo detection by directing the pinnae at the higher signal to noise ratio ( SNR ) components of echoes along the midline when target distance is larger [42] . At close target range , however , the bat’s acoustic field of view must be broad in order to ensure that the target remains in its spatial processing window , and increased inter-pinna separation serves to increase the bat’s acoustic field of view . Increasing the separation and flattening the pinna at shorter target distances also supports sonar localization by augmenting ILD and interaural time difference ( ITD ) cues [12 , 43] . Although the small size of the bat’s head limits its range of ITD , sound level differences at the ears effectively boost inter-aural time differences through time-intensity trading [43] . By flattening the pinna positions , bats alter ILD values in ways that could further enhance echo localization . In addition to the effects of target distance on inter-pinna separation , we observed that bats alter inter-pinna separation when tracking multiple targets . Bats in the current study decreased inter-pinna separation when tracking two targets as compared to one . Holding the pinna at a more upright and narrow orientation may also aid in the segregation of echoes from multiple targets by limiting the acoustic field of view to a more restricted azimuthal angle [12] . By altering the inter-pinna separation with target distance , the bats ultimately influence the acoustic information used to represent their acoustic scene [12] . Bats also dynamically modulate the sampling volume of their acoustic scene by altering the spectral content of their vocalizations [44] . Lower-frequency sounds are less directional than higher-frequency sounds , and when bats shift the frequencies of their vocalizations , they can increase or decrease their acoustic field of view [44] . Previous research has shown that bats typically downshift the frequency band of their vocalizations in the terminal buzz phase , an adjustment that increases their sonar field of view in the final attack phase of insect pursuit [45] . In our study , bats showed a similar downshift in the frequency of vocalizations with decreasing target distance ( S9 Fig ) while simultaneously increasing inter-pinna separation . By performing these coordinated behaviors , bats can correlate the width of the sonar beam with the width of their acoustic sensing . At larger target distances , the bats emit high-frequency , directional vocalizations while keeping inter-pinna separation narrow to focus hearing sensitivity along the midline . As the target approaches , sonar call frequency decreases and the sonar beam widens , while an increase in inter-pinna separation increases sensitivity to more peripheral locations . In this way , the acoustic view during echolocation is determined by both the change in the outgoing signal as well as a change in the positions of the external ears , and bats modify these parameters to meet specific task demands . It is noteworthy that big brown bats rapidly coordinated small deflections of the pinnae around the time of sonar vocalizations and echo arrival . Previous research has reported this behavior in CF bats echolocating a stationary target [14 , 16 , 46] . These bats produce long sonar vocalizations and enhance inter-aural differences by coordinated pinna movements , which serve to increase localization accuracy [13] . Our work extends this finding to bats that use very brief FM sonar signals ( 0 . 5–4 ms in this study ) to track a moving target . When tracking a moving target , the bat must continuously update information about target position from dynamic echo returns [9 , 47] , and our data suggest that motor programs for vocal production and pinna deflections are tightly coordinated to support high-resolution sonar localization . Through the synchronized timing of sonar vocalizations and pinna/head movements , the bat influences the cues for sonar localization to enable a highly precise and flexible distal sensing system . By quantifying the bat’s natural orienting behaviors in a novel target-tracking task , we discovered that echolocation shares with both human vision and audition general solutions to object localization in 3-D space . Much in the same way that humans move their eyes to foveate an object or move their heads to accentuate information about the position of an auditory or visual object , the bat controls sensory sampling through motor actions . The results of this study point to the importance of coordinated timing in active sensing systems that contribute to high-resolution spatial localization . The findings reported here not only hold importance to the field of systems neuroscience but also for engineering applications that implement adaptive controls for robotic sensory systems .
Three wild-caught big brown bats ( E . fuscus ) served as subjects in the following behavioral studies . The bats were collected in the state of Maryland under a permit issued by the Department of Natural Resources and were housed in an animal vivarium at the University of Maryland , College Park . All procedures employed were approved by the University of Maryland’s Institutional Animal Care and Use Committee ( IRBnet Protocol 413006–4; UMD Protocol R-13-04 ) . The bats were trained to rest on a platform and track a moving , tethered prey item ( mealworm ) using echolocation ( Fig 1A ) . The tethered prey item was suspended from a four-cornered loop of monofilament line that was connected to a set of four pulleys and a rotary stepper motor ( Aerotech ) . The velocity , acceleration , and deceleration parameters could be set independently and were controlled via a computer interface with custom software written in Matlab 2012b . By spinning the rotary motors , the target could be moved in multiple directions , velocities , and accelerations in front of the bat . Bats were initially trained by pairing a sound presentation with the delivery of a food reward immediately in front of the animal . Once the bat learned the association between the sound and food presentation , the starting position of the food reward , the time delay of its delivery , and the velocity of its travel were slowly increased as the animals learned to track the tethered target . The velocity of the target was steadily increased until it moved at 4 m/s , a velocity similar to the speed of a flying bat intercepting prey on the wing [48] . The final starting position of the target was 2 . 5 meters from the animal’s resting position on the platform . Catch trials were introduced to keep the bat actively engaged in the target tracking task rather than passively listening to sounds of the motors driving the target towards the bat . The catch trials involved suspending the target from a location adjacent to the target starting position on the loop of fishing line . The experimental apparatus therefore made the same sounds when propelling the target , but these sounds were decoupled from the target motion by having the target remain stationary at the starting position throughout the trial . Once the bat was trained on the task , approximately 25% of all trials were catch trials . Data collection began after animals exhibited reliable changes in pulse rate , pulse duration , and the production of the terminal “buzz phase” as the target approached . During the training phase , the target moved in one direction towards the animal at a velocity of 4 m/s . Once the bat reliably tracked the approaching target , more complicated target motions were presented to the animal , along with the addition of a second moving food reward . In total , four different types of target motion were presented to the animal ( Fig 1B , Table 1 ) , along with catch trials . The bat’s echolocation behavior , head aim , and pinnae movements were recorded as the animal tracked the moving tethered target . Time synchrony across video , audio , and other hardware devices was achieved through the generation of a TTL pulse by the computer controlling the stepper motor system . Data acquisition was collected on all systems starting with the TTL pulse at the beginning of target motion , and then continued for 10 s . This provided enough time for all the different target motion conditions to be completed with a buffer of at least 2 s . Sonar vocalizations were recorded with an ultrasonic microphone manufactured by Ultra Sound Advice ( SM2 ) connected to a pre-amplifier ( SP2 ) . The microphone was placed directly above the path of target motion and at a distance of 2 . 5 meters from the bat . The vocalizations were filtered between 15 and 110 kHz ( Kemo VBF44 ) before being digitized at 250 kHz by a National Instruments M-series data acquisition board and saved on a hard drive for offline analyses . Head and pinnae positions were tracked using a Vicon Nexus Motion Tracking System . Four T-40s series cameras were mounted above the bat on a custom built railing system . The cameras each had 18 mm lenses and were positioned to maximize overlap of each camera’s view of the platform area where the bat was tracking the target . The frame rate was set to 500 Hz and calibrated to sub-millimeter accuracy in XYZ position measurements with a moving wand-based calibration method . Three-millimeter diameter IR reflective markers were then affixed to the bat’s ears and nose-bridge using spirit gum . The positions of the reflective markers were tracked across the 2-D views of each Vicon camera and reconstructed in 3-D by the Vicon Nexus software ( version 1 . 85 ) . The x-dimension was set as left and right motion on the platform ( perpendicular to target motion ) , the y-dimension was set as front to back motion , and the z-dimension was set as vertical motion . In summary , the bat’s vocalizations were recorded at a sampling rate of 250 kHz; the target’s position was measured at mm/ms; and the 3-D position of the head and ears was measured at 2 ms intervals and interpolated linearly to 1 ms intervals at a sub-millimeter spatial resolution . Analysis of the 3-D positions of the head and pinnae was pre-processed using the Vicon Nexus platform . The positions of the left ear , right ear , and head markers were labeled in all frames where they were visible . Once the individual files were labeled , they were imported into Matlab 2012b . Matlab was first used to spline fill any gaps in the labeled marker trajectories . Gaps longer than 50 ms were not interpolated and were excluded from further analyses . Once the marker trajectories were post-processed , the Euclidian distances in three dimensions between the two tips of the outer ear pinnae markers and the rotational movements of the head were calculated . The inter-pinna separation over time was computed by measuring the absolute distance in 3-D space between each tip of the two pinnae . The rotational movement of the head was measured for periods of time when one pinna was moving down in the z-dimension while the other ear pinna was moving in an upwards direction for at least 20 ms . We term these asymmetrical rotational movements “head waggles , ” and they were further confirmed by measurements showing that each pinna was also moving in opposite directions with respect to the head marker . The analysis of the sonar vocalizations involved several steps , all of which were performed in Matlab 2012b . We first determined the onsets and offsets of each vocalization by drawing an amplitude threshold through a low-passed , rectified version of the original oscillogram . The onsets and offsets were then corrected for the travel time from the bat to the microphone ( 2 . 5 meters in front of the bat ) . Sonar pulse interval was calculated by taking the difference in time between onsets of successive sonar vocalizations , and sonar pulse duration was defined as the time between onset and offset time of each pulse . Echo arrival time was calculated by determining the distance of the target at the onset of each sonar vocalization and then using the speed of sound to estimate the travel time from the bat’s mouth to the target and then back to the bat’s ear . On each day , the temperature and humidity of the experimental room was recorded to accurately determine the speed of sound calculation . An ultrasonic microphone was also placed below the bat to record returning sonar echoes . The sensitivity of this microphone precluded recordings of sonar echoes returning from targets at larger distance , however . As a result , all echo arrival times were calculated as described above , and then these times were validated by examining when an echo was received on the microphone placed under the bat . Once the times of the vocalizations were computed , the vocal sequences were analyzed to extract “sonar sound groups . ” Sonar sound groups were determined by a statistical criterion defined in earlier work [9 , 22] . Briefly , sonar sound groups of three or more vocalizations were identified by a consistent pulse interval across at least three sonar vocalizations ( within 5% error of the mean PI of the sound group ) , with the pulse interval of the surrounding calls at least 1 . 2 times larger than the mean sonar sound group pulse interval . For sonar sound groups of two vocalizations , the pulse interval of the flanking calls of the doublet sonar sound group had to be 1 . 2 times longer than the pulse interval between the two calls within the sonar sound group . In this way , the term “sonar sound group” identifies vocalizations that are in relative temporal isolation , and at shorter pulse intervals , from the surrounding sonar vocalizations . The data on target motion , pulse timing , echo arrival , and sonar sound group production were then analyzed with respect to the pinnae and head movement data . The distance between the pinnae was measured as the bat tracked the moving target; the rotation of the head was analyzed with respect to target motion and sonar sound group production; and the timing of each vocalization and echo return was correlated with the timing of small pinna deflections . These small pinna movements are distinct from the global changes in inter-pinna distance that occur as the target approached . “Local peaks” in pinna separation were not observed when the target was in close proximity to the bat and sonar pulse rate was greater than 25 Hz , so this data was not included in subsequent analyses . In order to calculate statistically significant differences between pulse interval and pulse duration across target motions , a dʹ statistic was calculated across target distance for all pair-wise comparisons of target motion . The dʹ statistic is a measure of discriminability between two time-varying signals . A random sampling method was then used to construct 95% confidence intervals ( CI ) where data from each group were pooled , randomly sampled into two groups , and then a dʹ statistic computed from the random permutation . This was repeated 1 , 000 times , and a distribution of randomly sampled dʹ statistics was constructed to determine the 95% CI . Pulse interval and pulse duration values were binned into 10 cm bins with respect to target distance . A dʹ statistic was also used to determine time points of significant differences in inter-pinna separation for different motion conditions . As noted above , the dʹ statistic is a measure of discriminability and , in this case , the inter-pinna separation measurements . In order to determine 95% CI intervals , or when two data sets were significantly different , data were pooled and randomly shuffled into two groups , and a dʹ statistic was calculated across the randomly shuffled groups . This was performed 1 , 000 times , resulting in a distribution of dʹ values calculated across the randomly shuffled groups . The 95% confidence intervals on the dʹ statistic were determined from this distribution . For any instance of multiple comparisons , a Bonferroni correction was performed to determine the appropriate p-value . | As animals operate in the natural environment , they must detect and process relevant sensory information embedded in complex and noisy signals . One strategy to overcome this challenge is to use active sensing or behavioral adjustments to extract sensory information from a selected region of the environment . We studied one of nature’s champions in auditory active sensing—the echolocating bat—to understand how this animal extracts task-relevant acoustic cues to detect and track a moving target . The bat produces high-frequency vocalizations and processes information carried by returning echoes to navigate and catch prey . This animal serves as an excellent model of active sensing because both sonar signal transmission and echo reception are under the animal’s active control . We used high-speed stereo video images of the bat’s head and ear movements , along with synchronized audio recordings , to study how the bat coordinates adaptive motor behaviors when detecting and tracking moving prey . We found that the bat synchronizes changes in sonar vocal production with changes in the movements of the head and ears to enhance acoustic cues for target detection and localization . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"acoustics",
"vocalization",
"medicine",
"and",
"health",
"sciences",
"engineering",
"and",
"technology",
"ears",
"vertebrates",
"animals",
"mammals",
"animal",
"signaling",
"and",
"communication",
"echoes",
"animal",
"behavior",
"computer",
"vision",
"remote",
"sensing",
"zoology",
"computer",
"and",
"information",
"sciences",
"behavior",
"bioacoustics",
"target",
"detection",
"head",
"physics",
"anatomy",
"bats",
"biology",
"and",
"life",
"sciences",
"physical",
"sciences",
"amniotes",
"organisms",
"sonar"
] | 2016 | Action Enhances Acoustic Cues for 3-D Target Localization by Echolocating Bats |
Small-molecules that inhibit interactions between specific pairs of proteins have long represented a promising avenue for therapeutic intervention in a variety of settings . Structural studies have shown that in many cases , the inhibitor-bound protein adopts a conformation that is distinct from its unbound and its protein-bound conformations . This plasticity of the protein surface presents a major challenge in predicting which members of a protein family will be inhibited by a given ligand . Here , we use biased simulations of Bcl-2-family proteins to generate ensembles of low-energy conformations that contain surface pockets suitable for small molecule binding . We find that the resulting conformational ensembles include surface pockets that mimic those observed in inhibitor-bound crystal structures . Next , we find that the ensembles generated using different members of this protein family are overlapping but distinct , and that the activity of a given compound against a particular family member ( ligand selectivity ) can be predicted from whether the corresponding ensemble samples a complementary surface pocket . Finally , we find that each ensemble includes certain surface pockets that are not shared by any other family member: while no inhibitors have yet been identified to take advantage of these pockets , we expect that chemical scaffolds complementing these “distinct” pockets will prove highly selective for their targets . The opportunity to achieve target selectivity within a protein family by exploiting differences in surface fluctuations represents a new paradigm that may facilitate design of family-selective small-molecule inhibitors of protein-protein interactions .
Selectivity of a compound for its desired protein target—or targets—is an important property optimized in the course of small-molecule drug discovery [1] . Some diseases , such as chronic myeloid leukemia , can be traced to dysfunction of a single protein target ( BCR-ABL ) ; in such cases , drugs such as imatinib are sought to act selectively against that target [2] . Conversely , a number of drugs ( such as sunitinib and chlorpromazine ) have proven exceedingly successful because they act on multiple targets [3 , 4]; this has led to increased interest in “polypharmacology” to address complex disease states such as cancer and psychiatric conditions [5] . A clear downside of compound promiscuity , however , is the potential for adverse effects ( toxicity ) due to interactions with unrelated proteins , or even interactions with other proteins in the same family as the target [6] . One recent cautionary example is ABT-263 ( navitoclax ) , a Bcl-2 inhibitor that exhibited a dose-limiting adverse effect ( thrombocytopenia ) stemming from its inhibition of Bcl-xL [7 , 8] . Tuning the selectivity of a lead compound can sometimes be carried out by exploiting differences in shape and electrostatics between target and off-target proteins , using insights from structure-activity relationships ( SAR ) [9] or structural biology [10–12] to focus on features that will prevent specific undesirable interactions . While determinants of selectivity have been carefully mapped in a number of “traditional” drug targets , such as kinases , this has not yet been the case for emerging classes of “non-traditional” drug targets , most notably small-molecule inhibitors of protein-protein interactions . Though extensive efforts to identify inhibitors of protein interactions have only recently begun to bear fruit , structural analysis of available examples has revealed that often the inhibitor-bound protein is captured in a conformation that is distinct from both the unbound and protein-bound conformations [13] . In such cases , the unbound and protein-bound conformations could not have served to rationalize binding of the inhibitor its desired target , let alone explain selectivity against other members of the protein family . We recently described an approach for biased exploration of protein fluctuations , in order to better sample pocket-containing conformations at protein interaction sites [14] . We found that when starting from an unbound protein structure , we observe low-energy conformations that include deep surface pockets at druggable sites but not elsewhere on the protein surface . These findings supported a “conformational selection” model [15] , whereby inhibitors recognize low-lying excited states of the protein that are naturally visited under physiological conditions . A natural implication of this conformational selection model is that the particular variety of pocket shapes visited by the protein surface will dictate the regions of chemical space in which complementary inhibitors can be found: this would have clear implications for designing new inhibitors . Conversely , one may instead view this complementarity from the perspective of the ligand: under this model , an inhibitor is expected to be active against a given protein if , and only if , the protein surface includes a suitably complementary pocket among those that are sampled under physiological conditions . Within a protein family , then , only the subset of family members that sample the corresponding pocket will be inhibited by this compound . The Bcl-2 protein family in humans is comprised of about 25 members that together regulate apoptosis through a series of protein-protein interactions that induce either pro-death ( Bax , Bak , and others ) , pro-survival ( Bcl-xL , Bcl-2 , Mcl-1 , Bcl-w , and others ) , or derepression/sensitizing ( Bid , Bim , and others ) activity [16] . The critical step of disregulating apoptosis in tumor maintenance and chemoresistance has made certain members of this family exceptionally attractive targets for therapeutic intervention for many years , and in a variety of cancers [17] . Despite the overall dearth of examples of small molecules that inhibit protein interactions , a number of such compounds have been reported against assorted members of the Bcl-2 family . Further , selectivity of these inhibitors against a panel of Bcl-2 family members have also been reported in some cases . These data , together with the availability of multiple experimentally-derived structures of inhibitor-bound complexes , position the Bcl-2 family as a rich model system in which we can explore the determinants of selectivity for small-molecule inhibitors of protein-protein interactions . Here , we use the Bcl-2 family to explore whether the ensemble of pocket shapes sampled on protein surfaces can be used to explain ligand selectivity . We will use simulation to generate ensembles of pocket-containing conformations [14] , then directly probe whether a complementary pocket for a given ligand is present in the ensemble . If successful , this approach may allow us ultimately to predict , and by extension design , inhibitors with a desired selectivity profile; such compounds could serve both as a starting point for development of new therapeutics , and also as “tool compounds” to probe the underlying biological mechanism of disease .
Pockets were identified using the “pocket” protocol in Rosetta [14] . This is a local pocket detection approach adapted from the Ligsite algorithm [19] , and it uses geometric criteria to identify concave regions on a protein surface suitable for small-molecule binding . The approach entails building a small grid in the vicinity of one or two “target” residues , then mapping the van der Waals surface for the whole protein onto the grid; the exposed grid points are marked as “solvent . ” Next , the grid is screened to identify linear segments of solvent points that are bounded by protein points; the grid points along these segments are marked as “pocket” points . Pocket grid points are clustered into “pockets , ” and then any clustered pockets not in direct contact with the target residue ( s ) are discarded . The “deep volume” of a pocket is defined as the volume of the pocket that is well-sequestered from the solvent ( more than 2 . 5 Å from any solvent grid point ) . Generation of a representative protein surface pocket is demonstrated in Fig . 1 . A number of related approaches have been described for identifying potential small molecule binding sites on protein surfaces in this way [19–25]; the primary difference between these methods relative to the method described here is our use of “deep” pocket volume , which leads to improved discrimination when distinguishing known ligand binding sites from other shallow concave regions on the protein surface [14] . These “deep” pocket volumes are correlated to , but smaller than , pocket volumes identified by other representative methods [14]; in the context of this study , then , we expect that focusing our pocket description on regions in direct contact with the protein will allow the most critical features to be captured with minimal extraneous information . The selection of the target residue pairs is important in determining the location of the grid , and in turn may affect the resulting pocket shapes . To avoid biasing results towards any particular inhibitor—and to allow the method to be applied in cases for which no inhibitor has yet been identified—we did not use the structure of the protein in complex with any small-molecule ligand when defining the target residues . Instead , we developed an algorithm to select suitable target residues starting from the structure of the protein-protein complex . First , the “Robetta” server [26 , 27] is used to estimate the contribution to the binding free-energy of each interfacial sidechain ( ΔΔGres ) via computational alanine scanning . All pairs of interfacial residues are then exhaustively tested by building a 24 Å candidate grid placed at the residue pair’s center of mass , and summing ΔΔGres for each residue that falls on this grid . To ensure that the grid placement captures the key energetic contributors to the protein-protein interaction , the residue pair that captures the largest cumulative ΔΔGres is used in subsequent studies . In essence , this approach aims to align the pocket grid such that it is optimally close to the energetically dominant residues of the protein-protein interaction . This tool is implemented in the Rosetta software suite , and its use it demonstrated in the Supporting Methods section [18] . The effect of the particular target residues used will be examined further in the Results section . Ensembles of pocket containing conformations were generated using the “relax” protocol in Rosetta , which incorporates both backbone and side chain degrees of freedom in a Monte Carlo search . Starting from the standard Rosetta energy function , we added a term corresponding to the “deep pocket” volume multiplied by a proportionality constant . This biasing potential favors pocket-containing conformations that have more deep pocket volume , and thus drives sampling towards such conformations . In our previous work , we found that applying a proportionality constant of -0 . 25 Rosetta energy units per Å3 allowed identification of pocket-containing conformations with similar energies to those observed in unbiased simulations [14] . In the present study , therefore , we reused the same proportionality constant . A separate independent trajectory was used to generate each of the 1000 output conformations used for each protein included in our studies . It took about 5 minutes to generate each conformation on a modern CPU , though the independence of trajectories made it trivially scalable to multiple processors . To compare the shape and chemical similarity of surface pockets , we introduce the concept of an exemplar: a map of the “perfect” ligand that could complement a surface pocket , without the natural constraints of bond connectivity or chemistry . The lack of such physical requirements means that the exemplar does not correspond to any particular chemical entity , but instead simply consists of a collection of atoms in space . This collection occupies volume , and may include hydrogen bond donors and acceptors , but by construction does not imply any particular connectivity between the atoms . Exemplars are built from the “deep volume” that defines a protein surface pocket . First , the ideal location for a hydrogen bond partner of every donor/acceptor on the protein surface is determined; if this location lies within the pocket , then this region of the pocket volume is reserved for the polar group that will complement the protein surface . After placing these polar groups in the pocket , the remaining unoccupied volume is filled with hydrophobic ( carbon ) atoms using a greedy algorithm , such that the center of two atoms are no less than 1 . 7 Å apart . Exemplar points are then clustered together based on a proximity threshold of 5 Å , so that cases in which two small pockets flank the target residues are represented by a single exemplar . Generation of an exemplar from a representative protein surface pocket is demonstrated in Fig . 1; complete details of our implementation are included in S1 Text . These exemplars are analogous in philosophy to the popular pharmacophore maps used in medicinal chemistry to reflect consensus properties of known ligands [28 , 29]; while the latter approaches rely on mimicry of the existing binding partners [30–33] , however , exemplars are instead built purely from features of the protein surface . Recently , the utility of protein-based pharmacophores has been explored for pose prediction and virtual screening [34–36] , but such approaches have not been used for the comparison of pocket shapes as described below . Having represented the shape and chemical features of a protein surface pocket by its “perfectly complementary ligand” ( the exemplar ) , we are now positioned to quantitatively compare pockets using standard tools developed for comparison of chemical entities in 3D , provided that these tools do not require ( or assume ) knowledge of bond connectivity . A variety of ligand-based shape comparison methods have been developed for aligning pairs of molecules on the basis of volume overlap [37 , 38]; a convenient modern implementation of this approach is the ROCS software ( OpenEye Scientific Software , Santa Fe , NM ) [39 , 40] . ROCS represents molecular shape as the sum of Gaussians centered at each atomic position , and can rapidly calculate the near-optimal alignment that maximizes volume overlap between two molecules . Chemical groups are also included as separate Gaussians in the scoring/optimization step , to include electrostatic effects . The results we present below are built upon quantitative comparisons of the similarity in shape and hydrogen bonding patterns for pairs of protein surface pockets . In all cases we carry out this analysis by generating an exemplar for each of the two pockets , then using ROCS to align and score the overlap between these exemplars . We note that method to define pockets , and thus exemplars , makes use of a grid centered at a pair of target residues . To quantitatively examine the effect of the grid orientation on the exemplar similarity , we used four pocket-containing crystal structures and rotated each to 100 random orientations . Using one of these structures as a reference ( ABT-737 bound to Bcl-xL ) , we found that rotated variants of this structure produced similar exemplars , as did rotated variants of a related structure ( Bcl-xL in complex with a related ligand ) . In contrast , rotated variants of Bcl-xL bound to unrelated ligands , or rotated variants of Mcl-1 , produced highly dissimilar exemplars ( S1 Fig . ) . Collectively , these observations demonstrate that the orientation dependence of generating exemplars on a grid makes a negligible contribution to the exemplar distances we report below .
There have been dozens of reported inhibitors of Bcl-2 family members [7 , 8 , 41–59] , spanning a broad range of chemotypes . Unfortunately , experimentally-derived structures are not available for the majority of these inhibitors in complex with their cognate protein partner ( s ) ; this makes it difficult to gain insight into the detailed basis for molecular recognition in these cases . Instead , we began by compiling a comprehensive collection of all structures in the Protein Data Bank containing a Bcl-2 family member in complex with a non-fragment small-molecule inhibitor bound at the protein interaction site: these are listed in S1 Table . There are 28 such complexes , covering 26 unique inhibitors , and 3 different proteins are represented: Bcl-xL ( 14 structures ) , Bcl-2 ( 9 structures ) , and Mcl-1 ( 5 structures ) . The diversity of the inhibitors in this set is readily apparent from the Tanimoto similarity of fingerprints describing each chemical structure ( Fig . 2A ) . Unsurprisingly , groups of compounds that are similar by this measure typically represent a chemical series designed by a single research group ( e . g . compounds 1–3 from WEHI [60] ) . Given this collection of chemical scaffolds , we sought to ask how such diverse compounds could be recognized on the surface of a single protein family . There are two possibilities: either these compounds might adopt a shared three-dimensional structure not evident from their chemical structure ( i . e . a non-obvious example of “scaffold hopping” [61] ) , or else the protein surface must be sufficiently malleable to adopt different conformations when binding different ligands . To test whether these distinct compounds present a common structure to complement the protein surface , we carried out comparisons of the inhibitors’ shape and chemical features in their active ( bound ) conformations ( Fig . 2B ) . To examine the pattern of similarity between chemical structure and three-dimensional structure we sorted the all-vs-all Tanimoto scores for each set , and found a statistically significant non-zero Spearman correlation coefficient between these rankings ( p < 10-35 ) . This observation confirms that dissimilar chemical structures do not somehow adopt a shared three-dimensional structure . As expected , a given shape and pattern of hydrogen bond donors/acceptors is conserved within a chemical series , but not across different chemical scaffolds . To directly evaluate similarity of the conformations adopted by the protein to bind each ligand , we next carried out an analogous comparison using the exemplar derived from each inhibitor-bound pocket ( Fig . 2C ) . Here again we observe the same pattern of similarity , mimicking both that of the chemical structures ( p < 10-26 ) and their corresponding three-dimensional structures ( p < 10-20 ) . While each of the inhibitors within a given chemical class bind to a very similar pocket on the surface of the cognate protein , different chemical classes each take advantage of a dramatically different pocket on the protein surface . Moreover , the surfaces of different proteins bound to similar ligands ( e . g . Bcl-xL complex 7 vs Bcl-2 complex 19 ) resemble one another more closely than the surface of a single protein bound to chemically distinct ligands ( e . g . Bcl-xL complexes 3 vs 7 ) . These observations highlight the plasticity of this protein surface: multiple members of the Bcl-2 family can adopt similar conformations to bind a given ligand , yet a given protein can also form radically different surface pockets to accommodate different ligands . Together with the fact that the unbound structures of these Bcl-2 family members lack suitably deep surface pockets for inhibitor binding ( S3 Fig . ) , the observations presented here underscore the fact that molecular recognition in this protein family cannot be explained using a single conformation , but rather an explanation of ligand selectivity will instead require consideration of the many available conformations that this surface can adopt . To explore pocket-containing conformations of these proteins , we generated conformational ensembles using simulations in the presence of a biasing potential [14] . In essence , energy associated with the biasing potential in these simulations serves as a proxy for the binding energy of some ( unspecified ) ligand , which in turn serves to stabilize alternate conformations of the protein surface . The biasing potential takes account of the protein surface geometry ( “deep pocket volume” ) , but does not encode any information about the identity or features of any particular ligand . Thus , this approach allows efficient sampling of the conformational space available for protein reorganization in response to ligand binding . While other methods could have been employed to sample alternative conformations ( such as unbiased molecular dynamics simulations with retrospective analysis to identify pocket-containing conformations [62] ) , we chose this approach because it allowed us to rapidly generate a large ensemble of pocket containing conformations . As noted earlier , our approach defines the relevant protein surface on the basis of the protein/peptide interaction site , and not based on any known small-molecule inhibitor ( Methods section ) . In light of the fact that particular “target” residues are needed to produce the ensembles of pocket-containing conformations , we begin by examining the effect of the target residues on the resulting “pocket ensemble . ” Starting with the Bcl-2 family member Bcl-xL , we selected the two top-scoring pairs of target residues resulting from analysis of its peptide-bound structure . We then used each pair of target residues to generate separate ensembles of 2000 pocket-containing conformations , with each trajectory initiated from the unbound structure of Bcl-xL . To visually compare the conformational “pocket space” sampled by these two ensembles we built an exemplar from each pocket , and used multidimensional scaling analysis ( MDS ) to construct the two-dimensional projection that best reflects the pairwise distance between every pair of exemplars . The resulting map demonstrates that these two ensembles are strongly overlapping , and points to the robustness of the conformational space sampled to the particular target residues used ( Fig . 3A ) . In light of the similarity between the ensembles , we used conformations generated using the second-best scoring pair of target residues ( rather than the top-scoring pair ) , for consistency with our previous studies of Bcl-xL [14] . While application of the same approach for selecting target residues led to similar residues for most members of the Bcl-2 family , the residue pair selected on Mcl-1 was notably different . While this derived exclusively from different energetic contributions to the protein-protein interaction from each protein surface , it is interesting to note that known inhibitors of Bcl-xL and Mcl-1 bind at slightly different regions on the surface of these two proteins ( S4 Fig . ) . To ensure that the range of pockets across the Bcl-2 family would be fully captured in our studies , we used both the pair of residues derived from Bcl-xL and those derived from Mcl-1 to generate ensembles of pocket-containing conformations; in the analyses presented below , all such conformations are combined into a single ensemble regardless of the target residues used to generate them . To ensure that only physiologically relevant conformations were included in the subsequent analysis , we compared their energies to those obtained in equivalent unbiased simulations of the corresponding protein . Previously , we found that this method produced ensembles of pocket-containing conformations a slightly higher but overlapping distribution of energies than the unbiased ensemble [14] . With the caveats that conformations generated in our “pocket opening” protocol are not drawn from a Boltzmann distribution and the energy differences do not necessarily capture real differences in free energy , we instead simply collect pocket-containing conformations from the biased simulations that are within 15 Rosetta energy units of the unbiased ensemble . Thus , these represent conformations that are in principle available within the unbiased ensemble , but are simply not observed due to limitations of sampling . The distribution of energies observed in an unbiased simulation of each Bcl-2 family member is presented in Fig . 3B , along with the range of energies spanned by conformations from biased simulations . We next sought to compare the surface pockets on conformations comprising these ensembles to those pockets observed on experimentally-derived inhibitor-bound structures . We again turned to MDS analysis , this time including exemplars built from the experimentally-derived structures of the unbound , peptide-bound , and inhibitor-bound protein . For each of Bcl-xL , Bcl-2 , and Mcl-1 ( all family members for which structures of inhibitor-bound complexes are available ) , the resulting maps show that the pocket-containing conformational space sampled via simulation includes thorough coverage of experimentally-derived inhibitor-bound structures ( Fig . 4 ) . Though each simulation was initiated from either the unbound or the peptide-bound crystal structure , the resulting ensemble is notably distinct from these starting states; instead , the biasing potential drove sampling towards pocket-containing conformations that include examples very similar to the inhibitor-bound structures . In addition , each protein also samples a collection of conformations ( marked “D” ) with exemplars that differ from those observed in any available inhibitor-bound conformations: these will be discussed in more detail in the following section . In all three cases we find that these ensembles , generated without any prior information about the inhibitors , span the space of known inhibitors . This indicates that each protein is predisposed to adopt the particular pocket shapes observed in the corresponding inhibitor-bound structures: these are not conformational changes that are “induced” by the inhibitor , but rather these are among a suite of available conformations from which the inhibitor may select . Upon binding , meanwhile , smaller changes to the protein surface may then occur in response to particular features of the ligand ( such as reorientation of hydrogen bond donors and acceptors ) . To explore differences between ensembles of pocket-containing conformations of Bcl-2 family members , we carried out an analogous MDS analysis comparing exemplars from multiple family members; an important aspect of exemplar generation and comparison is that the description of a surface pockets is not tied to the sequence of the protein , allowing one to compare exemplars on the surfaces of different proteins . We compiled conformations generated from simulations of each Bcl-2 family member ( Bcl-xL , Bcl-2 , Mcl-1 , Bcl-w , Bax , Bid , and Ced-9 ) , and carried out MDS analysis using the complete set of exemplars ( Fig . 5A ) . Unsurprisingly , we observe that different family members sample different surface pockets; however , we do not observe such differences between the ( closely related ) human and mouse Mcl-1 sequences ( S6 Fig . ) . In these projections we again note that the ensemble generated for a given protein spans the majority of known inhibitors of that protein . Unsurprisingly , given the similarity we observed between inhibitor-bound Bcl-2 pockets and inhibitor-bound Bcl-xL pockets ( Fig . 2C ) , we find that most pockets observed in the inhibitor-bound structures occupy a similar region on this projection ( all except 1 , 5 , 6 , and 23 ) . Notably , many known inhibitors fall in a region of “pocket space” that is sampled by more than one protein . In the case of Bcl-xL , for example , many known inhibitors bind to pockets that are not only observed in simulations of Bcl-xL , but also in simulations of Bcl-2 . On this projection , these inhibitors ( all except 1 , 5 , and 6 ) reside in a region sampled by both Bcl-xL and Bcl-2 . Similarly , most of the Bcl-2 inhibitors ( all except 23 ) are in the same region sampled by both Bcl-xL and Bcl-2; in contrast , this other Bcl-2 inhibitor ( 23 ) overlaps with a regions sampled by Mcl-1 but not Bcl-xL . In addition to regions shared by more than one family member , each map contains a “distinct” region that is sampled exclusively by a single protein ( Fig . 5A , “D” ) . As expected , exemplars corresponding to these conformations are very different from those of the ( unbound or peptide-bound ) conformations from which the corresponding simulations were initiated ( Fig . 5B ) . Further comparison of the conformations themselves show that these alternate conformations are accessed primarily through concerted reorganization of the sidechains that comprise the surface pocket , though corresponding changes to the protein backbone—especially in the case of Mcl-1—are also required to enable this reorganization ( Fig . 5C ) . In summary , relatively modest structural changes to the protein conformation can produce radically different exemplars , and none of the inhibitors described to date bind to these particular protein conformations . To facilitate further study of these alternate protein conformations , we have made several such representatives publicly available on Proteopedia [63 , 64] ( http://proteopedia . org/wiki/index . php/User:John_Karanicolas/Selectivity_by_small-molecule_inhibitors_of_protein_interactions_can_be_driven_by_protein_surface_fluctuations , with the first model for each protein corresponding to the conformation we highlight here ) . The regions that we have described as “distinct” on these maps correspond to pockets that are only sampled by a single member of the Bcl-2 family . In other words , these “distinct” pockets are far from each of the pockets sampled by other family members . So far , we have identified these regions visually on the basis of a two-dimensional projection of “pocket space”; to avoid the loss of information associated with reduction of dimensionality , however , we can instead identify “distinct” pockets using exemplar distances directly . For each conformation comprising the ensembles used above , we found the exemplar distance of the closest pocket sampled by a different family member . To avoid describing rarely-sampled ( outlying ) conformations as distinct , we subtracted from this the exemplar distance of the closest pocket from one’s own ensemble . This measure , that we will call “distinctness , ” is largest for conformations taken from regions of pocket space that are well-sampled within a given ensemble , but not visited in the ensembles of the other family members . We evaluated the “distinctness” of each conformation in the ensembles presented above , and compiled the results for each family member into a histogram ( Fig . 6 ) . As expected , these results are consistent with our observations from the MDS analysis presented above: all of the pockets sampled by Bcl-w strongly resemble pockets from other family members and are therefore not “distinct , ” whereas Bcl-xL and Mcl-1 each sample certain highly distinct regions . Particularly striking from this analysis is the “distinctness” of experimentally-derived inhibitor-bound structures of Bcl-2 family members: none of these inhibitors take advantage of the “highly distinct” pockets available on the surfaces of Bcl-xL or Mcl-1 . Rather , each of these compounds targets a pocket that is sampled not only by the cognate binding partner , but also by at least one other member of the Bcl-2 family . As noted in the MDS analysis , the majority of known Bcl-xL inhibitors fall in a region of “pocket space” that is sampled by both Bcl-xL and Bcl-2 ( Fig . 5A ) . Given that Bcl-2 is found to form pockets that are similar to these inhibitor-bound Bcl-xL pockets , one might expect that these compounds would inhibit Bcl-2 in addition to Bcl-xL . Conversely , in light of the conformational selection model presented earlier , one would expect that the lack of similar pockets in the ensembles generated for other Bcl-2 family members might suggest that these proteins would not be inhibited by these compounds . To explore this idea , we measured the exemplar distance between each known inhibitor bound pocket and each pocket observed in a given ensemble ( starting from an unbound or peptide-bound structure ) . We then compared the most similar exemplar distances for each protein/inhibitor pair to experimentally-derived binding data , to evaluate whether the pockets sampled in these ensembles dictate the Bcl-2 family members that will be inhibited by a given compound . The results from this analysis are presented in Fig . 7A . Using pockets observed in inhibitor-bound crystal structures of Bcl-xL ( 1–14 ) , we find in many cases that highly similar pockets are sampled in ensembles generated from simulation of Bcl-xL and Bcl-2 ( green ) , but not in the corresponding ensembles from Mcl-1 or Bcl-w ( yellow/red ) : this represents a quantitative recapitulation of our observation that these ligands occupy surface pockets accessible only to Bcl-xL and Bcl-2 ( Fig . 5A ) . In light of this finding such compounds would be expected to bind Bcl-xL and Bcl-2 , but not Mcl-1 or Bcl-w; available experimental binding data confirm that indeed this is generally the case . We observe a similar pattern for most inhibitor-bound crystal structures of Bcl-2 ( 15–21 ) , which is again unsurprising given the similarity of these inhibitor-bound pockets to those of Bcl-xL ( Fig . 2C ) . The sole exceptions to this pattern are complexes 22 and 23 , for which corresponding pockets are observed in either the Bcl-2 or Bcl-xL ensembles but not both . The former ( 22 ) is indeed a dual inhibitor , but binds in a shallow pocket that is not well-described by any Bcl-2 exemplar in our ensemble . The latter ( 23 ) is indeed selective for Bcl-2 over Bcl-xL , as anticipated from the lack of overlap with the Bcl-xL ensemble from simulation , and therefore represents successful recapitulation of the binding data . While we noted earlier that Mcl-1 can adopt conformations with highly distinct surface pockets , we also noted that inhibitor-bound crystal structures do not make use of these pockets ( Fig . 6 ) . Rather , we find that these compounds ( 24–26 ) instead bind to pockets that are very similar to those included in the Bcl-2 ensemble ( Fig . 7A ) , and indeed experimental observations confirm that these compounds also inhibit Bcl-2 , but not Bcl-xL . Among the most notable incorrect predictions are those involving Bcl-w: a number of these compounds inhibit Bcl-w ( 11 , 12 , 13 , 16 , 17 ) , but corresponding pocket shapes were not included in our sampling . In retrospect , this may have arisen due to structural features of the starting conformation from which this conformational ensemble was generated: when evaluated by Molprobity [65] , this member of the NMR ensemble contained only 72% of residues in favorable regions of Ramachandran space . These unfavorable structural features may in turn have led to a lack of convergence of our simulations . Overall , however , there is a striking relationship between the pockets visited by a given ensemble and the experimentally-derived ligand selectivity . To quantitatively examine the ability of the pockets sampled in these ensembles to recapitulate the selectivity profile of a given ligand , we asked how well this approach could be used to distinguish the most tightly binding protein-ligand pairs ( those marked with ✔✔✔ ) from those pairs that bind too weakly to be detected/quantified ( those marked with ✗ ) . For each inhibitor-bound crystal structure , we normalized the exemplar distances to the variation across the corresponding row; this essentially expressed ligand selectivity as a Z-score indicating how closely a given ensemble approached the inhibitor-bound pocket . Using these Z-scores to rank the likelihood of interaction for a given protein-ligand pair , we used a receiver operating characteristic ( ROC ) plot to show performance at this binary prediction problem ( Fig . 7B ) ; we find that the predictions from this method far outperform those of a random classifier ( p < 4x10-7 ) . The absence of experimental binding data for many protein-ligand pairs is a natural shortcoming associated with culling this information from available reports in the literature . The complete maps of ligand selectivity for each compound , as inferred from ensembles generated by simulation , thus stand as completely new predictions in many cases ( S8 Fig . ) . Because the pockets should match the shape of the ligand , a reasonable assumption would be that similar results would be found by comparing the shape of the inhibitor to the ensemble of pocket shapes . To explore this idea we measured the exemplar distance between each native inhibitor conformer and the most similar pocket observed in a given ensemble of Bcl-2 family members , and then compared these exemplar distances to experimentally-derived binding data . As expected , the predictions ( S9 Fig . ) are very similar when comparing to the ligand directly , as opposed to comparing against the ligand-bound pocket . The abundance of reported inhibitors of Bcl-2 family members , including their selectivity across the Bcl-2 family , enabled the detailed comparison presented above . Due to the challenges encountered to date in identifying small-molecule inhibitors of protein interactions , however , there do not yet exist any further examples that we know of in which multiple compounds target different members of a protein family . Nonetheless , we were able to identify a separate example of a small-molecule inhibitor of a protein interaction that has been shown to inhibit select members of a protein family ( for which at least one ligand-bound crystal structure is available ) : this is ( + ) -JQ1 , a compound shown to selectively inhibit a subset of human bromodomains [10] . While only one compound , binding data are available against many family members . Here , we have generated ensembles of pocket-containing conformations for 16 bromodomains , and measured the extent to which each family member samples a pocket similar to that observed in the crystal structure of ( + ) -JQ1 bound to the first bromodomain of BRD4 . In this single additional example , here again we find that the presence of complementary surface pockets in the ensemble generated by simulation can accurately predict the ligand selectivity across a protein family ( Fig . 8 ) , whereas we do not observe complementary pockets for family members that do not tightly bind ( + ) -JQ1 .
Identification of small-molecule inhibitors of protein interactions immediately raised the question of how these compounds might interact with proteins that appeared to lack complementary surface pockets; the answer came through structural studies of an interleukin-2 complex that showed the ligand can occupy a hydrophobic groove not present on the unbound protein surface [13] . These structural studies , together with analysis of binding thermodynamics , first pointed to the “adaptivity” of this protein surface: the protein can adopt multiple conformations , one of which presents a surface complementary to the ligand . Our analysis of Bcl-2 family complexes supports this view , and extends it further: we find that the plasticity of the protein surface allows multiple distinct surface pockets to be presented , and these different pockets can be recognized by inhibitors with dramatically different chemotypes . Upon generating ensembles of pocket-containing conformations by simulation , we find that these ensembles span all the pockets used by known inhibitors—even though no information about any inhibitor was used to influence the simulations in any way . This observation provides strong and direct evidence for an underlying model of conformational selection [15]: the protein surface is predisposed to adopt certain pocket shapes , and these shapes in turn restrict the range of complementary ligands . Upon binding , the protein may then undergo further smaller changes in response to particular features of the ligand . By comparing the regions of “pocket space” explored by several Bcl-2 family members , we find that certain shapes are available to all family members: we expect that compounds complementing such pockets will show very broad specificity across this family . We further find that most inhibitors reported in the literature bind to pockets that are shared by more than one family member , and accordingly most are found to be active against more than one family member . Conversely , we also find that many Bcl-2 family members sample pockets that are not accessible to any other family member: here lies a tremendous opportunity , since we expect that a compound built to complement such a pocket will prove highly selective for its target . Development of compounds that target these highly “distinct” pockets represents a tantalizing new strategy for drug discovery: by building target selectivity into the broad features of the chemical scaffold itself , selectivity may be more robustly preserved in the course of optimization of the compound for other orthogonal desirable properties ( bioavailability , pharmacokinetics and pharmacology ) . Despite the existence of these pockets on the surface of Bcl-2 family members and extensive interest in identifying selective inhibitors , however , not a single crystal structure reported to date includes a compound that targets any of these highly “distinct” pockets . How then can we identify compounds that achieve target selectivity by explicitly targeting these “distinct” pockets ? We anticipate the solution may lie with the exemplars themselves . As noted earlier , the exemplar is essentially a map of the “perfect” ligand to complement a given pocket , albeit a ligand that is not physically realizable . Accordingly , we expect that the exemplar will serve as an ideal template for ligand-based screening of ( virtual ) compound libraries; tools such as ROCS [39 , 40] that evaluate volume and chemical overlap may be used to find compounds that closely mimic the shape and chemical features of the exemplar . Indeed , the ability to assess the selectivity profile by comparing the native ligand conformer to the exemplars derived from the ensembles implies that ROCS is capable of identifying compounds with the desired shape and chemical features needed to strongly interact with a member of a protein’s ensemble . Together , this set of tools may provide both a means to identify pockets that “encode” a desired selectivity profile within a protein family , and also a means to connect the resulting pockets to specific compounds that exhibit this selectivity profile . The overall paucity of examples of small-molecule inhibitors of protein interactions necessitated our focus for this study be largely restricted to the Bcl-2 family . As selective inhibitors of other protein families involved in protein interactions are reported , it will be exciting to refine the insights presented here . In light of the fact that many of the other small-molecule inhibitors of protein interactions described to date also bind to similarly “adaptable” binding sites [66] , meanwhile , we are optimistic that the perspectives presented here will prove extensible to these therapeutic targets as well . | Despite intense interest and considerable effort , there are few examples of small molecules that directly inhibit protein-protein interactions . Crystal structures of early successes have highlighted the plasticity of the protein surface , as some inhibitor-bound proteins are captured in conformations that are distinct from both their unbound and their protein-bound conformations . The lack of a single well-defined structure presents a challenge for predicting the members of a protein family to which a given compound will show activity ( ligand selectivity ) . Here we generate ensembles of conformations from simulation , and show that we can predict ligand selectivity based on which family members sample conformations complementary to the ligand . This approach may present a new avenue for designing highly-selective inhibitors of protein-protein interactions . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [] | 2015 | Selectivity by Small-Molecule Inhibitors of Protein Interactions Can Be Driven by Protein Surface Fluctuations |
Congenital cytomegalovirus ( CMV ) infection is the most common non-hereditary cause of sensorineural hearing loss ( SNHL ) yet the mechanisms of hearing loss remain obscure . Natural Killer ( NK ) cells play a critical role in regulating murine CMV infection via NK cell recognition of the Ly49H cell surface receptor of the viral-encoded m157 ligand expressed at the infected cell surface . This Ly49H NK receptor/m157 ligand interaction has been found to mediate host resistance to CMV in the spleen , and lung , but is much less effective in the liver , so it is not known if this interaction is important in the context of SNHL . Using a murine model for CMV-induced labyrinthitis , we have demonstrated that the Ly49H/m157 interaction mediates host resistance in the temporal bone . BALB/c mice , which lack functional Ly49H , inoculated with mCMV at post-natal day 3 developed profound hearing loss and significant outer hair cell loss by 28 days of life . In contrast , C57BL/6 mice , competent for the Ly49H/m157 interaction , had minimal hearing loss and attenuated outer hair cell loss with the same mCMV dose . Administration of Ly49H blocking antibody or inoculation with a mCMV viral strain deleted for the m157 gene rendered the previously resistant C57BL/6 mouse strain susceptible to hearing loss to a similar extent as the BALB/c mouse strain indicating a direct role of the Ly49H/m157 interaction in mCMV-dependent hearing loss . Additionally , NK cell recruitment to sites of infection was evident in the temporal bone of inoculated susceptible mouse strains . These results demonstrate participation of NK cells in protection from CMV-induced labyrinthitis and SNHL in mice .
Cytomegalovirus ( CMV ) is the most common infectious cause of congenital sensorineural hearing loss ( SNHL ) in humans [1] with between 15–30% of pediatric hearing loss attributable to this infection [2–4] . The consequences of hearing loss for affected children include speech and language delay , poor education attainment , and poor occupational performance in adulthood [5] . The total cost for each child with hearing loss is estimated to be over three hundred thousand dollars accounting for the lost productivity , the need for special education , vocational rehabilitation , assistive devices and medical costs [6] . One study estimates the total costs to the United States associated with congenital CMV infection to be $4 billion a year [7] . Despite the known significant health burden caused by congenital CMV induced hearing loss , very little is known about its pathogenesis including considerable uncertainty regarding the roles of direct viral replication in the cochlea and the contribution of host immune responses . An animal model that accurately recapitulates human CMV-induced hearing loss has been developed to evaluate more effective strategies for prevention and treatment [8 , 9] . Our group and others have successfully demonstrated that murine CMV ( mCMV ) -induced labyrinthitis in BALB/c murine newborn pups occurs when green fluorescent protein ( GFP ) expressing mCMV was used to inoculate newborn mice via an intracerebral ( IC ) injection [10 , 11] . These studies recapitulate viral mediated hearing loss in human infant because a critical factor for effective correlation between the mouse model and the clinical condition is that the mouse auditory system at birth is analogous to the human fetal auditory system and does not achieve stable thresholds until 4 weeks of age [12] . When infected at birth , fifty-five percent had profound hearing loss ( ≥ 80 dB ) at 4 weeks of age , while the other forty-five percent initially showed moderate hearing loss that progressed to profound hearing loss by 6–8 weeks . These findings mirror the longitudinal human clinical studies that show that 50% of children with hearing loss have worsening thresholds over time [13 , 14] . Moreover , asymmetric hearing loss occurred in 40% of the mice , similar to the rate of 50% among children with congenital CMV infection observed by Fowler and colleagues [15] . In addition , we also showed that this susceptibility to CMV-induced hearing loss was age dependent . While all of the postnatal day three ( P3 ) infected mice showed elevated auditory brainstem response ( ABR ) thresholds at 4 weeks of age , only fifteen percent of the P8 , and none of the P14 infected mice had hearing loss . Pass et al . demonstrated an age dependent effect in children with congenital CMV infection who were more likely to develop hearing loss when infected during the first trimester rather than later in pregnancy [16] . The mouse model is relevant to evaluate the viral host interactions that involve the innate and adaptive immune response [17 , 18] . Previous studies have established that the response to mCMV infection is strain-dependent in mice , with susceptibility determined by expression of the Ly49H gene ( Klra8 killer cell lectin-like receptor , subfamily A , member 8 ) in the distal portion of the natural killer cell gene complex [19–21] . During the early phase of infection , the Ly49H receptor recognizes m157 , a viral-encoded class I major histocompatibility complex ( MHC ) homologue expressed at the cell surface of mCMV infected cells . Despite the adaptation of viral expression of the MHC as an immune evasion tactic [22] , the interaction of m157 with the Ly49H receptor triggers NK cell activation and elimination of the infected cells in mice [23 , 24] . Although Ly49H recognition of m157 is critical to controlling viral titers in the spleen and lung , it is much less effective in controlling virus replication in the liver . Here , we evaluated the role of NK cells , specifically Ly49H recognition of m157 , in mCMV-induced labyrinthitis and subsequent hearing loss .
To analyze the physiological consequences of mCMV infection on hearing , we measured the minimum electrophysiological input required to evoke a threshold response of our experimental animals using auditory brainstem response ( ABR ) , and distortion product otoacoustic emission ( DPOAE ) . ABR samples evoked potentials in the auditory nerve and brainstem , thus measures the physiological response of the entire auditory pathway . DPOAE measures sound produced by the structures of the inner ear , in particular , the amplification function of the outer hair cells . In both cases , increased thresholds required to elicit measurable responses indicate hearing deficiency . ABR and DPOAE thresholds showed a marked difference between the two mouse strains . BALB/c pups inoculated at P3 with mCMV-GFP showed profound hearing loss by 4 weeks of age ( Fig 1 and S1 Table ) . In contrast , infected C57BL/6 showed a slight , but significant , threshold elevation for DPOAE measurements compared to uninfected control mice ( P < 0 . 005 , Fig 1A and S1 Table ) and no evidence of hearing loss based on ABR measurements ( P = 0 . 664 , Fig 1B and S1 Table ) . Comparison between infected C57BL/6 and BALB/c groups yielded a significant difference between the two strains over all DPOAE thresholds ( P < 0 . 0001 , S1 Table ) and ABR thresholds ( P < 0 . 0001 , S1 Table ) . These data indicate that C57BL/6 mice were resistant to mCMV-induced hearing loss compared to BALB/c mice . The outer hair cells ( OHC ) in the Organ of Corti function to enhance cochlear sensitivity and frequency selectivity and are responsible for amplification of sound vibrations measured by DPOAE . Since mCMV infection resulted in differential hearing loss , we evaluated outer hair cell loss in the two mouse strains . Outer hair cell loss was evident by scanning electron microscopy ( SEM ) in both BALB/c and C57BL/6 mice due to mCMV-GFP infection compared to uninfected controls ( Fig 2A and 2B ) . However , total OHC loss was more than two-fold higher in infected BALB/c mice compared to C57BL/6 mice ( Fig 2C ) . Compared to uninfected controls , OHC loss was evident in all cochlear turns , but BALB/c infected mice had more severe OHC loss in the basal cochlear turn than C57BL/6 infected mice ( Fig 2C ) , perhaps reflecting the greater DPOAE and ABR thresholds seen at higher frequencies . Cochleograms reflected the differential OHC loss seen by SEM in the two strains ( S1 Fig ) and indicated greater loss of outer hair cells for BALB/c infected mice compared to C57BL/6 infected mice ( S1B Fig ) . Although differences at individual time points did not reach the level of significance between the mouse strains , hair cell loss was progressive in that there was a significant overall time-dependent outer hair cell loss in BALB/c mice ( P = 0 . 0053 by ANOVA ) . A similar time-dependent trend was noted for C57BL/6 mice that did not reach the level of significance ( P = 0 . 109 ) , but consistent with the resistance to mCMV hearing loss in C57BL/6 mice . These data indicate that strain-dependent hearing loss can be at least partially explained by OHC loss . Consistent with other forms of ototoxicity , mCMV infection did not markedly alter the appearance of inner hair cells when examined by SEM or immunohistochemistry and compared to uninfected controls ( see images provided at https://doi . org/10 . 7278/S5V69GR6 ) . To assess the direct participation of the NK cell Ly49H receptor/m157 ligand interaction in protecting C57BL/6 mice from mCMV-induced hearing loss , we first tested the effect of a Ly49H blocking antibody on DPOAE and ABR . Increased thresholds were seen at 4 weeks post-injection for all frequencies in mice receiving both the Ly49H blocking antibody and 200 pfu mCMV compared to mice receiving the IgG isotype control antibody and 200 pfu mCMV for DPOAE and ABR ( P < 0 . 0001 ) ( Fig 3A and S1 Table ) although increases were most pronounced for ABR at the 32 kHz tone frequency ( Fig 3B and S1 Table ) . This hearing loss worsened over time in that mCMV infected mice showed increases over all thresholds at 6 weeks compared to 4 weeks after inoculation ( P < 0 . 0001 ) . These data show that blocking the Ly49H receptor in otherwise resistant C57BL/6 mice resulted in at least mild hearing loss by 4 weeks of age that progressed to moderate-to-profound loss by 6 weeks after mCMV inoculation . We next tested the effect of the mCMV-encoded m157 ligand by inoculating C57BL/6 pups with a mCMV virus deleted for the m157 gene ( mCMV Δm157 , described in [25] ) . C57BL/6 mice infected with mCMV-Δm157 showed significant increases in both DPOAE ( Fig 3C and S1 Table ) and ABR ( Fig 3D and S1 Table ) thresholds over all tested frequencies relative to mice infected with wild-type mCMV virus . Taken together , these data indicate that a competent Ly49H receptor/m157 ligand interaction protects mice from mCMV-induced hearing loss . In the absence of Ly49H/m157 disruption , mCMV infected cells in the C57BL/6 mouse cochlea were rare within the first week after infection , whereas mCMV infected cells were plentiful in cochlea from BALB/c mice up to one week post-infection [11] . Viral-encoded GFP was routinely detected in the spiral ganglion , and occasionally in perilymphatic epithelia by fluorescent and immunofluorescent microscopy ( Fig 4 and S2 Fig ) in C57BL/6 mice treated with Ly49H blocking antibody . Viral DNA was detected in temporal bones from infected mice ( S3 Fig ) . Furthermore , activated caspase-3 was localized in the vicinity of GFP signals ( S4 Fig ) indicating mCMV infection resulted in activation of the apoptotic cascade . Caspase activation was only seen in coordination with GFP signal , which were largely confined to the spiral ganglion with rare individually infected cells seen in the scala tympani . Coordinated cleaved caspase signal dramatically increased within the spiral ganglion in mCMV infected C57BL/6 mice after Ly49H/m157 blockade . These data demonstrate active mCMV infection of cells within the mouse cochlea . However , consistent with our previous data in BALB/c mice [11] , GFP signals , indicating active mCMV infection , were not seen in the hair cells of C57BL/6 mice at any of the examined time points up to 28 days post-infection suggesting that hair cells are not the direct target of mCMV . The effect of mCMV infection on NK cell localization was examined in a C57BL/6 mouse strain that constitutively expressed td-Tomato , a red fluorescent protein , in NK cells ( NK1 . 1-tdTomato knock-in mice ) . NK cells , as indicated by RFP fluorescence ( N = 6–12 mice per group ) or RFP immunofluorescence ( N = 3–4 mice per group ) , were rarely seen in in uninfected C57BL/6 mice and only occasionally seen in mCMV infected C57BL/6 mice ( Fig 4B and 4D ) . In these cases , RFP signal appeared to be randomly localized . In contrast , pretreatment of NK1 . 1-tdTomato knock-in mice with Ly49H blocking antibody prior to mCMV-GFP inoculation showed a dramatic increase in both mCMV-GFP infected cells and associated NK cells three days after infection ( Fig 4C and 4E ) . These data suggest that , in the presence of a competent Ly49H/m157 interaction , NK cells can effectively attenuate mCMV infection in the cochlea .
Natural killer cells are a fundamental component of the immune response to virally infected cells and are among the first immune cells to respond to pathogen challenge . NK cells function in the early innate response via cytolytic activity and affect the adaptive immune response through release of cytokines . In mice , recognition of virally infected cells is largely coordinated by the Ly49 C-type lectin family of homodimeric receptors , which have both activating and inhibitory isoforms . Numerous studies in mice have established that NK cell expression of the Ly49H receptor confers resistance to mCMV infection through early recognition of the virally-encoded m157 cell surface antigen [19 , 21 , 23 , 24 , 26 , 27] . NK cell mediated protection from mCMV infection has been demonstrated by decreased viral load in murine spleen and lung [19 , 25 , 26 , 28] and decreased cell lysis and tissue disintegration in spleen [26] . However , Ly49H/m157 interactions are less effective in other tissues such as the liver , so it was unclear if these interactions are relevant for mCMV-induced hearing loss in neonatal mice as we now demonstrate . DPOAE and ABR testing consistently showed elevated hearing thresholds and cochleograms demonstrated greater outer hair cell loss in the susceptible BALB/c mouse strain as compared to the near normal thresholds and minimal outer hair cell loss in the resistant C57BL/6 mouse strain . The fact that NK cell depletion by IP injection of neonatal mice ( P2 ) with anti-asialo GM1 antibody increases mCMV titer and reduced cytokine production in brain indicates that the NK cell response participates in mCMV infection in newborn mice after IC injection [29] . Our data demonstrate that mCMV susceptibility to hearing loss is mediated , at least in part , by NK cells . We recognize that Ly49H receptor expression on NK cells has not been explicitly established and it remains possible that the blocking antibody interacts with an alternate target that results in hearing loss , however , that both addition of the Ly49H blocking antibody and infection with an m157 deficient viral strain rendered C57BL/6 mouse susceptible to mCMV hearing loss and cochlear damage suggests the Ly49H receptor/m157 interaction participates in mCMV-induced hearing loss . Additionally , NK cell co-localization with GFP expressing mCMV infected cells dramatically increased in the cochlea of normally resistant C57BL/6 mouse strain after blockade of the Ly49H receptor consistent with the requirement for physical interaction between NK cells and mCMV-infected cells for effective Ly49H engagement of m157 . Although a competent NK cell Ly49H receptor interaction with virally encoded m157 ligand in infected cells was required for otoprotection against viral infection in the cochlea , hearing was not completely preserved as DPOAE thresholds increased modestly , but significantly , in mCMV infected resistant C57BL/6 mice compared to uninfected controls . The increased DPOAE thresholds in resistant C57BL/6 mice were consistent with outer hair cell loss , which eventually reached about 50% of the loss seen in susceptible BALB/c mice , although overall OHC loss is minor and likely does not explain the full extent of hearing loss . These data indicate that a competent NK cell response is not sufficient to protect inner ear structures from damage after mCMV infection and that either NK cell clearance of infected cells was incomplete or that sequelae of infection contributed to subsequent SNHL . Furthermore , since direct evidence of mCMV infection in the hair or supporting cells within the organ of Corti was not seen suggests that secondary effects of mCMV infection are responsible for hair cell loss . Our observation of outer hair cell loss without evidence of direct cochlear hair cell infection is consistent with previous studies of mCMV infected mice [10 , 30 , 31] . Similarly , mCMV infection favored cells of the spiral ganglion and perilymphatic epithelial cells , which is largely consistent with previous results [10 , 30–32] . Apoptosis of spiral ganglion neurons has been identified as a component of hearing loss in the susceptible BALB/c mouse strain [32] . Our data demonstrates similar activation of the apoptotic cascade in the previously resistant C57BL/6 mouse strain after interruption of NK cell recognition signals suggesting that early clearance of mCMV infection by NK cells and protection from spiral ganglion apoptosis contributes to protection from hearing loss . The fact that spiral ganglion cells appear to be the major site of mCMV infection at 3-days post-infection after Ly49H receptor blockade indicates that protection of spiral ganglion cells was the main contributor to NK hearing loss protection . It is known that individuals with defects affecting NK cell function are particularly susceptible to human CMV disease [33 , 34] . Although the killer cell lectin-like receptor , subfamily A genes , of which Ly49H is a member , appear to be lacking in humans [35] , human NK cells express a range of inhibitory and activating surface receptors , including lectin-like receptors and Ig-like receptors that could be explored in the context of CMV-induced hearing loss [36] . For example , a human CMV-encoded immunoevasin , UL18 , has been shown to activate NK cells lacking leukocyte Ig-like receptor 1 [37] . However , it is unlikely that modulation of the NK cell response in the clinical setting will be a viable intervention target given the paucity of information about NK cell developmental status and receptor complement in utero or in newborns . Nevertheless , our results further delineate mechanisms of CMV-induced hearing loss in the mouse and provide additional evidence of the correlation to the clinical presentation of congenital CMV sensorineural hearing loss .
All animal studies were approved by the University of Utah Institutional Animal Care and Use Committee ( protocol number 14–07006 ) , performed in compliance with relevant institutional policies , local , state , and federal laws , and conducted following National Research Council Guide for the Care and Use of Laboratory Animals , Eighth Edition . Animals were anesthetized with hypothermia or ketamine/xylazine and euthanized by exsanguination after a surgical plane of anesthesia was reached . Recombinant mCMV ( strain K181 MC . 55 ( ie2- GFP+ ) ) expressing green fluorescent protein ( GFP ) was supplied by Dr . Mark R Schleiss ( Minneapolis , MN , USA ) . A mCMV mutant with a functional deletion of the m157 gene ( Δm157 ) and its parental wild-type strain ( WT1 ) were previously described [25 , 38] . To grow viral stock , M2-10B4 murine fibroblast cells ( cat# CRL-1972 , American Type Culture Collection , Manassas , VA , USA ) were cultured in complete medium ( minimal essential medium , 10% ( v/v ) fetal bovine serum ( FBS ) , 2mM l-glutamine , 100 U/ml penicillin , 0 . 1 mg/mL streptomycin , 10nM HEPES ) . Once 70% confluent , the viral inoculum was added and the cells were incubated to 80% to 100% cytopathic effect . Each plate was then subjected to three freeze–thaw cycles and the resulting viral supernatant was collected . Virus purification was carried out by a commercial contract laboratory ( Virapur , San Diego , CA , USA ) using the following protocol: cellular debris were removed by centrifugation ( 1 , 000×g ) at 4°C , and the virus were pelleted through a 35% sucrose cushion ( in Tris-buffered saline [50 mM Tris–HCl , 150 mM NaCl , pH 7 . 4] ) at 23 , 000×g for 2 h at 4°C . The pellet was resuspended in Tris-buffered saline containing 10% FBS . Viral stock titers were determined on M2-104B cells as 50% tissue culture infective doses ( TCID50 ) per milliliter . BALB/c , C57BL/6 mice ( Jackson Labs , Sacramento , CA , USA ) and a C57BL/6 mouse strain that constitutively expresses red fluorescent protein ( NK1 . 1-tdTomato knock-in mice ) in NK and NKT cells [39] were used for experiments as indicated . Animals were housed and bred under pathogen-free conditions at the Central Animal Facility at the University of Utah . Mice were injected via an intracerebral route at post-natal day 3 ( P3 ) of life as previously described [11] . Briefly , the pups were momentarily placed on ice to induce anesthesia . The mouse was manually restrained , and a 10 μl Hamilton syringe with a 30G needle was inserted past the calvarium in the mid parietal region to inject 200 plaque forming units ( pfu ) of virus in a volume of 1 μl . Control groups received 1 μl phosphate-buffered saline ( PBS ) carrier unless otherwise indicated . Mice were monitored for adverse effects , including mortality , behavioral abnormalities , and developmental delay . Experimental and control animals were housed separately . For groups pretreated with Ly49H blocking antibody , 20 μg of purified mouse anti-Ly49H monoclonal antibody ( clone 3D10 , cat# 14-5886-82 , eBiosciences , San Diego , CA , USA ) in 50 μl of PBS was injected into the peritoneal cavity twelve hours before the inoculation of mCMV . Control mice received 20 μg of purified mouse IgG isotype control antibody ( cat# 02–6502 , ThermoFisher , Rockford , IL , USA ) in 50 μl of PBS . For auditory brainstem response ( ABR ) and distortion product otoacoustic emission ( DPOAE ) testing , mice were anesthetized with a combination of ketamine and xylazine at 100 and 10 mg/kg body weight , respectively . ABRs/DPOAEs were performed in a double-walled sound chamber ( IAC Acoustics , North Aurora , IL , USA ) . The body temperature was maintained at ~37°C via a heating pad . A small incision was made at the tragus to allow better access to the ear canal . For ABR testing , an electrostatic speaker ( EC-1 , Tucker-Davis Technology , Alachua , FL , USA ) fitted with a 1 . 5 cm long polyethylene tube was placed abutting the ear canal . Needle electrodes were placed subcutaneously at the mastoid of the tested side and vertex , with a remote ground electrode placed in the rump area . ABR thresholds were measured bilaterally in all mice . ABR signals were amplified with a TDT RA4 pre-amplifier ( Tucker-Davis Technology ) , filtered from 100 to 3000 Hz , averaged and digitized with a TDT RA16BA processor controlled by BioSigRP software ( Tucker-Davis Technology ) . Acoustic stimuli were digitally generated and processed by a RX6 real-time processor and passed through a PA5 attenuator prior to delivery to the speaker amplifier at a rate 24–32 times/sec . Responses to 1 , 000 sweeps were averaged for a series of responses to tone pips ranging from 8 to 32 kHz ( 5 ms with 0 . 5 ms cos2 rise and fall ) using 5 or 10 dB intensity steps , over a 15–90 dB of sound pressure level ( dB SPL ) range . ABR traces were visually inspected after plotting the amplitude of each peak against stimulus intensity . Thresholds typically corresponded to a level one step below that at which the peak-to-peak response amplitude began to rise . An ABR threshold of 90 dB SPL ( i . e . , the highest stimuli presented in this study ) was assigned to cochleae that failed to stimulate an ABR waveform at 90 dB SPL . The DPOAEs were measured using an ER-10B+ ( Etymotic Research , Elk Grove , IL , USA ) microphone coupled with two EC1 speakers . Stimuli of two primary tones f1 and f2 ( f2/f1 = 1 . 2 ) were presented with f2 = f1–10 dB . Primary tones were stepped from 30 to 80 dB SPL ( for f1 ) in 10 dB increments and swept from 8 to 32 kHz in octave steps . Stimuli were generated and attenuated digitally ( 200 kHz sampling ) . The ear canal sound pressure was pre-amplified and digitized . A fast Fourier transformation was computed , and the sound pressures at f1 , f2 , and 2f1- f2 were extracted after spectral averaging from 50 serial waveform traces ( each corresponding to 84 ms of digitized ear canal sound pressure waveform ) . The noise floor ( average of 10 points in the FFT on either side of 2f1-f2 ) was also measured: it ranged between -25 and 0 dB SPL , depending on the test frequencies . All data were shown in mean ± SEM . The mice hearing reaches mature thresholds by 3 weeks of age [12] . To avoid potential confounding results due to immaturity of the auditory system in mCMV infected mice , we chose to test our animals beginning 4 weeks of age . Mice were anesthetized with ketamine/xylazine and exsanguinated by transcardial ( left ventricle to right atrium ) perfusion with 0 . 1 M phosphate-buffered saline ( PBS ) , pH 7 . 4 containing 100 U/ml heparin followed by 20 ml of 2% paraformaldehyde in phosphate buffer ( PB ) at room temperature ( RT ) . The excised cochleae were immersed in fixative ( 2% paraformaldehyde in 0 . 1 M PB ) overnight at 4°C , washed with PBS , and decalcified by immersion in 0 . 12 M EDTA , pH 7 . 0 for 1–5 days at 4°C . The decalcified cochleae were infiltrated with sucrose and then embedded in 7 . 5% gelatin/15% sucrose/1 X PBS . Serial 10 μm thick mid-modiolar sections were cut on a freezing microtome and mounted on poly-L-lysine-coated glass slides . Slides were stored at −20°C until further use . The slides were dried at RT for 10 min , washed in PBS , permeabilized with 0 . 2% Triton X-100 in PBS for 1 hour at RT , washed in PBS , and transferred to blocking buffer ( BlockAid Blocking Solution , ThermoFisher , Waltham , MA , USA ) for 1 hour at RT , prior to application of primary antibodies . Primary antibodies to GFP ( goat polyclonal anti-GFP , cat# AF4240 , R&D Systems , Minneapolis , MN , USA ) and tdTomato ( rabbit polyclonal anti- RFP , cat#600-401-379 , Rockland Antibodies , Limerick , PA , USA ) were also used to validate the fluorescent labels . Rabbit anti- Myosin VIIa ( polyclonal , cat# 25–6790 , Proteus BioSciences , Ramona , CA , USA ) was used to visualize hair cells in whole-mount cochleograms . Rabbit anti- active caspase-3 ( clone C92-605 , cat# 559565 , BD Pharmingen , San Jose , CA , USA ) was used to monitor apoptosis . All antibodies were incubated in blocking buffer . Following overnight incubation of the sections in primary antibodies at 4°C , the sections were rinsed in PBS and secondary antibodies applied for 1 hour at RT . Secondary antibodies used were donkey anti-goat IgG , Alexa Fluor 488 conjugate ( cat# A11055 , ThermoFisher ) for the GFP primary , chicken anti-rabbit IgG , Alexa Fluor 594 conjugate ( cat# A21442 , ThermoFisher ) for the RFP and Myosin VIIa primary antibodies , and donkey anti-rabbit IgG , Alexa Fluor 568 conjugate ( cat# A10042 , ThermoFisher ) for the caspase-3 primary . Sections were counterstained with 4’ , 6-diamidino-2-phenylindole ( DAPI ) in PBS for 5 min at RT in the dark . Scanning electron microscopy ( SEM ) was performed as described previously [40] . Briefly , mice were anaesthetized and then exsanguinated using 2 . 5% glutaraldehyde in PBS via transcardial perfusion . Temporal bones were harvested and fixed in 2 . 5% glutaraldehyde/PBS overnight at 4°C . Cochleas were dissected and coated using 1% osmium tetroxide and thiocarbohydrazide , followed by critical point drying using hexamethyldisilazane . Samples were then imaged using Hitachi-4800 scanning electron microscope . Mice and temporal bones were treated and processed as described for histochemistry in the main text , through the decalcification step . Mouse cochlear whole mounts were prepared as described in ( http://www . bio-protocol . org/e332 ) . Hair cells were visualized using a primary antibody to myosin VIIA ( rabbit anti-myosin VIIA , Proteus Biosciences cat# 25–6790 , Ramona , California ) and chicken anti-rabbit Alexa Fluor 594 ( Life Technologies cat# A-21442 ) secondary antibody . A standard cochleogram was prepared for each ear using a 20X objective on a Nikon A1R confocal microscope . In each section , the number of present and absent hair cells was assessed throughout the entire section thickness and plotted as fractional loss . Viral DNA was measured by quantitative PCR ( qPCR ) using the viral Immediate Early response gene 1 ( IE1 ) as the target relative to β-actin expression . DNA was extracted from the crushed temporal bones using QIAmp MinElute Virus Spin Kit ( #57704 , Qiagen , Valencia , CA ) . Each sample was assayed in duplicate using Taqman Gene Expression Master Mix ( # 4370048 , Life Technologies , Carlsbad , CA ) and the Applied Biosystems QuantStudio 12K Flex Real Time PCR System ( Life Technologies ) . Amplification conditions were initial denaturation for 95°C for 10 minutes , followed by 45 cycles of denaturation at 95°C for 15 seconds and anneal/extension at 60°C for 1 minute . Primer sequences were: IE1 primer 1: CCC TCT CCT AAC TCT CCC TTT , IE1 primer 2: TGG TGC TCT TTT CCC GTG , ActinB primer 1: AGC TCA TTG TAG AAG GTG TGG , Actin B primer 2: GGTGGG AAT GGG TCA GAAG ( Integrated DNA Technologies , Prime Time Standard qPCR Assay 6FAM/Zen/ABFQ ) . Delta Ct values were determined for IE1 DNA levels normalized relative to β-actin levels and comparisons between groups were carried out using nonparametric rank sums tests . A standard curve was generated using DNA extracted from known quantities of mCMV . Efficiency of the PCR reaction was 99 . 66% as determined from the slope of the CT vs . log ( mCMV DNA ) standard curve . Statistical analysis was carried out in IBM SPSS Statistics for Windows ( V . 21 , IBM , Armonk , NY , USA ) and GraphPad Prism for Windows ( V . 6 , GraphPad Software , La Jolla , CA , USA ) . Data are expressed as mean ± SEM . Differences in ABR and DPOAE thresholds were analyzed by means of the nonparametric Kruskal-Wallis test . OHCs loss from cochleas imaged using SEM was analyzed by means of Mann–Whitney U test . Time-dependent cochleogram data were analyzed by means of 2-way ANOVA . The a priori significance level was set at P < 0 . 05 . | Cytomegalovirus ( CMV ) transmission from an infected mother to her fetus is a leading cause of permanent hearing loss in children , but the contributing processes are not clear . In this report , we utilized a mouse model , which recapitulates many features of congenital CMV mediated childhood hearing loss , to demonstrate that natural killer cells ( NK ) , a component of early host immune response to infection , play a critical protective role in CMV-induced hearing loss . Specifically , we determined that NK cells interact with CMV infected cells through binding of the NK cell receptor , Ly49H , with a virally-encoded protein , m157 , expressed on the cell surface of CMV infected inner ear cells , to mediate the protective effect . Findings from this study provide insight into the host immune response during CMV-induced hearing loss in mice . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] | [
"blood",
"cells",
"medicine",
"and",
"health",
"sciences",
"immune",
"cells",
"immune",
"physiology",
"ears",
"immunology",
"cytomegalovirus",
"infection",
"neuroscience",
"animal",
"models",
"outer",
"hair",
"cells",
"otology",
"inner",
"ear",
"model",
"organisms",
"microscopy",
"experimental",
"organism",
"systems",
"hearing",
"disorders",
"antibodies",
"research",
"and",
"analysis",
"methods",
"immune",
"system",
"proteins",
"infectious",
"diseases",
"white",
"blood",
"cells",
"animal",
"cells",
"proteins",
"scanning",
"electron",
"microscopy",
"mouse",
"models",
"head",
"otorhinolaryngology",
"biochemistry",
"cellular",
"neuroscience",
"deafness",
"cell",
"biology",
"anatomy",
"nk",
"cells",
"cochlea",
"physiology",
"neurons",
"electron",
"microscopy",
"biology",
"and",
"life",
"sciences",
"cellular",
"types",
"afferent",
"neurons",
"viral",
"diseases"
] | 2017 | Natural killer cells attenuate cytomegalovirus-induced hearing loss in mice |
Schistosome parasites cause schistosomiasis , one of the most prevalent parasitemias worldwide affecting humans and animals . Constant pairing of schistosomes is essential for female sexual maturation and egg production , which causes pathogenesis . Female maturation involves signaling pathways controlling mitosis and differentiation within the gonads . In vitro studies had shown before that a Src-specific inhibitor , Herbimycin A ( Herb A ) , and a TGFβ receptor ( TβR ) inhibitor ( TRIKI ) have physiological effects such as suppressed mitoses and egg production in paired females . As one Herb A target , the gonad-specifically expressed Src kinase SmTK3 was identified . Here , we comparatively analyzed the transcriptome profiles of Herb A- and TRIKI-treated females identifying transcriptional targets of Src-kinase and TβRI pathways . After demonstrating that TRIKI inhibits the schistosome TGFβreceptor SmTβRI by kinase assays in Xenopus oocytes , couples were treated with Herb A , TRIKI , or both inhibitors simultaneously in vitro . RNA was isolated from females for microarray hybridizations and transcription analyses . The obtained data were evaluated by Gene Ontology ( GO ) and Ingenuity Pathway Analysis ( IPA ) , but also by manual classification and intersection analyses . Finally , extensive qPCR experiments were done to verify differential transcription of candidate genes under inhibitor influence but also to functionally reinforce specific physiological effects . A number of genes found to be differentially regulated are associated with mitosis and differentiation . Among these were calcium-associated genes and eggshell-forming genes . In situ hybridization confirmed transcription of genes coding for the calcium sensor hippocalcin , the calcium transporter ORAI-1 , and the calcium-binding protein calmodulin-4 in the reproductive system pointing to a role of calcium in parasite reproduction . Functional qPCR results confirmed an inhibitor-influenced , varying dependence of the transcriptional activities of Smp14 , Smp48 , fs800 , a predicted eggshell precursor protein and SmTYR1 . The results show that eggshell-formation is regulated by at least two pathways cooperatively operating in a balanced manner to control egg production .
Blood-dwelling endoparasites of the genus Schistosoma are the only trematodes that have evolved a gender dimorphism [1] , [2] . These parasites cause schistosomiasis , which is of worldwide significance for humans and animals in tropical and sub-tropical areas [3] . About 780 million people live in endemic areas being at risk of schistosomiasis , of which 200 million are infected generating annual losses of 1 . 7 to 4 . 5 million disability adjusted life years ( DALYs ) of humans as determined by the Global Burden of Disease Programme [4] , [5] . Living in the abdominal veins of their vertebrate hosts , adult paired females produce up to 300 eggs per day . Half of these eggs penetrates the epithelia and reach the gut lumen ( e . g . S . mansoni ) or the bladder ( S . haematobium ) to be transported to the environment for continuing the life cycle . The remaining eggs migrate via the blood stream to different organs such as spleen and liver , where they get trapped causing granuloma formation and liver cirrhosis [6] , [7] . A unique biological feature of schistosomes is the dependency of the sexual maturation of the female on a constant pairing contact with the male . Following pairing , mitoses and differentiation are induced in the female leading to the differentiation of the reproductive organs , ovary and vitellarium [8] , [9] . Regarding the importance of eggs for continuing the life cycle and provoking pathogenesis , a number of studies focused on the identification and characterization of genes controlling reproductive development of this parasite [8] , [10]–[14] . Furthermore , genome and transcriptome projects have unravelled the parasite's genetic repertoire [15]–[17] . Several members of the TGFβ signaling pathway were identified , which is a highly conserved pathway throughout the animal kingdom . Among these , the type I TGFβ receptor ( SmTβRI; Smp_049760; [18] ) , a type IIb activin receptor called SmRK2 ( SmActRIIb; Smp_144390; [19] ) , SmSmad4 ( Smp_033950; [20] ) , SmSmad2 ( Smp_085910; [21] ) , and SmFKBP12 ( Smp_079230; [22] ) were identified . Localisation studies revealed a preferential expression of the listed molecules within the reproductive organs . SmTβRI transcripts were localised within the vitellarium and ovary as well as the parenchyma of both genders [23] . Furthermore , expression of SmTβRI protein was also found at the surface of male parasites [18] . SmActRIIb was found to be expressed in males and females , and localised on the tegumental surface of the gynaecophoral canal and some parenchymatic cells [19] . SmSmad4 was detected within epithelia surrounding the gut and vitellarium , as well as the subtegument and muscles of males [20] and SmSmad2 within the vitellarium , the developing egg , and the ovary of the female worm , but also in the testes and tubercles of the male [24] . SmFKBP12 co-localised with SmTβRI in the female gonads as well as the parenchyma of the adult schistosomes [23] . In the animal kingdom , the TGFβ pathway controls proliferation and differentiation processes [25] . First studies to elucidate the functional meaning of the schistosome TGFβ pathway have included ligand-induction and inhibitor-suppression approaches [11] , [26] . Using human TGFβ ( hTGFβ ) to induce the TGFβ pathway in adults in vitro , a recent study showed that genes related to morphology , development , and cell cycle were differentially transcribed [27] . Earlier studies , based on the use of a specific TβRI kinase inhibitor ( TRIKI ) in adult schistosomes in vitro to suppress the TGFβ pathway , provided first evidence for its role in regulating mitotic activity and egg production in paired S . mansoni females [11] . Using a similar inhibitor approach with adults in vitro indicated the additional influence of ( a ) Src kinase-containing pathway ( s ) on these processes in paired S . mansoni females . Based on the discovery of the gonad-specific expression of the cellular Src tyrosine kinase SmTK3 ( Smp_151300; [28] ) , inhibition experiments with the Src-kinase inhibitor Herbimycin A ( Herb A ) led to reduced mitotic activity and egg production in paired females as well [29] . The comparison of both inhibitor treatments pointed to a stronger reduction of both parameters following Herb A treatment [11] . The strongest influence on the mitotic activity and egg production was observed by combining both inhibitors . In this study , we investigated the inhibitory impact of TRIKI , Herb A , or the combined compounds on the transcriptome of female schistosomes using a microarray approach and comprehensive qPCR analyses . Besides the identification of a large number of genes , which were differentially transcribed upon inhibitor treatment , the results provide strong molecular evidence for the participation of both TβRI and Src kinase-containing pathways controlling the transcription of genes involved in eggshell formation in a cooperative and balanced manner .
The predicted inhibition of SmTβRI by TRIKI ( also known as LY-364947 ) was confirmed by expressing the recombinant intracellular active kinase domain of SmTβRI in Xenopus laevis oocytes [30] , a suitable system for the expression and detection of kinase activity of schistosome proteins [31]–[34] . In X . laevis stage VI oocytes naturally blocked in prophase I of meiosis I , the kinase potential of an exogenous recombinant active kinase triggers resumption of meiosis and thus germinal vesicle breakdown ( GVBD ) , a process easily monitored by the appearance of a characteristic white spot at the animal pole of the oocyte [30] . To functionally analyze the kinase potential of SmTβRI , a constitutively active variant ( SmTβRI7D ) [35] and an inactive one ( SmTβRIVVAAAVV ) were generated by site-directed mutagenesis , and appropriate cRNAs were injected into Xenopus oocytes . Results shown in Figure 1 demonstrated that expression of the active SmTβRI7D version induced GVBD in ≥80% of oocytes whereas the inactive one SmTβRIVVAAAVV had no effect on the fate of the oocytes . In the presence of TRIKI , oocytes expressing SmTβRI7D underwent no more GVBD when drug concentrations ≥30 nM were used , confirming the inhibitory effect of TRIKI on SmTβRI kinase . For comparison , previous experiments showed that in this cellular system a complete inhibition of the Src kinase SmTK3 was obtained using 10 nM Herb A [34] . GVBD induced in control oocytes by the natural stimulus progesterone [30] was not affected by TRIKI-treatment ( data not shown ) , demonstrating a specific effect of TRIKI on the SmTβRI kinase in injected oocytes . Based on previous findings of reduced mitotic activity and egg production following inhibitor treatment of adult female schistosomes in vitro [11] , [29] , we devised an approach to unravel the molecular mechanisms affected by these inhibitors . To this end , large-scale transcriptional analyses were performed using a microarray platform representing nearly the complete S . mansoni transcriptome [27] , [36] , [37] . Our experimental design comprised adult schistosomes that were cultured in vitro for 48 h with either TRIKI ( 300 nM ) , Herb A ( 4 . 5 µM ) , the combination of both inhibitors ( H+T ) , or with DMSO only as control . A total number of 8745 genes were detected as expressed in the TRIKI-treatment assays . According to subsequent significance analysis of the microarray data ( SAM , q-value≤0 . 03 ) 2595 genes were found to be differentially transcribed when compared to the control . Of these , 2330 were protein-coding genes , while 265 had an antisense-orientation relative to the protein-coding gene in the given locus ( Supplementary Table S1 ) . A hierarchical clustering of the replicate data was performed ( Fig . 2A ) , showing that the transcription of 1765 protein-coding genes was enhanced ( up-regulated , red ) , and of 565 repressed ( down-regulated , green ) . The opposite tendency was detected for antisense transcripts , where 67 showed enhanced , and 198 repressed transcription . A list with all differentially transcribed genes is available in Supplementary Table S2 . Further functional analyses were done only with protein-coding genes . GO analyses of differentially transcribed genes revealed ontology categories enriched with genes being up- or down-regulated ( BH adjusted p-value≤0 . 05; Supplementary Table S3 ) . The GO category ncRNA metabolic process belonged to the ontology biological process , and contained genes with enhanced transcription . For genes with repressed transcription the ontologies biological process , including the category mRNA metabolic process , and cellular component were detected . Using IPA , a number of networks enriched with proteins coded by differentially transcribed genes were identified , of which the five most significant are presented ( summarised in Supplementary Table S4 ) . The first network included molecules involved in gene expression , protein synthesis , and amino acid metabolism . The second network contained molecules necessary for small molecule biochemistry , lipid metabolism , and amino acid metabolism . Molecules of the third network were associated with cell cycle , cellular function and maintenance , and molecular transport . The fourth network included molecules of nucleic acid metabolism , small molecule biochemistry and DNA replication , recombination and repair . The fifth network consisted of molecules belonging to lipid metabolism , nucleic acid metabolism , and small molecule biochemistry . A total number of 8016 genes were detected as expressed in the Herb A-treatment assays . SAM identified 1181 genes to be differentially transcribed with a q-value≤0 . 03 . Among these , 1021 represented protein-coding genes and 160 antisense-oriented transcripts ( Supplementary Table S5 ) . A hierarchical clustering of the protein-coding genes was performed ( Fig . 2B ) , showing that a majority of 719 genes had an enhanced transcription , whereas 302 exhibited a repressed transcription . The opposite pattern was found for the antisense-oriented genes . Here , 57 genes were enhanced in their transcription and 103 genes repressed ( see also Supplementary Table S1 ) . Further functional analyses were done only with probes representing protein-coding genes . GO analysis resulted in the identification of GO categories ( BH adjusted p-value of ≤0 . 05 ) significantly enriched only with genes showing enhanced transcription ( Supplementary Table S6 ) . The identified categories of the ontology biological process were negative regulation of molecular function , cellular carbohydrate metabolic process , protein folding , and glycoprotein metabolic process . The categories belonging to the ontology molecular function were peptidase regulator activity , enzyme inhibitor activity , and peptidase inhibitor activity . The cellular component ontology included genes belonging to the categories nuclear membrane-endoplasmic reticulum network and endoplasmic reticulum membrane . Among the five networks detected by IPA as most significantly enriched with proteins coded by genes with altered transcription , network 1 was comprised of molecules related to post-translational modification , protein folding , and molecules known from humans to be involved in cancer ( among other signal transduction proteins ) . The second network included molecules , whose functions are associated in humans with cancer , gastrointestinal disease , and genetic disorder . For the third network molecules involved in RNA post-transcriptional modification , DNA replication , recombination , and repair and energy production were enriched ( Supplementary Figure S1A ) . The fourth network contained molecules involved in endocrine system development and function , small molecule biochemistry as well as cellular function and maintenance . The fifth network comprised molecules involved in carbohydrate metabolism , cellular function and maintenance , and molecular transport ( Supplementary Table S7 ) . A total number of 11 , 668 genes were detected as expressed in schistosome parasites in the assays with Herbimycin A and TRIKI . Using SAM statistics , 521 genes were identified as differentially transcribed with a q-value of ≤0 . 03 ( Supplementary Table S8 ) . From these , 411 were protein-coding genes and 110 antisense-oriented transcripts , respectively ( Supplementary Table S1 ) . A hierarchical clustering was performed ( Fig . 2C ) , and showed that a higher number of protein-coding genes had an enhanced transcription ( 254 ) compared to those with repressed ( 157 ) transcription . The same picture was obtained for the antisense messages . Among these , 70 genes showed an enhanced transcription , 40 a repressed transcription upon the combined inhibitor treatment . Further functional analyses were conducted only with protein-coding genes . By GO analyses a significant enrichment ( BH adjusted p-value≤0 . 05 ) of genes with an enhanced , as well as repressed transcription was identified for different GO categories ( Supplementary Table S9 ) . In summary , enriched genes with enhanced transcription were part of GO categories including multicellular organismal process , localisation of cell , nucleobase , nucleoside , nucleotide and nucleic acid transport , RNA localisation , metabolic process , regulation of cellular metabolic process , regulation of biosynthesis , and cellular nitrogen compound metabolic process ( including the corresponding subcategories ) for the ontology biological process . The identified categories ( and corresponding subcategories ) of the ontology molecular function were: hydrolase activity , acting on acid anhydrides , structural molecule activity , binding , and calmodulin binding . Categories comprised of genes showing repressed transcription following the combined inhibitor-treatment included ion binding and peptidase activity , for the ontology molecular function . The ontology biological process included the categories membrane lipid biosynthetic process , protein modification process , and melanin biosynthetic process . Using IPA , two networks with significantly enriched proteins coded by differentially transcribed genes were identified . The first included molecules associated with post-translational modification , protein folding and cellular comprise . The second network comprised cellular function and maintenance , nucleic acid metabolism , and small molecule biochemistry ( Supplementary Table S10 ) . To get an overview of genes that were influenced by at least one inhibitor treatment , a data comparison was performed analysing the overlapping differentially transcribed protein-coding genes ( q≤0 . 03 ) and a Venn diagram was created ( Fig . 3; Supplementary Table S11 ) . Transcription of a large number of genes was affected by one treatment only . 1933 genes were exclusively found to be regulated by TRIKI , while the transcription of 659 genes was influenced by Herb A , and the transcription of 238 genes was affected by the combined treatment only . Regulation of 302 genes was affected by TRIKI as well as Herb A , but these genes were not detected as differentially transcribed in the data set from the combined approach . Of these , 87% were transcriptionally regulated in the same direction ( 213 genes were up- and 50 down-regulated in each of the two treatments ) , and 13% in inverse directions ( 18 genes were down-regulated and 21 up-regulated by Herb A treatment , with an opposite pattern in TRIKI treated worms ) . 113 genes were affected in their transcription by TRIKI or the combined treatment , of which only 32% were equally regulated ( 34 genes were up- and 2 genes down-regulated in either condition ) , and 68% of these genes were inversely regulated ( 13 genes were up- and 64 down-regulated by the combined treatment and showed an opposite change upon treatment with TRIKI alone ) . Treatment of worms with Herb A or the combined inhibitors resulted in the differential transcription of 78 genes . In contrast to the proportions of genes whose transcription was influenced by TRIKI-treatment or the combined inhibitors , here 72% were regulated in the same direction ( 46 genes were up- and 10 genes down-regulated in either condition ) , and 28% had an inverse pattern ( 21 genes were up- and 1 gene was down-regulated by the combined treatment and showed an opposite change upon Herb A treatment alone ) . The intersection of all three experiments comprised 18 genes , which included e . g . an hsp70-interacting protein , an immunophilin homolog , different hypothetical proteins , and calmodulin-4 . GO , IPA , and literature-based research were the basis of the selection of differentially transcribed genes , whose inhibitor-influenced transcription was validated by qPCR . Additionally , a manual classification of differentially transcribed genes was used for selecting further candidates . A huge number of signaling molecules were identified to be affected by at least one inhibitor treatment . Among these were several members of the schistosome TGFβ pathway [12] , [21] , [38] . Thus two TGFβ superfamily receptors , e . g . SmTβRI ( Smp_049760 ) , as well as SmActRIIb ( Smp_144390 ) were transcriptionally enhanced , whereas SmSmad4 ( Smp_033950 ) showed a repressed transcription following TRIKI-treatment . SmActRIIb and SmSmad 4 were transcriptionally regulated in the same direction following Herb A-treatment . Also genes involved in eggshell formation were found to be influenced by all inhibitor approaches . Selected candidates were the predicted eggshell precursor protein ( Smp_000430 ) and an eggshell protein similar to fs800 ( Smp_000270 ) [39] . Both showed enhanced transcription following TRIKI-treatment , which was unexpected due to the slight negative effect of TRIKI on egg production of female schistosomes shown before [11] . In contrast to this result , the combined treatment led to a repressed transcription for both genes , which was expected according to the strong reduction of the egg production following this dual inhibitor approach [11] . As another eggshell gene Smp48 ( Smp_014610 ) [40] was chosen , whose transcription was repressed following the Herb A-treatment . This inhibitor was shown before to negatively influence egg production of treated couples in vitro [11] , [29] . Finally , tyrosinase 1 ( SmTYR1 , Smp_050270 ) was selected , a gene which was localised in the vitellarium and shown to be responsible for cross-linking processes of eggshell precursor proteins [41] . SmTYR1 was transcriptionally repressed following the combined inhibitory treatment . As a candidate for surface proteins tetraspanin 18 ( Smp_174190 ) was chosen due to the strongest repressed transcription among the regulated tetraspanins according to TRIKI data analysis . Another member of this protein class was tetraspanin-1 ( Smp_011560 ) , which showed the same transcription tendency following Herb A-treatment . Furthermore , the combined inhibitor treatment revealed another tetraspanin 1 homolog ( Smp_155310 . 1 ) to be transcriptionally repressed . This molecule was also detected within the GO categories membrane and membrane parts . The mentioned tetraspanin homologs have not been characterized in schistosomes yet . Many heat shock protein genes were identified within the data sets of the individual Herb A- and the combined treatment . Among these was a hsp70 homolog ( Smp_106930 ) , which we selected as a representative of transcriptionally enhanced hsps following both inhibitor treatments . Furthermore , IPA of the data sets of both treatments identified hsp70 within network 1 , which includes molecules involved in protein folding , in each case . An impact of Herb A on protein folding processes was shown previously to result in the transcriptional activation of hsp70 [42] . Other molecules identified by IPA were the sodium/potassium-pump ( Na/K-pump; Smp_015020 ) and cathepsin S ( Smp_139240 ) . As a member of network 3 the Na/K-pump was detected with an enhanced transcription following TRIKI-treatment . Cathepsin S showed a repressed transcription following the combined inhibitor treatment , and it belonged to the network 1 . Further candidates were small nuclear ribonucleoprotein ( snurp; Smp_069880 ) , which we identified within the GO category “mRNA metabolic process” , and within the IPA network 2 of TRIKI data analysis . As a representative gene , whose transcription was affected by all three inhibitor treatments , calmodulin-4 ( Smp_032990 ) was selected for validation . It was additionally identified within the category binding of the GO analysis of the combined treatment data set . For the validation of the microarray data following TRIKI-treatment the genes coding for SmTβRI , SmActRIIb , a protein similar to fs800 , a predicted eggshell precursor protein , and Na/K-pump were finally selected as candidates for enhanced transcription ( Fig . 4A ) . Calmodulin-4 , tetraspanin 18 , SmSmad 4 , and a snurp were chosen as representatives of transcriptionally repressed genes ( Fig . 4B ) . In contrast to SmTβRI , SmActRIIb , Na/K-pump , and snurp the results of qPCR and microarray analyses correlated well for all other genes . The different results of the qPCR of both receptors , the Na/K pump , and snurp may be explained by biological variance between worm batches and/or by cross-hybridization . Contradictory results of qPCR and microarray for a small fraction of false-positive genes have already been documented in the literature [43] . Because of the high standard deviations of the microarray log2ratio-values , the contradictory results of qPCR and microarray for the Na/K-pump suggested that this gene might belong to the false-positive genes predicted by SAM . Due to the biological variance of the used worm batches , the calculated correlation coefficient according to Spearman was not significant comparing all qPCRs and the corresponding microarray data directly . For the validation of the microarray data following Herb A-treatment the genes coding for SmActRIIb , calmodulin-4 and hsp70 were chosen as representatives for enhanced transcription ( Fig . 5A ) . Tetraspanin-1 , SmSmad4 and Smp48 showed a repressed transcription in the microarray data ( Fig . 5B ) . The results of both analyses correlated well for calmodulin-4 , hsp70 , tetraspanin-1 , SmSmad 4 and Smp48 . The qPCR result obtained for SmActRIIb was contradictory to that of the microarray , which may have resulted from biological variations of the used worm batches . Nevertheless , the results obtained for these genes with both analyses significant correlated according to Spearman's Correlations Coefficient ( rs = 0 . 886 ) . The validation of the microarray data of females treated with both inhibitors included the transcriptionally enhanced genes calmodulin-4 and hsp70 ( see Fig . 6A ) . A repressed transcription was detected for the genes coding for a protein similar to fs800 , the eggshell precursor protein , tetraspanin 1 , SmTYR1 and cathepsin S ( Fig . 6B ) . An increase of transcript levels was confirmed for calmodulin-4 and for hsp70 , at least in one qPCR experiment . The identified reduction of transcripts was confirmed for tetraspanin 1 , SmTYR1 , and cathepsin S . For the gene similar to fs800 an increased transcript level was shown for two biological replicas by qPCR , but for the third qPCR replica and the microarray data a reduction of transcripts was determined . Both results correlated significantly according to Spearman's Correlations Coefficient ( rs = 0 . 7234 ) . During the study we obtained multiple evidence for inhibitor-induced differential transcription of different genes with known function in eggshell formation . In light of this and of previous evidence of a strong negative effect of Herb A as well as a moderate negative effect of TRIKI on egg production [11] , functional qPCR experiments were performed focusing on the analysis of a variety of candidates for protein-coding genes involved in this decisive step of the schistosome life cycle . The selected genes were the two well-characterized eggshell precursor protein Smp14 ( Smp_131110 . x; [44] ) and Smp48 ( Smp_014610; [40] ) as well as a predicted eggshell precursor protein ( Smp_000430 ) , which is still uncharacterized . Furthermore , we included the eggshell protein cross-linker SmTYR1 ( Smp_050270; [41] ) and the eggshell component , which was identified as similar to fs800 ( Smp_000270; [39] ) into this analysis . Towards this end paired female schistosomes were cultured in vitro under the same conditions as before for the microarray approaches to obtain a comparable basis for the transcriptional analyses . For this analysis the microarray data were used independent of their significance values . All three inhibitor treatments influenced the transcription of all genes selected ( Fig . 7 ) . In the majority of cases TRIKI-treatment resulted in an increase of transcript levels , which corresponded to the available microarray data . Due to the filtering criteria , the microarray data for Smp14 and Smp48 were not present within the microarray data sets although representative oligonucleotide probes existed . In contrast , SmTYR1 passed filtering during microarray data processing , but the transcriptional changes were not detected to be significant . Significant transcriptional changes were observed by microarrays for the gene similar to fs800 and the eggshell precursor protein following TRIKI-treatment . For Herb A-treated worms , a reduction of the transcripts of all genes was found by qPCR and this effect was stronger , when compared with the transcriptional changes of the combined inhibitor treatment . Furthermore , the reduced transcription of the qPCR-validated genes in Herb A-treated worms was in accordance with the findings of the microarray analysis . Here a repressed transcription was also determined for Smp14 , the eggshell precursor protein , and at least for two biological replicas for SmTYR1 and the protein similar to fs800 . The repressed transcription of Smp48 was the only significant transcriptional change within the microarray analysis , although the second replica showed a log2ratio of nearly 0 . The combined treatment , finally , led to a repressed transcription of all genes in both analyses , except two biological replicas used for the qPCR analysis for the fs800 gene . Although the tendency of down-regulation of the transcription of the selected genes corresponded well with the results obtained for Herb A-treated females , the effect of the combined treatment was found to be not as strong . Here , the detected decrease of transcripts was significant for the eggshell precursor protein , SmTYR1 and the gene similar to fs800 in the microarray analysis . In summary , the results confirmed a strong influence of the used inhibitors on transcriptional regulation of chosen genes involved in eggshell-formation processes . Furthermore , the data demonstrated a stronger effect of Herb A compared to TRIKI , which was in accordance with the observation of the physiological effects reported before [11] . Thus the results of our study provide the first molecular evidence that transcriptional regulation of genes involved in eggshell-formation processes is under the control of Src- and TβRI-containing signaling pathways . One additional finding of this study was that a number of genes with predicted calcium-associated functions were among those found to be differentially regulated ( see also Supplementary Figure S1B ) . To provide evidence for their potential contribution to egg production processes we performed localization studies investigating the tissue-specific transcription of candidate genes ( Fig . 8 ) . By in situ hybridization experiments , transcripts of hippocalcin , a neuronal calcium sensor [45] , [46] , were detected within the ovary , in the vitellarium and around the ootype ( Fig . 8 A , B ) , where according to classical literature the Mehlis' gland is located [47]–[49] . Transcripts of the potential calcium-influx channel protein ORAI-1 [50] were found to be expressed in ovary and vitellarium of the female and testes of the male ( Fig . 8 D , E ) . Transcripts of the predicted eggshell precursor protein gene were detected within the vitellarium ( Fig . 8 G ) . This was expected with respect to previous findings of the expression of similar genes such as p14 [44] , which was used in our study as positive control ( data not shown ) . Additionally , signals of the predicted eggshell precursor protein gene were also observed in the ovary ( Fig . 8 H ) , which was unexpected with regard to p14 that is predominately expressed in the vitellarium . Finally , calmodulin-4 transcripts were observed within the vitelloduct and around the ootype ( Fig . 8 J ) . In this case , the gene prediction indicated a small gene with the consequence that the probe used was relatively short compared to others normally used for this technique . Thus we cannot exclude being close to the detection limit in this case and that calmodulin-4 may be also transcribed in other organs . Furthermore , evidence for the presence of antisense transcripts of hippocalcin ( Fig . 8 C ) and calmodulin-4 ( Fig . 8 L ) was found since signals were obtained with sense RNAs in the same tissues . To confirm these transcription patterns , organ-specific RT-PCRs were performed ( Supplementary Figure S2 ) with template RNA of purified testes and ovaries obtained by a novel method for the isolation and enrichment of ovaries and testes [Hahnel et al . , submitted] . The results obtained confirmed and complemented the in situ findings providing additional evidence for two splice forms of ORAI-1 in testes and ovaries as well as calmodulin-4 transcription in testes .
Inhibitor experiments in previous studies and in this study have indicated that Src kinase- and TβRI-containing pathways influence mitotic activity and egg production in paired schistosome females [11] , [29] , and that SmTK3 [29] , [34] and SmTβRI are targets of Herb A or TRIKI , respectively . Furthermore , a yeast-two-hybrid ( Y2H ) cDNA-library screening and subsequent qualitative and quantitative analyses identifying and characterizing binding partners acting “downstream” of SmTK3 detected besides others a homolog of the BAF60 subunit of the SWI/SNF complex ( SmBAF60 ) and a diaphanous homolog ( SmDia ) as the strongest interacting partners [51] . The SWI/SNF complex is involved in chromatin-remodeling activities , DNA-damage responses , transcriptional activation , sliding of nucleosomes , and alteration of histone-DNA contacts [52] . Diaphanous proteins belong to the big group of formin-homology proteins known to play roles in actin-mediated processes controlling cell and tissue architecture , cell-cell interactions , gastrulation , and cytokinesis [53] . To detect genes being controlled by SmTK3- and SmTβRI-containing pathways , presumably playing roles for reproductive processes in adult schistosomes , microarray analyses were performed with RNA of inhibitor-treated paired females as template . A number of genes were detected to be differentially transcribed by individual inhibitors as well as their combination . Among these was a minor amount of antisense RNAs . Besides the possibility that some of these could have protein-coding function , the majority of these RNAs belong to the big group of non-coding RNAs ( ncRNAs ) , and their detection , especially during transcriptome analyzes , has opened a new research field since evidence has accumulated that ncRNAs may have regulatory functions [54] . After first evidence for the occurrence of antisense RNAs in the S . mansoni genome was obtained [36] , a recently performed detailed analysis estimated that around ≥10% of the transcribed genome may represent non-coding RNAs [37] . According to comparative life-stage analyses , differences in the transcription of schistosome ncRNAs were found indicating their potential roles in diverse biological and physiological processes . Thus it was no surprise to find a fraction of antisense RNAs also in our analysis as being differentially transcribed following treatment with individual or both inhibitors . Among these , some may interfere with regulatory processes . As soon as more knowledge about this class of molecules exists in schistosomes , the findings of our studies may contribute in the future to unravel their function . The highest number of differentially transcribed genes with protein-coding function was detected for treatment with TRIKI compared to the other treatments . The same tendency was observed when intersections were generated between single treatments and combined treatment , again more genes were found to be differentially transcribed when TRIKI was used . Within the intersection of the individual inhibitor treatments ( 302 genes; Fig . 3 ) , the majority of differentially transcribed genes were regulated in the same direction , a smaller part in the opposite direction . This indicates that genes within this intersection may be targets of both signaling pathways , which were previously hypothesised to cooperate during cell division and egg production processes [11] . The identity of genes found within this intersection indirectly support this assumption since genes well known for their role in mitosis such as e . g . the cell cycle check point protein rad 17 , cyclin 1 , or glypican ( Supplementary Table S11 ) were found . Unexpectedly , these genes were all found to be up-regulated by each inhibitor; this was not expected due to the previously observed inhibitory effect of Herb A and TRIKI on mitotic activity [11] . However , other genes contributing to mitosis regulation may have been negatively affected . Indeed , among the genes down-regulated by each inhibitor is a dynactin homolog ( Supplementary Table S11 ) . Dynactin is known to direct and coordinate the activities of the dynein motor , which is required for several cellular functions including cell division [55] . Indirect support for these conclusions comes from recent laser-microdissection microscopy ( LMM ) and oligonucleotide microarray analysis , which detected genes up-regulated two-fold or more in the gastrodermis , the ovary , the vitellarium , and the testes in S . mansoni and S . japonicum [56] , [57] . Among these were rad 17 ( ovary , vitellarium ) , cyclin 1 ( ovary ) , glypican ( ovary ) , and dynactin ( ovary , vitellarium ) as representatives of genes with elevated transcript levels indicating their functional relevance within the gonads . Furthermore , an IPA analysis of the combined inhibitor treatment predicted that the c-myc protein might have been activated , since genes theoretically regulated by c-myc were differentially transcribed . Along the same line IPA predicted the protein p53 to be inhibited following Herb A treatment . Since c-myc is able to cause proliferation inhibition ( mitoinhibition; [58] ) its putative activation may have contributed to the previously observed reduction of mitotic activity following treatment with these inhibitors . Moreover , c-myc expression and p53 inactivation were described as two cell-cycle events regulated by Src during mitosis [59] . Detailed clarifications of these points will be the subject of further studies . Interestingly , a homolog of Bcl2-associated athanogene ( BAG1 ) was discovered as being up-regulated by each single inhibitor treatment . BAG1 binds to Bcl2 , an oncogene inhibiting apoptosis , enhancing its anti-apoptotic effect . In this way BAG1 connects growth factor receptors with anti-apoptotic mechanisms [60] . Indeed , a recent publication provided first evidence that the proliferation of vitelline cells is independent of pairing , but their survival is male-dependent , being pairing-dependently controlled via apoptosis regulation [61] , and LMM-microarray analysis showed enhanced Bcl2 transcript-levels in the ovary and the vitellarium [57] . Furthermore , it is noteworthy that a number of genes involved in calcium regulation were differentially regulated such as homologs of hippocalcin [45] , [46] , [62] , [63] , different calmodulins [55] , or the calcium-influx channel protein ORAI-1 [50] . To provide further evidence for their role in reproduction localization studies were performed demonstrating that the schistosome homologs of hippocalcin , ORAI-1 , the predicted eggshell precursor protein gene and calmodulin-4 were transcribed within the reproductive system . Hippocalcin belongs to the calmodulin superfamily and exerts putative sAHP ( slow afterhyperpolarization ) function in the brain of higher eukaryotes [46] . To our knowledge , there is no information available yet about homologs in invertebrates . Its particular transcription pattern in schistosomes indicates a function of hippocalcin in the ovary , the vitellarium , and within the Mehlis' gland which is known to contribute to egg formation [49] . Since in trematodes Mehlis' glands are connected to the nervous system as shown by the expression of neuropeptides within these glands in S . mansoni as well as Fasciola hepatica [64] , [65] , it is tempting to speculate that schistosome hippocalcin may among further functions represent another neuronal player contributing to neurophysiological processes during egg formation . A neurophysiological function was also shown for the calcium channel protein ORAI-1 of D . melanogaster , which is required for normal flight and associated patterns of rhythmic firing of the flight motoneurons [66] . In C . elegans ORAI-1 knockdown caused complete sterility confirming its essential role in calcium signaling in the gonads [67] . With respect to the localization data obtained in our study , this may apply also to schistosome ORAI-1 . Especially interesting was a calmodulin-4 homolog , which was ( i ) present in the GO enrichment analysis of Herb A/TRIKI double-treated worms , ( ii ) a member of the group of 18 genes representing differentially expressed transcripts detected by all three inhibitor approaches , and ( iii ) seemed to be a target of both pathways being up-regulated following Herb A treatment and down-regulated following TRIKI-treatment as confirmed by our qPCR results ( see Figs . 4–6 ) . Links of both pathways to calcium mobilisation and signaling exist . Besides its role as a regulator of the type I TGFβ receptors , the immunophilin FKBP12 regulates the functional state of calcium channel receptors by altering their conformation and coordinating multi-protein complex formation [68] . Among others , Src-kinase signaling can lead to calcium mobilisation contributing to oocyte maturation and fertilisation [69] , [70] . A yet uncharacterized immunophilin was also among the 18 differentially transcribed genes identified by all three inhibitor approaches . Although not in each case ( ORAI1 , no observed up-regulation in the gonads ) , the above mentioned LMM analysis revealed Ca2+-metabolism-associated genes such as hippocalcin ( ovary , vitellarium ) as transcriptionally enhanced [57] . This and the findings from our study suggest that calcium may also contribute to the physiological processes controlling egg production . Besides the major intersection of the individual inhibitor treatments , the other intersections between the combined treatment and the individual treatments revealed further groups of genes ( 113/78; Fig . 3 ) being representatives of potential targets affected by either TGFβ ( 113; Fig . 3 ) or Src pathways ( 78; Fig . 3 ) . This indicates that one pathway may have a more dominant effect . With few exceptions such as an ( yet uncharacterized ) eggshell precursor protein within the TRIKI/combined treatment intersection ( 113; Fig . 3 ) , the majority of both intersections represented hypothetical proteins from S . mansoni ( Supplementary Table S11 ) , which may be novel , schistosome-specific targets of these pathways . The tendencies of transcriptional regulations between these two groups ( 113/78 ) were inversely correlated ( group 113: 32%/68% regulated in the same/opposite direction , respectively; group 78: 72%/28% regulated in the same/opposite direction , respectively ) . We interpret this as another evidence for a stronger response to the inhibition of the Src-kinase containing pathway ( s ) . A closer look on differentially transcribed genes following TRIKI-treatment revealed a lot of signaling molecules , which included members of the schistosome TGFβ pathway . All type I receptors of TGFβ superfamily were up-regulated by treatment with TRIKI , although only SmTβRI and SmBMPRI were significant for these data , in contrast to the SmActRI . Nevertheless , all three type I receptors showed enhanced transcription , which suggests feedback regulation . This is indirectly supported by the opposite regulation of follistatin , a regulator of TGFβ signaling [25] . It is part of the major overlap of the single inhibitor treatments , being up-regulated by TRIKI and down-regulated by Herb A ( 302; Fig . 3 , Supplementary table S11 ) . Beyond that SmSmad2 and SmSmad4 showed repressed transcription following individual TRIKI or Herb A treatments . The qPCR results indicated some additional influence by biological variability among the worm batches used affecting at least SmTβRI and SmActRI transcription , for which in one of three cases each the same regulatory tendency , as detected in the microarray data , was confirmed . LMM-microarray analysis showed enhanced transcript levels for SmSmad4 ( ovary ) , but not for SmTβRI [56] . A recently published study investigated the transcriptome of adult schistosomes following stimulation with hTGFβ using the same microarray platform [27] . The comparison of the differential transcription of genes upon hTGFβ ( ∑ 381 genes ) or TRIKI-treatment ( ∑ 1766 ) revealed 77 genes present in both analyses , of which 58 were regulated in the opposite direction ( listed in Supplementary Table S12; Supplementary Figure S3 ) . The higher overall number of differentially transcribed genes of the inhibitor treatment is probably caused by a lower stringency criterion in our study compared to the study of Oliveira et al . [27] , in which only genes with a log2ratio value ≥|1| were analyzed . Finally , hTGFβ but not BMP7 was shown to be able to bind SmTβRI [26] . Assuming that hTGFβ is not able to activate other type I receptors of the TGFβ superfamily , which is indirectly supported by the identification of the ligands SmBMP [71] and SmInAct [13] , the response may be specific leading to a narrow response window . Although previous experiments had shown that TRIKI acts specifically in comparison with others inhibitors ( like those of the BMP pathway; [72] ) , it is nonetheless able to inhibit also TGF-β RII , p38 MAPK , or mixed lineage kinase-7 [73] , [74] . Since only homologs of TGFβ RII and p38 MAPK exist in S . mansoni , and since inhibiting these would require 10–15 times higher TRIKI concentrations , we expected no alternate target effects . However , we cannot exclude effects on yet unknown targets in schistosomes . Of the 77 genes present in both analyses , 19 genes showed the same regulation pattern . This could be explained by the different sources of RNA used in both studies; whereas in our study RNA of paired females was used , RNA obtained from parasite couples was used in the other study [27] . Since the TGFβ pathway may fulfil different functions in both genders , the regulation of transcription affected by stimulation with hTGFβ can be expected to be different between and within the genders . This could lead to a bias in the transcriptome analyses , which would not be present if both analyses were done with RNA obtained from the same source , e . g . paired females only . Nevertheless , by comparing differentially transcribed genes detected in both treatments , the majority ( 75% ) of genes were found to be affected in the opposite direction , which is in line with the expected outcome of these inversely correlating approaches ( correlation of rs = −0 , 259 ) ( Supplementary Figure S3 ) . The analysis indicated that more genes were transcriptionally repressed than enhanced by hTGFβ , while more genes were enhanced than repressed by TRIKI such as the eggshell protein gene similar to fs800 ( Smp_000270; [39] ) . Herb A-treatment influenced the transcription of different members of the schistosome TGFβ pathway , although not all were significant as defined by our criteria . The homolog of the SmActRIIb belonging to the TGFβ superfamily as well as the SmSmad4 homolog showed the same regulation as observed by TRIKI-treatment ( for both homologs a significant influence on the transcription was detected ) . With respect to the hypothesised pathway cooperation , these findings supportively complement the TRIKI data . Additional support for pathway cooperation was obtained by the Y2H approach identifying SmTK3-interacting molecules , which besides SmBAF60 and SmDia identified a Smad 2/3 homolog as binding partner [51] . SRC/Smad binding during TGFβ-cooperative pathway activities was also found in other cellular systems [75] . The importance of both pathways for egg production in schistosomes was finally confirmed by qPCR experiments focusing on the question whether genes with proven and hypothesised functions for egg formation are influenced in their transcription by these inhibitors . This analysis also included genes from the microarray analysis not high-lighted as significant according to our criteria , such as Smp14 . Nonetheless , the qPCR results confirmed enhanced transcription of all genes following TRIKI-treatment , supporting the evidence of a negative influence of the TGFβ pathway on transcription of these genes . This finding was not expected due to the results of a former study , which indicated a slight decrease of egg production upon TRIKI-treatment [11] . However , eggshell formation is a highly complex process and depends on many genes , of which some -yet unknown and/or not represented by this analysis- may be negatively affected by TRIKI-treatment leading to the observed slight decrease in egg numbers . The complexity of egg production-associated processes was also demonstrated by a recent study , in which a direct link between egg production and the mitochondrial oxygen consumption was presented [76] . One of the crucial steps during this process seems to be fatty acid β-oxidation , which is initially catalysed within the mitochondria by acyl CoA dehydrogenase ( SmACAD ) . Within the microarray data obtained for Herb A-treatment , SmACAD transcription was significantly repressed . Since the decreased SmACAD activity was shown to be associated with a decrease of egg production [76] , and since SmACAD transcription was negatively influenced by Herb A , we conclude that this gene is also under the control of ( a ) Src kinase pathway ( s ) . Furthermore , Herb A treatment led to a strong negative effect on the transcription of the genes involved in eggshell formation , which perfectly correlated to the previous finding of a remarkably reduced number of eggs after treatment with this inhibitor . Although the decrease of Smp14 transcription was in contrast to Northern blot data from an elder study [29] , which may have been caused by biological ( batch ) variation , the more sensitive qPCR data obtained here corresponded to the microarray data . Furthermore , all other analyzed genes exhibited the same tendency of transcription regulation . This was also observed following the combined inhibitor treatment . Here , down-regulation of all studied genes was detected , however , it was not as strong as determined for worms treated exclusively with Herb A . From this we conclude that the transcriptional values determined for the combined inhibitor treatment represent an average of the values of both single-inhibitor treatment approaches , which revealed contrary transcription values with a bias towards the Herb A effect . This supports the view of cooperating pathways , but suggests opposing effects of SmTβRI and Src kinase pathways regulating egg production in a balanced way . Concerning the investigated eggshell-forming genes , the influence of both pathways was not equal , since the effect of Herb A dominated that of TRIKI ( Fig . 9 ) , which corresponded of the physiological data obtained previously [11] . Although not directly compared in one qPCR experiment , the same tendency was also observed for calmodulin-4 transcription upon treatment with Herb A ( strong up-regulation ) , or TRIKI ( weak down-regulation ) , or both inhibitors ( up-regulation ) . Thus calmodulin-4 may represent another target molecule of the cooperative pathway activity . The findings presented in this study extend our knowledge on mechanisms controlling reproduction in schistosomes , which is an important , but not yet understood process against the background of understanding basic principles leading to female maturation and egg production in schistosomes and other trematodes . For the first time it is shown that the cooperative actions of TβRI and Src kinase-containing pathways are involved in the control of egg formation , a process not only essential for schistosome life-cycling , but also for the pathological consequences of the disease . In light of the necessity to find alternatives to Praziquantel , the only drug applied worldwide to fight schistosomiasis [77] , our findings may also open novel perspectives for alternative concepts to interfere with the transmission of schistosomes by negatively influencing egg production thus interrupting the parasite life cycle . Here kinases have already suggested their potential as targets by in vitro experiments [78] .
A Liberian isolate of Schistosoma mansoni was maintained in the intermediate host Biomphalaria glabrata , and Syrian hamsters ( Mesocricetus auratus ) as final host [79] , [80] . 42 days post infection adult worms were obtained by hepatoportal perfusion . All animal experiments have been done in accordance with the European Convention for the Protection of Vertebrate Animals used for experimental and other scientific purposes ( ETS No 123; revised Appendix A ) and have been approved by the Regional Council ( Regierungspraesidium ) Giessen ( V54-19 c 20/15 c GI 18/10 ) . After perfusion with M199 medium ( Gibco; including glucose , sodium bicarbonate , 4- ( 2-hydroxyethyl ) -1-piperazineethane sulfonic acid ) , the worms were washed twice with this medium and subsequently cultured in M199 supplemented with FCS ( Gibco; 10% ) , HEPES ( Sigma; 1M , 1% ) , and antibiotic/antimycotic mixture ( Sigma; 1% ) at 37°C and 5% CO2 , as described previously [33] . For inhibitor treatments , worm couples were left in culture for 2 days adapting to the in vitro environment to restore full egg production capacity [11] , [29] . Both Herbimycin A ( Herb A , Enzo Life Science; CAS: 70563-58-5; 1 mg/ml ) and the TβRI kinase inhibitor ( TRIKI , or LY-364947; purchased from Calbiochem; CAS: 396129-53-6; 5 mg/ml ) were dissolved in dimethyl sulfoxide ( DMSO ) . TRIKI ( C17H12N4 ) is a pyrazole-based inhibitor ( 3- ( pyridin-2-yl ) -4- ( 4-quinonyl ) -1H-pyrazole ) with an IC50 value of 51 nM for human TβRI ( Calbiochem; [73] , [74] ) . It binds to the active site of the TβRI kinase domain and is less effective inhibiting TGF-β RII ( IC50 = 400 nM ) , p38 MAPK ( IC50 = 740 nM ) , or mixed lineage kinase-7 ( MLK-7; IC50 = 1 , 400 nM ) . For transcriptome studies the cultivation of adult schistosomes was done for 2 days with either 4 . 5 µM Herb A , 300 nM TRIKI , or the combination of both inhibitors ( H+T ) with the same concentrations . As control , worms were cultivated in medium containing DMSO . Medium and additives were refreshed daily . Pairing stability and vitality were checked each day . As vital and useful for further experiments we considered couples , whose males sucked with their ventral suckers to the Petri dish , while keeping the female within the gynaecophoric canal . Vital worms performed uniform , wave-like movements , showed regular gut peristaltic , and produced eggs . Separation of couples and/or failing of males to suck to the Petri dish , resting on the side and showing reduced wave-like movements were considered as signs of decreasing vitality . Such worms were removed from the dishes and not considered for further analyses . Worms used for the experiments ( treated and control samples likewise ) were carefully separated by pipetting or using featherweight tweezers , immediately shock-frozen in liquid nitrogen , and stored at −80°C for further use . To test whether TRIKI was able to inhibit SmTβRI ( Smp_049760 ) , its intracellular part was cloned into the expression vector pcDNA 3 . 1/V5-His B . To this end this region was amplified by PCR with cDNA as template and the primers TGFβRI_intra_BamHI-5′ ( 5′-GGATCCTACTTCCTCTGGAGAAGGAAATC-3′ ) and TGFβRI_intra_EcoRV-3′ ( 5′-GATATCTAAATGCTTTGAATTACTATTGTTATTGG-3′ ) . Both primers contained specific restriction sites ( 5′ primer BamHI; 3′ primer EcoRV ) , which were used for directed cloning of the amplification product into the vector . The obtained wild type ( wt ) SmTβRI-pcDNA 3 . 1/V5-His construct was commercially sequenced ( LGC Genomics , Berlin ) to confirm the correct open reading frame ( ORF ) . To convert wt SmTβRI into a constitutively active variant , the Ser and Thr residues of the GS motif ( position 284–290 ) as well as the Thr ( position 299 ) and Gln ( position 303 ) were mutated into 7 Asp ( SmTβRI7D ) as described earlier [35] . For negative control , an inactive kinase variant was generated by exchanging Thr residues to Val , as well as by exchanging Ser and Gln residues to Ala ( SmTβRIVVAAAVV ) . These mutations were done successively by site-directed mutagenesis , using the SmTβRI-pcDNA 3 . 1/V5-His wt construct as template ( 25 ng ) and the following primers: TGFβRI_Mut-1 ( 5D ) -5′ ( 5′-GATGGACCACGATGACGATGGGGACGGTGACGGAAAACCTTTACT-3′ ) + TGFβRI_Mut-1 ( 5D ) -3′ ( 5′-AGTAAAGGTTTTCCGTCACCGTCCCCATCGTCATCGTGGTCCATC-3′ ) ; TGFβRI_Mut-2 ( 7D ) -5′ ( 5′-CCTTTACTAGTTCAGCGAGATGTCGCTAGGGACGTTCAGTTGG-3′ ) + TGFβRI_Mut-2 ( 7D ) -3′ ( 5′-CCAACTGAACGTCCCTAGCGACATCTCGCTGAACTAGTAAAGG-3′ ) ; TGFbRI_Mut-1 ( VVAAA ) -5′ ( 5′-GATGGACCACGTTGTCGCTGGGGCAGGTGCCGGAAAACCTT-3′ ) + TGFβRI_Mut-1 ( VVAAA ) -3′ ( 5′-AAGGTTTTCCGGCACCTGCCCCAGCGACAACGTGGTCCATC-3′ ) ; TGFβRI_Mut-2 ( VV ) -5′ ( 5′-CCTTTACTAGTTCAGCGAGTGGTCGCTAGGGTAGTTCAGTTGG-3′ ) + TGFβRI_Mut-2 ( VV ) -3′ ( 5′-CCAACTGAACTACCCTAGCGACCACTCGCTGAACTAGTAAAGG-3′ ) . For PCR the proofreading Pfu DNA-Polymerase ( Promega; 3 U/µl ) was used followed by DpnI digestion for 1 h at 37°C to digest remaining wt vector DNA [81] . Following bacteria transformation and plasmid-DNA isolation , sequencing confirmed the correct ORFs of the constructs . Subsequently , cRNA synthesis was done as previously described [32] , and 60 ng each used for microinjection into Xenopus laevis stage VI oocytes [31] , [32] . The oocytes were cultured in ND96 medium at 19°C for 18 h . Germinal vesicle break down ( GBVD ) , a witness of meiosis progression dependent on kinase activity [30] , is characterized by the development of a white spot at the animal pole of the oocyte . As a positive control for GVBD , oocytes were stimulated with progesterone . As negative controls , non-injected oocytes as well as oocytes transfected with the inactive kinase variant SmTβRIVVAAAVV were used . For inhibitor studies , oocytes were cultured in ND96 medium supplemented with different concentrations of TRIKI ( 3 nM , 30 nM , and 300 nM ) for 18 h . RNA from inhibitor-treated or from control females was isolated using Trizol reagent ( Invitrogen ) . Subsequently , a DNAse digestion was done with the RNAeasy kit ( Qiagen ) according to the manufacturer's manual . The quality of the isolated RNA was determined using Bioanalyzer microfluidic electrophoresis ( Agilent Technologies ) . For microarray experiments a S . mansoni custom-designed oligonucleotide platform ( 60-mers ) was used , with approximately 44 , 000 probes representing nearly the complete S . mansoni transcriptome based on available cDNA sequence data from S . mansoni and S . japonicum . This platform was produced by Agilent Technologies ( described in [36] , and all associated information ( probes , annotation ) is available at Gene Expression Omnibus ( GEO ) under the accession number GPL8606 . From each sample of the three inhibitor treatments , 300 ng RNA were used for cDNA amplification followed by Cy3 and Cy5 labelling during in vitro transcription using the Quick Amp Labelling Kit , two colors ( Agilent Technologies ) . Labelling included a dye-swap approach as an internal technical replica for each sample . Thus , six microarray hybridizations were performed per inhibitor treatment and corresponding controls , including two technical replicas for each of the three biological replicas . For hybridization , 825 ng cRNA of each labelled inhibitor sample were used and combined with a control sample labelled with the opposite dye . Hybridization was done at 65°C for 17 h with rotation . The slides were washed and scanned with the Gene Pix 4000B Scanner ( Molecular Devices ) according to the Agilent manual . The obtained raw data were extracted using Feature Extraction software ( Agilent Technologies ) and are available under GEO study number GSE39732 . For subsequent analyses , a gene was considered as expressed only if its corresponding probe exhibited signals that were significantly higher than background ( employing default parameters from the Feature Extraction software and recovering the column “IsPosAndSig” from the output ) . A probe had to fulfil the criterion to have detectable expression in at least 75% of all replicas in at least one of the two conditions ( inhibitor-treated or control ) . LOWESS algorithm was used for normalisation of the intensities [82] , and the log2ratios between inhibitor-treated and control groups were calculated . Subsequently , an adjustment of these filtered data was done using an updated genome annotation to eliminate redundancy of the probes per gene [37] . A low overall correlation was observed for the technical replicates Herb/DMSO 4 and H+T/DMSO 1 , leading to the exclusion of these experiments from further analyses . As a consequence , the technical replicate TRIKI/DMSO 4 treatment was removed additionally to achieve a comparable basis for the analysis of all three microarray approaches; the latter technical replicate was selected for exclusion based on its lowest total intensities of the fluorescence signals . To identify genes with a significant change in the levels of transcribed message , SAM ( Significance Analysis of Microarrays ) was used [83]; genes with a q-value≤0 . 03 were considered to have a significantly differential level of transcribed message between inhibitor-treated and control samples . The subsequent functional analyses focused on probes representing protein-coding genes ( labelled as “to be used in GO = YES” in the updated annotation of the array [37] ) , although antisense-oriented oligonucleotide probes were present on the microarray platform as well . Hierarchical clustering was done using Spotfire [84] . Functional analyses were performed for differentially transcribed genes; Gene Ontology ( GO ) analyses [85] of all three inhibitor data sets were done with the software tool Ontologizer [86]; parent child union [87] was selected to identify categories containing enriched genes , and the p-value was adjusted according to Benjamini-Hochberg ( BH ) correction [88] . Further functional analyses were performed by Ingenuity Pathway Analysis ( IPA; http://www . ingenuity . com; [89] ) , which represents a tool providing curated information from the literature for human , mouse and rat models about canonical pathways , regulated transcription factors and their targets , and possibly regulated molecular networks , including signal transduction cascades ( of which some contribute to human cancer and other diseases ) . To this end , all S . mansoni genes were annotated with the corresponding human homolog ( determined according to a search with the following blast parameters: e-value<10−10 and at least 60% coverage ) ; genes that fulfilled these criteria were annotated with the label “to be used in IPA = YES” in the updated annotation [37] , and they were uploaded to IPA along with their corresponding microarray transcription measurements . These putative homologs represented the basis for the identification of pathways and networks enriched with proteins encoded by significantly differentially transcribed genes; default settings were used , except for presentation of networks ( 140 molecules per network , number of networks: 10 ) and for “user dataset” as reference set . To validate the transcriptional changes caused by inhibitor treatment of selected genes quantitative PCRs ( qPCRs ) were performed on a Rotor Gene Q ( Qiagen ) . RNAs from inhibitor-treated or control females were isolated using TriFast reagent ( PeqLab ) , and 1 µg each was reverse transcribed using QuantiTect Rev . Transcription kit ( Qiagen ) according to the manufacturer's instruction . The amplified cDNA was diluted 1∶20 and used for subsequent qPCR analyses . The detection of synthesised DNA double strands was based on the incorporation of SYBRGreen using PerfeCTa SYBR Green Super Mix ( Quanta ) . To distinguish between the specific amplification product and unspecific primer dimers following each qPCR analysis a melting point analysis was done . Primer 3 Plus software was used for primer design ( http://www . bioinformatics . nl/cgi-bin/primer3plus/primer3plus . cgi ) . The amplification products had a size between 140 and 160 bases . Primers were designed to have melting points at 60°C . Furthermore , differentiation of amplification products of cDNA and genomic DNA was obtained by designing primers that flanked predicted introns of the appropriate genes . A list of all used primers is available in Supplementary Table S13 . All primers were commercially synthesised by Biolegio ( Netherlands ) . Since standard reference genes normally used for relative quantification analyses such as α-tubulin , Cu/Zn SOD ( superoxide dismutase ) , or histone showed regulation following inhibitor treatment , we decided to perform absolute quantification on the basis of standard curves generated by purified PCR products ( used in dilution series ) [90] . These quantifications were done for both inhibitor- and DMSO-treated samples . Fold changes are available in Supplementary Tables S2 , S5 , and S8 . Subsequently , log2ratios ( treated/control ) were calculated according to a previous study providing a solid basis for comparison of microarray and qPCR data [91] . The efficiency of each qPCR was determined to be between 90–100% . Finally , the validity of the obtained ratios of qPCRs and microarrays was determined by Spearman's rank correlation coefficient ( rs ) as well as the correlation of the intersection of the microarray data following hTGFβ and TRIKI-treatment [92] , [93] . To detect the occurrence of transcripts of ORAI-1 , hippocalcin , the predicted eggshell precursor protein gene , and calmodulin-4 of S . mansoni , in situ hybridizations were performed as described earlier [33] , [51] . To this end , adult worm pairs were fixed in Bouin's solution ( picric acid/acetic acid/formaldehyde; 15/1/5 ) and embedded in paraplast ( Histowax , Reichert-Jung ) before sections of 5 µm were generated and incubated in xylol . Following re-hydration , proteins were removed by proteinase K treatment ( freshly prepared , final concentration 1 µg/ml ) , and the sections were dehydrated . For hybridization , transcripts were generated in vitro by RT-PCR , checked for their identity by sequencing , and labeled with digoxigenin following the manufacturers' instructions ( Roche ) . The following primer combinations were used for amplification ( hippocalcin , Smp_085650: forward 5′-GCTATTTATGCGATGGTTGGC-3′ , reverse 5′-GACTCTGAGGTATCAGGAATGAC-3′; ORAI-1 , Smp_076650 . 1: forward 5′-GTTGTCGTGCATATAATGGCT-3′ , reverse 5′-CTGGACTCCACTTCTAAGAAAGG-3′; eggshell precursor protein gene , Smp_000430: forward 5′-GTTCCAATTACCAACCAACGTC-3′ , reverse 5′-GTTTCCGTTACCACCATAATTACC-3′; calmodulin-4 , Smp_032990: forward 5′-ATGAATGTTCCAATAACTCGTGAAG-3′ , reverse 5′-AAGTGCTCTTGTTAATTCTGGTAAAC-3′ ) . Primers were 5′-tagged by the addition of the T7-sequence ( 5′-TAATACGACTCACTATAGGGAGA-3′ ) to allow RT-PCR-based product synthesis of antisense or sense probes using T7 polymerase ( Roche ) . PCR conditions to amplify calmodulin-4 were: denaturation 95°C 45 s , annealing 56°C 45 s , elongation 72°C 45 s; 30 cycles; for all other transcripts: denaturation 95°C 45 s , annealing 60°C 45 s , elongation 72°C 30 s; 30 cycles . Labeled transcripts of hippocalcin ( 359 bp ) , ORAI-1 ( 428 bp ) , the hypothesized eggshell precursor protein gene ( 533 bp ) , and calmodulin ( 210 bp ) were size-controlled by gel electrophoresis . To prove their quality , transcript blots were made confirming digoxigenin-incorporation by alkaline phosphatase-conjugated anti-digoxigenin antibodies ( Roche ) with naphthol-AS-phosphate and Fast Red TR ( Sigma ) . All in situ hybridizations were performed for 16 h at 42°C . Sections were washed up to 1× SSC , and detection was achieved as described for transcript blots . Testes and ovaries were isolated by a recently established organ-isolation method [Hahnel et al . , submitted] . In short , adult males or females ( 50–60 each ) were treated with 500 µl of tegument solubilisation ( TS ) -buffer ( 0 . 5 g Brij35 , 0 . 5 g Nonidet P40 , 0 . 5 g Tween80 , and 0 . 5 g TritonX-405 per 100 ml PBS ( 137 mM NaCl , 2 . 6 mM KCl , 10 mM Na2HPO4 , 1 . 5 mM KH2PO4 in DEPC-H2O , pH 7 . 2–7 . 4 ) ) at 37°C and 1 , 200 rpm in a shaker for 5 min to solubilise the tegument . This step was repeated for females ( 1× ) and males ( 2× ) followed by washing steps ( 3× ) with M199 medium ( 2 ml ) . To remove the muscles , elastase IV from pancreas ( Sigma , #E0258 ) was used ( 5 units/ml , in M199 medium ) , and 500 µl added to each sample followed by slight agitation ( 600 rpm ) in a shaker at 37°C for about 30 min . During incubation , the worms were swirled up manually every 5 min . This reaction was stopped when the medium turned opaque and the worms were fragmented but not completely digested . At this point liberated intact organs were observed . Testes and ovaries were identified by their characteristic morphology and carefully transferred by pipetting to fresh M199 medium . For quality inspection bright field microscopy was performed and if necessary , the organs were separated from remaining tissue rests by additional washes and transfers . Finally , the organs were collected by a pipette , transferred to 1 . 5 ml tubes , and concentrated by centrifugation for 5 min at 1 , 000 g ( testes ) or 1 min at 8 , 000 g ( ovaries ) . After removal of the supernatant the organs were frozen in liquid nitrogen and stored at −80°C for further use . For cDNA synthesis the QuantiTect Reverse Transcription Kit ( Qiagen ) was used with 500 ng of total RNA as template following the instruction of the manufacturer . PCR was done with 2 µl of a 1∶40-dilution ( testis cDNA ) or 1∶80-dilution ( ovary cDNA ) of each cDNA-sample in a total volume of 50 µl containing 1× reaction buffer ( 80 mM Tris-HCl , 20 mM ( NH4 ) 2SO4 , 0 . 02% w/v Tween20 , 2 . 5 mM MgCl2 ) , 200 µM dNTPs , 400 nM of each primer and 2 . 5 units Fire-Pol taq polymerase ( Solis BioDyne ) . The reactions were performed in a MasterCycler ( Eppendorf ) programmed as follows: 1 cycle , 95°C , 2 min; 35 cycles , 95°C , 45 sec; 60°C , 45 sec; 72°C , 45 sec . The primers used to amplify hippocalcin , eggshell precursor protein gene , and calmodulin-4 were mentioned above , primers used for ORAI-1 amplification were newly designed to be able to detect both splice variants of this gene ( Smp_076650 . 1 , Smp_076650 . 2; forward 5′- ACGTTGTTACTTCTTCAGTACTCC-3′; reverse 5′-ACTTTGTAGGTAGTAAGCGCAC-3′ ) . As positive control for similar amounts of cDNA of each organ-sample the S . mansoni heat shock protein 70 gene ( SmHSP70 accession number L02415; forward 5′-TGGTACTCCTCAGATTGAGGT-3′; reverse 5′-ACCTTCTCCAACTCCTCCC-3′ ) was used since as it was described to be expressed throughout diverse life stages and tissues , and it turned out to be a suitable control [94; Hahnel et al . , submitted] . The following public domain tools were used: SchistoDB ( http://schistodb . net/schisto/ ) , Gene DB ( http://www . genedb . org/Homepage ) , Welcome Trust Sanger Institute S . mansoni OmniBlast ( http://www . sanger . ac . uk/cgi-bin/blast/submitblast/s_mansoni/omni ) , BLAST ( http://blast . ncbi . nlm . nih . gov/ ) , Clustal W2 ( http://www . ebi . ac . uk/Tools/msa/clustalw2/ ) , InterPro ( http://www . ebi . ac . uk ) , SMART ( http://smart . embl-heidelberg . de/ ) . TRIKI , TβRI kinase inhibitor; Herb A/Herb , Herbimycin A; H+T , Herbimycin A combined with TRIKI , TGFβ , transforming growth factor beta; hTGFβ , human TGFβ; SmTβRI , S . mansoni type I TGFβ receptor; SmActRIIb , S . mansoni type IIb Activin receptor; SmSmad , S . mansoni Smad; SmTYR1 , S . mansoni tyrosinase 1; SmTK3 , S . mansoni tyrosine kinase 3; SmFKBP12 , S . mansoni FK506-binding protein 12; fs800 , female specific protein 800; ORAI-1 , calcium release-activated calcium channel protein 1; qPCR , quantitative PCR; CLSM , confocal laser scanning microscopy; GVBD , germinal vesicle break down; SAM , Significance Analysis of Microarrays; GO , Gene Ontology; IPA , Ingenuity Pathway Analysis; BH , Benjamini-Hochberg correction; snurp , small nuclear ribonucleoprotein; hsp , heat shock protein | As one of the most prevalent parasitic infections worldwide , schistosomiasis is caused by blood-flukes of the genus Schistosoma . Pathology coincides with egg production , which is started upon pairing of the dioeciously living adults . A constant pairing contact is required to induce mitoses and differentiation processes in the female leading to the development of the gonads . Although long known , the molecular processes controlling gonad development or egg-production in schistosomes or other platyhelminths are largely unknown . Using an established in vitro-culture system and specific , chemical inhibitors we have obtained first evidence in previous studies for the participation of signal transduction processes playing essential roles in controlling mitoses , differentiation and egg production . In the present study we applied combinatory inhibitor treatments combined with subsequent microarray and qPCR analyses and demonstrate for the first time that cooperating Src-Kinase- und TGFβ-signaling pathways control mitoses and egg formation processes . Besides direct evidence for managing transcription of eggshell-forming genes , new target molecules of these pathways were identified . Among these are calcium-associated genes providing a first hint towards a role of this ion for reproduction . Our finding shed first light on the signaling mechanisms controlling egg formation , which is important for life-cycling and pathology . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"gene",
"networks",
"animal",
"genetics",
"gene",
"regulation",
"microbiology",
"parasitology",
"dna",
"transcription",
"gene",
"function",
"parasite",
"physiology",
"genome",
"analysis",
"tools",
"genome",
"databases",
"molecular",
"genetics",
"microbial",
"growth",
"and",
"development",
"quantitative",
"parasitology",
"gene",
"expression",
"comparative",
"genomics",
"biology",
"pathogenesis",
"drug",
"discovery",
"biochemistry",
"rna",
"nucleic",
"acids",
"gene",
"identification",
"and",
"analysis",
"genetics",
"microbial",
"control",
"genomics",
"genetics",
"of",
"disease"
] | 2013 | Transcriptome Analyses of Inhibitor-treated Schistosome Females Provide Evidence for Cooperating Src-kinase and TGFβ Receptor Pathways Controlling Mitosis and Eggshell Formation |
Mutations affecting the ribosome lead to several diseases known as ribosomopathies , with phenotypes that include growth defects , cytopenia , and bone marrow failure . Diamond-Blackfan anemia ( DBA ) , for example , is a pure red cell aplasia linked to the mutation of ribosomal protein ( RP ) genes . Here we show the knock-down of the DBA-linked RPS19 gene induces the cellular self-digestion process of autophagy , a pathway critical for proper hematopoiesis . We also observe an increase of autophagy in cells derived from DBA patients , in CD34+ erythrocyte progenitor cells with RPS19 knock down , in the red blood cells of zebrafish embryos with RP-deficiency , and in cells from patients with Shwachman-Diamond syndrome ( SDS ) . The loss of RPs in all these models results in a marked increase in S6 kinase phosphorylation that we find is triggered by an increase in reactive oxygen species ( ROS ) . We show that this increase in S6 kinase phosphorylation inhibits the insulin pathway and AKT phosphorylation activity through a mechanism reminiscent of insulin resistance . While stimulating RP-deficient cells with insulin reduces autophagy , antioxidant treatment reduces S6 kinase phosphorylation , autophagy , and stabilization of the p53 tumor suppressor . Our data suggest that RP loss promotes the aberrant activation of both S6 kinase and p53 by increasing intracellular ROS levels . The deregulation of these signaling pathways is likely playing a major role in the pathophysiology of ribosomopathies .
Diseases linked to mutations affecting the ribosome include inherited disorders such as Diamond-Blackfan anemia ( DBA ) , Shwachman-Diamond syndrome ( SDS ) , and dyskeratosis congenita ( DC ) [1] . They may also be acquired as with 5q-myelodysplastic syndrome ( 5q-MDS ) [2] . While the phenotypes of these disorders vary extensively in clinical features and severity , the majority of them share some form of cytopenia . DBA , for example , is a pure red cell aplasia linked predominantly to mutations in ribosomal protein ( RP ) genes [3] . Patients with DBA experience a block in erythroid progenitor cell division and expansion in the bone marrow leading to the characteristic erythroblastopenia [4] , [5] . Growth defects , which are routinely observed in animal models of RP gene haploinsufficiency , are also common clinical features of patients with DBA as well as SDS and very severe forms of DC [6]–[10] . While it has been speculated that the hematopoietic phenotype at least in DBA patients is linked to the activation of the p53 tumor suppressor [11] , the mechanistic understanding of the pathophysiology underlying DBA and other diseases linked to mutations affecting the ribosome remains incompletely understood . Autophagy is the highly conserved cellular process of self-digestion that involves the formation of double-membrane structures termed autophagosomes engulfing cytoplasmic proteins and organelles [12] . These autophagosomes then fuse with lysosomes to become autolysosomes , wherein the proteins and organelles are degraded and then either recycled or exocytosed [13] . Autophagy is commonly observed during times of nutrient depletion or starvation and is up regulated in response to oxidative stress or the presence of deleterious organelles and protein aggregates [14] . Autophagy also plays a critical role in erythrocyte maturation . Conditional knockout of Atg7 , essential for the formation of the autophagosome membrane , yields hematopoietic stem cell failure , erythrocyte cell death , and severe anemia in mice due to the defective removal of mitochondria by autophagy ( mitophagy ) [15] . Additionally , the targeted deletion of the BCL-2 family member Nix results in defective erythroid maturation through impaired mitophagy during terminal erythroid differentiation and also causes anemia in a murine model [16] . The nutrient-sensitive AKT/target-of-rapamycin ( TOR ) pathway plays a critical role in controlling cell growth and size by stimulating the transcription of a number of factors required for protein translation including RP genes [17] , [18] . TOR-dependent autophagy induced by starvation or rapamycin treatment occurs through decreasing the TOR-dependent phosphorylation of Atg13 , an event required for association of Atg13 with Atg1 and subsequent autophagosome formation [19] . One important downstream effector of TOR signaling is S6 kinase , whose substrates include the translation machinery elements RPS6 , eukaryotic initiation factor 4B ( eIF4B ) , and eukaryotic elongation factor 2 kinase ( eEF2K ) [20]–[22] . S6 kinase phosphorylation also promotes the formation of autophagosomes , which in combination with the negative regulation of autophagy by TOR provides a balance to prevent cells from excessive self-digestion during prolonged periods of starvation [23] . One of the many activators of AKT and TOR is a phosphorylation cascade initiated by the stimulation of cells with the extracellular growth factor insulin and signaling through the insulin receptor substrate ( IRS1 ) and PI3-kinase to promote the turnover of PIP2 to PIP3 [24] . The highly conserved insulin pathway is required for the cellular import of glucose , the most vital carbohydrate for activation of the glycolysis pathway and the generation of ATP . AKT activation is as effective as insulin stimulation in inducing the expression of the glucose transporter 1 ( GLUT1 ) , the most highly expressed glucose transporter on glycolysis-dependent erythrocytes , while in other cells such as adipocytes insulin-activated AKT promotes the translocation of GLUT4 to the plasma membrane to increase glucose import [25] , [26] . Resistance of the insulin pathway to stimulation by extracellular ligands can occur through the over activation of S6 kinase , that in turn directly phosphorylates IRS1 leading to its degradation [27] , [28] . This mechanism of insulin resistance is commonly found in obese individuals with type II diabetes , due in large part to increased caloric intake and constitutive activation of the TOR pathway and S6 kinase . A loss of insulin signaling is also known to induce autophagy [29] , [30] . While insulin signaling can repress autophagy through TOR phosphorylation of Atg13 , a TOR-independent mechanism of repression has been described involving the AKT-dependent phosphorylation and functional inhibition of beclin-1 , the yeast homolog of Atg6 , an important initiator of autophagosome formation [31] , [32] . In this study we set out to determine if the mechanism inducing autophagy from RP loss is linked to the insulin pathway in a TOR-independent manner .
RPS19 is the gene most commonly mutated in DBA patients [3] . We therefore selected siRNAs against RPS19 to study the effects of its knock down in HEK cells stably expressing GFP-LC3 , a widely used reporter construct for measuring autophagy levels [33] . Figure 1A shows a western blot analysis of RPS19 expression in GFP-LC3 cells transfected with siRPS19 or a scrambled siRNA control ( siScr ) compared to the effects of treatment with 100 nM rapamycin , an inhibitor of TOR that is known to decrease the expression of RPs [18] . The quantification of Figure 1A in 1B indicates that the level of RPS19 knock down obtained is approximately 50% , equivalent to levels expected in DBA patients with RPS19 haploinsufficiency . Knock down of RPS19 in GFP-LC3 cells induces autophagy , as shown by western blot analysis with an antibody that recognizes the 18 kD LC3-I and the 16 kD LC3-II protein ( Figure 1C ) . LC3-I is converted to LC3-II upon autophagosome formation due to the conjugation of phosphatidylethanolamine ( PE ) that enables LC3-II to migrate faster on SDS/PAGE gels , and comparing the levels of LC3-II to actin is a reliable method to measure the level of autophagy [33] . The formation of autophagosomes upon RPS19 knock down can also be visualized by the increase of GFP-LC3 puncta found in the cytoplasm of siRPS19 transfected cells compared to cells transfected with siScr ( Figure 1D top panels and quantified in Figure 1E as the number of puncta/cell , N>100 , p = 0 . 002 ) . Bafilomycin A ( bafA ) is a drug that inhibits vacuolar H+ ATPase and is commonly used to block the fusion of autophagosomes with autolysosomes [34] . The lower left panel of Figure 1D shows a marked increase in autophagosome formation in GFP-LC3 HEK cells transfected with siScr and treated with bafA as compared to untreated cells , indicating a high degree of steady-state autophagic flux in normal HEK cells . However , a large amount of GFP-LC3 remains cytoplasmic . In cells transfected with siRPS19 and treated with bafA ( Figure 1D , lower right panel ) the cytoplasmic GFP-LC3 is almost completely gone and practically all the GFP-LC3 is distributed in a puncta fashion , indicating localization in autophagosomes and/or autolysosomes ( quantified in Figure 1F as the percent of cells with cytoplasmic GFP ) . We confirmed the presence of these structures by immuno-electron microscopy of HEK cells with reduced levels of RPS19 ( Figure 1G ) . These micrographs reveal autolysosomes that contain degraded organelles , including mitochondria ( left panel ) , where we find a strong presence of LC3 using immuno-gold labeling ( right panel ) . Because DBA-linked mutations in RPS19 primarily affect the function of CD34+ erythroid progenitor cells in humans , we infected CD34+ cells isolated from cord blood with shRNAs and cultured them in erythroid culture medium . On day nine after isolation , we compared the levels of LC3-II/actin with cells that were not infected ( NI ) , infected with a control shRNA ( shScr ) , or infected with shRNAs against RPS19 ( shRPS19 ) . Figure 1H shows that the successful knock down of RPS19 is coupled to an increase of the ratio of LC3-II/actin , suggesting an increase in autophagy as a result of RP knock down in CD34+ erythroid progenitor cells . Another method that allows for visualization of autophagosome formation is immunofluorescence ( IF ) with antibodies against LC3 [35] . We thus performed confocal IF analysis using lymphoblastoid cell lines ( LCLs ) derived from DBA patients carrying mutations causing haploinsufficiency of RPS17 , RPL11 , or RPS7 . These IF experiments , illustrated in Figure 2A , reveal distinct puncta representative of LC3-II incorporation into developing autophagosomes and autolysosomes in all three DBA-derived LCLs compared to very few puncta detected in normal cells ( quantified in Figure 2B , N>100 , p<0 . 01 ) . Additional evidence of increased autophagy in DBA LCLs is provided by western blot analysis revealing an increase in LC3-II/actin ratio in the RP haploinsufficient cells ( Figure 2C ) . The p62 ( sequestosome-1/SQSTM1 ) protein is a major substrate for autophagy that becomes degraded upon increased autophagy [36] . Figure 2D shows a decrease in p62 protein levels by western blot analysis in DBA-derived LCLs compared to normal control cells , 3 independent experiments of which are quantified in Figure 2E . A significant decrease of p62 is also detected by IF in cells derived from DBA patients , which is highly significant when the total amount of p62 expression per cell area is measured ( N>100 , p<0 . 01 ) ( Figures 2F-G ) . Lastly , electron micrographs in Figure 2H reveal the presence of many autolysosomes in the RPS17+/- mutant cells compared to only small lysosomes in the normal control cells . Together , these data indicate that mutation of the RP genes in cell lines derived from DBA patients have increased levels of autophagy . The zebrafish lines we selected to study carry viral inserts in the introns of RP genes that , coupled to the gradual loss of maternally contributed RPs , result in a progressive decrease of RP expression [37] . Although eventually lethal to the embryos , we emphasize that homozygosity of these inserts is synonymous with knock down models of RP gene loss ( the only animal models that faithfully recapitulate the anemia phenotype of DBA [38]–[40] ) and not deletion mutants ( hereafter referred to as RP-deficient embryos ) . Figure 3A illustrates this using western blot analysis of rpS7 expression in wild type or rpS7-deficient embryos at 1 and 2 days post fertilization ( dpf ) . When normalized to actin , rpS7 expression at 1 dpf in the rpS7-deficient embryos is ∼57% ±5 ( N = 3 ) of the expression level in wild type embryos . By 2 dpf , this drops to less than 5% . We therefore examined the red blood cells ( RBCs ) at 1 dpf , when rpS7 levels in the mutant embryos are characteristic of RPS7 haploinsufficiency seen in DBA patients [3] . Figure 3B shows representative electron micrographs of zebrafish RBCs at 1 dpf ( which retain their nuclei , unlike human RBCs ) . The upper left panel is a RBC from a wild type embryo showing mildly condensed chromatin in the nucleus and uniformly distributed , unadulterated cytoplasm . In contrast , the cytoplasm of rpS7-deficient mutant RBCs is mottled with circular double-membrane circular structures , a characteristic feature of autophagosomes ( arrowheads in Figure 3B lower left panel , upper right panel , and highlighted in the lower right panel ) . Additionally , autolysosomes are visible in these mutant RBCs ( Figure 3B , small arrows in lower left panel ) . While small double-membrane structures are occasionally observed in 1 dpf wild type RBCs , they appear much more frequently in the mutant embryos , as quantified in Figure 3C . By 2 dpf the majority of the RBCs in rpS7-deficient embryos have disappeared , evident by staining the embryos for hemoglobin with o-dianisidine in Figure 3D . The results are quantified in Figure 3E by scoring each embryo in the total clutch as either “Normal” , “Mild” , or “Severe” depending on the level hemoglobin loss , followed by genotyping of the individual embryos to confirm true homozygotes . Despite the loss of RBCs in 2 dpf embryos , electron microscopic analysis of rpS7-deficient embryos reveals the presence of characteristic double-membrane structures in other tissues at this stage . These structures originating from both the mitochondria ( Figure 3F , upper right panel ) and Golgi apparatus ( Figure 3F , lower right panel ) are shown engulfing cytoplasmic material . Representative images of these organelles in wild type embryos are shown in the left panels of Figure 3F . The mitochondrial structures we visualize in the rpS7-deficient embryos very closely resemble the previously reported mitochondria that supply membranes for autophagosomes in starved mammalian cells [41] , and the circular mitochondria that are formed in response to mitochondrial oxidative damage [42] . It has been previously reported that the knock down of rpS19 or rpS14 in zebrafish embryos using morpholino oligonucleotides increases phosphorylation of S6 kinase [43] . In agreement with these previous results , we detect an increase in p70 phospho-S6 kinase signaling at Thr389 in the GFP-LC3 HEK cells transfected with the siRNAs against RPS19 compared to siScr , a signal that is completely abolished by the addition of rapamycin ( Figure 4A ) . Given that rapamycin is well known to directly inhibit TOR , this suggests that TOR is likely phosphorylating S6 kinase in this model . An increase in phosphorylation of S6 kinase is also observed upon knock down of RPS19 in CD34+ cells ( Figure 4B ) . We additionally detect an increase in phospho-S6 kinase signaling in the LCLs with haploinsufficient RP gene mutations , including an increase in p85 S6 kinase phosphorylation ( Figure 4C ) . Similar to human cells with RP loss or mutations , we observe a large increase in S6 kinase phosphorylation in our genetic models of RP deficiency ( Figure 4D ) . Unfortunately , despite several attempts we were unable to find a commercially available S6 kinase antibody that cross-reacted with zebrafish . However , given the strength of the S6 kinase phosphorylation signal and the normal expression of actin , we are confident that this increase is not due to a large increase in S6 kinase expression . Taken together , these results demonstrate that the loss or mutation of several RPs in both human and zebrafish cells results in an increase of S6 kinase phosphorylation . Overactive S6 kinase phosphorylation causes the inhibition of the insulin pathway , in which AKT plays a central role [27] , [28] . One known mechanism of this inhibition is through a direct phosphorylation step of IRS1 by S6 kinase that in turn targets IRS1 for degradation [28] . To determine if RP loss affects downstream mediators of the insulin pathway , we performed western blot analysis on cells with reduced RPs to measure levels of IRS1 and phosphorylated AKT substrates . A decrease of IRS1 is observed in siScr transfected GFP-LC3 HEK cells when stimulated with insulin for 6 hours , similar to what has been shown previously in adipocytes treated with insulin over time [44] ( Figure 5A ) . We also observe by densitometer analysis an ∼50% decrease in IRS1 expression in cells transfected with siRPS19 compared to siScr , suggesting that the increase in S6 kinase phosphorylation observed in the siRPS19 cells results in degradation of IRS1 . To determine the effect of RP loss on AKT activity , we performed western blotting with an antibody that recognizes the phosphorylation motif of AKT substrates , ( RXXS*/T* ) . Transfecting cells with siRPS19 resulted in a significant reduction of phosphorylated AKT substrates compared to siScr , while treatment with insulin had the expected effect of increasing the AKT substrate phosphorylation in cells transfected with either siRNA ( Figures 5B-C ) . Similarly , western blot analysis of several different RP-deficient zebrafish embryos reveals a substantial decrease in the phosphorylation of AKT substrates ( Figure 5D ) . Taken together the data suggest that RP loss results in inhibition of AKT phosphorylation activity , likely driven by the degradation of IRS1 . In order to determine if activation of the insulin pathway could override the signals inducing autophagy in models of RP loss , we treated siRNA-transfected GFP-LC3 HEK cells with 350 nM insulin and performed confocal analysis and western blotting . Figures 6A-B show that the addition of insulin significantly reduced the number of GFP-LC3 puncta in the siRPS19-transfected cells to the same levels as siScr cells ( N>100 , p<0 . 01 ) . This reduction of autophagy is also revealed in the LC3 western blots shown in Figure 6C , where the increase of LC3II/actin ratio in the siRPS19 cells is reduced upon insulin stimulation to the ratio observed in the cells transfected with siScr . When we performed a titration of insulin on LC3-GFP HEK cells transfected with siRNAs against RPS19 we found that as little as 10 nM insulin was sufficient to significantly reduce the number of autophagosomes compared to untreated cells ( Supporting Figure S1 ) . We therefore conclude that the down-regulation of insulin pathway signaling , evident in the data presented in Figure 5 , is coupled to the induction of autophagy by RP loss . The rapamycin experiment in Figure 4A shows that the increased phosphorylation of S6 kinase linked to RP loss is dependent on TOR . Given the mitochondrial modifications revealed by electron microscopy , we hypothesized that one potential mechanism for this activation may be linked to the TOR function of sensing abnormal levels of intracellular ROS [45] . Trolox , a vitamin E derivative , functions as an anti-oxidant by the intracellular scavenging of ROS [46] . The addition of Trolox to GFP-LC3 HEK cells resulted in a significant reduction in the number of autophagosomes , abolishing the basal level observed in siScr-transfected cells and diminishing the abnormally high level observed in the siRPS19 cells ( N>100 , p<0 . 02 ) ( Figures 7A-B ) . This reduction of autophagy is coupled to a decrease in the phosphorylation of S6 kinase , shown by western blot analysis in Figure 7C . Moreover , when we treated RP-deficient zebrafish embryos with 10 mM Trolox overnight we observed a substantial decrease in the levels of S6 kinase phosphorylation in embryos carrying mutations in rpS7 or rpS3a , which in turn alleviated the inhibition of AKT phosphorylation activity ( Figure 7D ) . Interestingly , the Trolox treatment while clearly blocking autophagosome formation in the GFP-LC3 HEK cells was unexpectedly coupled to an increase of LC3-II by western blotting analysis ( Supporting Figure S2 ) , the latter a feature of Trolox that has been previously reported [47] . However , in the same cells we also find that Trolox results in an increase of p62 expression and a decrease of phosphorylated S6 kinase , supporting the confocal results indicating that there is a block of autophagy due to a reduction of S6 kinase signaling ( Supporting Figure S2 ) . The wild type and rpL11 embryos are not represented in Figure 7D because they do not survive the 10 mM Trolox treatment . This is compared to a survival rate of over 80% of the rpS3a embryos and ∼40% of the rpS7 embryos ( Figure 7E ) . Interestingly , these survival rates correlate with the severity of the morphological phenotypes typically observed with RP mutants including smaller heads/eyes , inflated hindbrain ventricles , and the presence of pericardial edemas [48] . Embryos are morphologically affected most acutely by the rpS3a mutation , followed by the rpS7 mutation , while the rpL11 mutation is comparatively milder ( Supporting Figure S3A ) . In other words , the increased survival of the rpS3a and rpS7 deficient mutants upon the Trolox treatment is likely due to the their higher initial ROS levels compared to wild type or rpL11 deficient mutants . This is illustrated in Supporting Figure S3B . Many cellular stresses , including oxidative damage and ROS , result in stabilization of the p53 tumor suppressor and the induction of apoptosis [49] . In zebrafish embryos , a homozygous mutation in the DNA binding domain of the p53 gene ( M214K ) allows for an intensified and sustained stabilization of p53 in response to stress [50] , [51] . We therefore crossed the rpS7-deficient embryos with the p53 mutant background in order to enhance p53 detection by western blotting ( Figure 7F ) . These double mutant embryos treated with 100 µM of Trolox overnight revealed a considerable reduction of the level of p53 stabilization ( Figure 7G ) . These data suggest that the phosphorylation of S6 kinase , the inhibition of AKT activity , the subsequent induction of autophagy , and the stabilization of p53 due to RP loss are due to an increase of intracellular ROS . This is represented by a graphic illustration in Figure 8 . Our study so far focuses on DBA-linked RP gene mutations , however deregulated autophagy may be a more general phenotype of ribosomopathies . To explore this possibility we examined cells derived from SDS patients , most of which carry a mutation in the SBDS gene that is important for one of the final maturation steps of the 60S ribosomal subunit [52] , [53] . Primary mononuclear cells ( MNCs ) isolated from the peripheral blood of two independent SDS patients revealed an increase in expression of LC3-II to actin compared to MNCs from a healthy individual ( Figure 9A ) . Moreover , we found evidence of increased autophagy in non-hematopoietic primary fibroblasts derived from four other independent SDS patients . Figures 9B-D illustrate with confocal and western blot analysis that levels of the autophagy substrate p62 are substantially diminished in cultured fibroblasts derived from SDS patients compared to those from a healthy individual . Taken all together , the data suggest that the deregulation of autophagy is a common aspect of ribosomopathies that are linked to bone marrow failure . Moreover , given that the majority of the experiments were performed in non-erythroid cells , they suggest a general relevance of autophagy and inhibition of the insulin pathway that likely extends beyond the cytopenia phenotype of ribosomopathies .
This study demonstrates conclusively that mutations in RP genes induce autophagy through ROS-mediated S6 kinase phosphorylation . Other animal models , including Drosophila and zebrafish , have shown previously that different mutations affecting the processing of rRNA induce autophagy , however the mechanisms have remained incompletely understood [54] , [55] . The study of the zebrafish defective in rRNA processing by mutation of pwp2h suggests that the mechanism is likely p53- and TOR-independent , despite the fact that the authors also note in their mutants an increase in phosphorylation of rpS6 ( a substrate of S6 kinase downstream of TOR ) [55] . The loss of RPs in zebrafish and mouse models of DBA has also been recently linked to major changes in metabolism , particularly glycolysis [40] . We therefore speculated that a link exists between the autophagy induced by mutations affecting the ribosome and the insulin pathway . Similar to other zebrafish models of RP loss using morpholinos [56] , we observe an increase in phosphorylation of S6 kinase in our genetic RP-deficient zebrafish embryos and in human cells with RP loss which is dependent on TOR and independent of AKT . Since reducing ROS levels with the antioxidant Trolox decreases this S6 kinase phosphorylation and reduces the autophagy activated by RP loss , we propose that increasing ROS levels is the link between RP loss and S6 kinase phosphorylation by TOR . This relationship between ROS , TOR , and S6 kinase has been previously reported as the result of nutrient overload , raising the likelihood that metabolic stress sensed by mitochondria is communicated to the TOR pathway directly through ROS [28] , [45] . ROS is increasingly becoming recognized as a critical signaling molecule and not just a damaging agent [57] . Our data underscore this by showing the importance of maintaining ROS levels at a particular threshold during the developmental process , and how reducing this threshold with Trolox is lethal to wild type , but not RP-deficient embryos with severe morphological phenotypes . Interestingly , other recent reports are beginning to reveal links between different inherited bone marrow failure disorders with increasing levels of intracellular ROS including DC , SDS , and Fanconi anemia [58]–[60] . These reports along with our study raise the possibility that the phenotypes of these bone marrow failure disorders may share a common basis that is increased ROS released from stressed mitochondria . Our study goes on to suggest that the stabilization of p53 , long regarded as the mechanism behind the early death of erythroid progenitor cells in DBA patient bone marrow , is also triggered by increasing ROS . Since hematopoietic progenitor cell death in Fanconi anemia has been definitively linked to p53 stabilization [61] , our results suggest a general mechanism of cell death induced by ROS may underlie these congenital bone marrow failures . The increase in S6 kinase phosphorylation that we observe upon RP loss is coupled to a decrease in the phosphorylation of AKT substrates , suggesting a mechanism similar to the over activation of S6 kinase promoting insulin resistance . The study mentioned previously is in line with this showing rpL11-/- zebrafish embryos and whole blood of rpL11+/- adult fish contain higher levels of glucose compared to wild type controls , an accumulation that would be expected if cells are unable to import and metabolize glucose properly [40] . This same study also measured a decrease of glycolytic enzymes in the rpL11-/- embryos and in fetal mouse liver cells with knocked down Rps19 [40] . It has been reported that insulin signaling is important for the proliferation of late erythroblasts and that impaired insulin signaling can cause severe growth defects in animal models [62]–[64] . Therefore insulin pathway inhibition may provide an additional layer of regulation underlying the cytopenia phenotypes , and suggests a putative mechanism of the growth defect phenotypes that are both present in DBA and common to other ribosomopathies . The conflicting results regarding the Trolox treatment in Supporting Figure S2 are worth noting . As mentioned , the increase of LC3-II that we observe with western blotting analysis despite a clear block of autophagosome formation has been previously reported in cells treated with Trolox , but the results in this report are in the absence of any confocal or IF data showing LC3 localization [47] . Consistent with our results it has been reported that many other antioxidants block autophagy [65] . Therefore it appears antioxidants inhibit autophagosome formation while simultaneously increasing LC3-II accumulation . These data underscore the limitations of using LC3 western blotting as the sole readout of autophagy activity , and highlight the importance of performing concurrent LC3 localization experiments . Our study describes how the impairment of insulin signaling and deregulation of autophagy result from mutations of RPs . Moreover , we demonstrate that these pathways may be involved in the etiology of a broader range of ribosomopathies that are linked to other mutations affecting the ribosome . While it seems clear that increased levels of ROS as a result of these mutations induces deregulation of autophagy , insulin signaling , and activation of the p53 , it is of course difficult to conclusively state the contribution of each impairment to the early death of hematopoietic cells in ribosomopathy patients . Most likely this early death is a combination of all these and other deregulated pathways . Nonetheless , our study suggests that therapeutics targeting intracellular ROS levels may be an effective treatment for patients with bone marrow failures that are linked to mutations affecting the ribosome .
Animal experiments were conducted in accordance with the Dutch guidelines for the care and use of laboratory animals , with the approval of the Animal Experimentation Committee ( Dier Experimenten Commissie [DEC] ) of the Royal Netherlands Academy of Arts and Sciences ( Koninklijke Nederlandse Akademie van Wetenschappen [KNAW] ( Protocol # 08 . 2001 ) . Zebrafish were raised and embryos obtained through natural spawning as previously described ( Westerfield , The Zebrafish Book . A Guide for the Laboratory Use of Zebrafish , 3rd Edition , 1995 ) in accordance with all Dutch regulations and guidelines . The GFP-LC3 HEK cells were a kind gift from Dr . Paul Coffer's lab . For the lymphoblastoid cell lines ( LCLs ) , RPS17 cells carry a deletion in exon 3 ( c . 200_201delGA ) resulting in a frameshift at codon 67 and an early stop codon at 86 . RPL11 cells carry an insert in exon 3 ( c . 160_161insA ) resulting in a frameshift at codon 54 and an early stop at codon 66 . RPS7 cells carry a donor splice site mutation in intron 3 ( IVS3+1 g>a ) . SDS MNCs were obtained by blood samples collected in a heparin vacutainer , diluted 1∶1 in PBS , layered over half the total volume of Ficoll ( GE Healthcare ) and centrifuged at 287 RCF for 20 min . The MNC and monocyte layer was then removed and washed 2x in 1xPBS . Cell lines were incubated at 37°C +5% CO2 , the LCLs in RPMI medium +10% FCS +1% pen/strep and fibroblasts in DMEM media +10% FCS +1% pen/strep . All primary cells and patient-derived cell lines used in this study were from patients and families that gave informed consent according to the Declaration of Helsinki . siRNAs against RPS19 were purchased from Invitrogen ( #4392420 ) . “Stealth RNAi” scrambled siRNAs from Invitrogen were used as a control ( #12935-400 ) . GFP-LC3 HEK cells were plated in 24-well plates ( or on glass-bottom confocal dishes , Greiner Bio-One GmbH , #627870 ) the day before transfection with DMEM media . Transfections were done with Oligofectamine ( Invitrogen , #12252-011 ) according to the manufacturer's instructions using the maximum recommended amount of siRNAs . Bafilomycin A ( Sigma , #B1793 ) was added at a final concentration of 50 nM for 4 hours . Rapamycin ( Sigma , #R8781 ) was added at a final concentration of 100 nM for 6 hours . 10 mM of Trolox ( Sigma #238813 ) was added overnight . Insulin ( Sigma , #I6634 ) was added at a final concentration of 350 nM for 6 hours . Confocal analysis was performed on a Leica SPE microscope . At least 8 shots per transfection were taken , and at least 3 transfections per condition were performed . The counter in ImageJ was used to quantify the number of puncta per cell and the number of cells with cytoplasmic GFP ( the images were unlabeled to prevent bias ) . Zebrafish embryos were lysed as previously described [66] . Trolox was added to embryos in E3 media overnight at 10 mM . LCLs and fibroblasts were lysed on ice using 1% Triton X-100 buffer containing protease and phosphatase inhibitor tablets ( Roche ) and normalized to 25 µg per sample . Lysates were loaded and fractionated by SDS-PAGE ( 8% for IRS1 and pAKT substrates; 10% gels for p62 , pS6 kinase , and S6 kinase; and 12% for LC3 and RPS19 ) under reducing conditions and immunoblotted on PVDF membranes . Primary antibodies used were RPS7 ( Santa Cruz #sc-100834 ) at a dilution of 1∶250 , RPS19 ( Santa Cruz #sc-100836 ) at 1∶1000 , LC3 for LCLs at 2 µg/ml ( Novus Biologicals #NB100-2331 ) , LC3 for MNCs , HEK and CD34+ cells at 1∶1000 ( Millipore #ABC232 ) , phospho-S6 kinase ( Cell Signaling #92345 ) at 1∶1000 , S6 kinase ( Cell Signaling #92025 ) , p62 at 1∶1000 ( Progen #GP62-C ) , and actin antibodies ( Santa Cruz #sc-1616 ) at 1∶200 . The zebrafish-specific p53 antibody ( zp53 ) was raised as previously described [66] . Secondary antibodies were diluted 1∶5000 including rabbit-HRP ( GE Healthcare ) for LC3 , pS6 kinase , and S6 , mouse-HRP ( GE Healthcare ) for RPS7 , RPS19 , and zp53 blots , guinea pig-HRP ( Abcam #ab97155 ) for p62 , and goat-HRP ( Santa Cruz #sc-2020 ) for actin blots . ECL reagent ( Amersham Biosciences ) was used for detection either with Kodak Biomax XAR film or with ImageQuant LAS 4000 ( GE Healthcare ) . Densitometer analysis was performed using a GS-800 densitometer ( BioRad ) and Quantity One software ( BioRad ) . Embryos were left in their chorions and treated with 10 mM Trolox in E3 media overnight . The survival of the embryos was assessed by the presence of a heartbeat . For electron microscopy analysis , LCLs , GFP-LC3 HEK cells , and 1 and 2 dpf zebrafish embryos were fixed in a mixture of 2% paraformaldehyde and 0 . 2% glutaraldehyde in 0 . 1 M phosphate buffer . The cells or embryos were washed in PBS-glycine to quench free aldehydes , then embedded in gelatin and infiltrated in 2 . 3 M sucrose , followed by rapid freezing in liquid N2 . 50 nM thick cryosections were cut at −120°C using an Ultracut-S ultra microtome ( Leica Microsystems , Vienna , Austria ) . Sections were either directly viewed in a JEOL 1200CX electron microscope ( Jeol Ltd . Japan ) , or after immuno-gold labeling with rabbit anti-LC3B ( Abgent , #AP1805a or AP1806a , Oxfordshire UK ) , followed by 10 nm protein-A gold . For LCLs , Shandon cytospins ( 5 min at 35 RCF ) were used to place cells on 12 mm coverslips ( VWR ) , cells were then fixed and permeabilized with methanol-acetone ( 1∶1 volume ) . Fibroblasts were grown overnight in DMEM +10% FCS +1% pen/strep on coverslips , then fixed as above . Primary PBMCs were initially fixed in 2% paraformaldehyde in PBS , then placed on coverslips with Shandon cytospins and permeabilized with 0 . 1% Triton X-100 . Cells were blocked with PBS and 0 . 2% fish skin gelatin ( Sigma ) at RT twice for 10 min each . All antibodies were diluted in this blocking buffer . Primary antibodies against LC3B ( Abgent , Oxfordshire UK ) were diluted 1∶100 and p62 ( Progen ) at 1∶300 , then added to the blocked cells for 30 min at RT . Secondary antibodies GaR-Alexa488 ( against LC3B , Invitrogen ) and DkaGP-Alexa488 ( against p62 , Biotium ) were all diluted 1∶100 and incubated on the cells in the dark for 30 min at RT . Cells were mounted in Prolong Gold + DAPI ( Invitrogen ) . Subcellular distribution patterns were analyzed using a Zeiss LSM-510 confocal microscope . Individuals counting the number of cells in each confocal slice that displayed the LC3-positive puncta carried out quantifications of the puncta staining . Average numbers were displayed as a percentage of the total number of cells per image ( unlabeled images were used to prevent bias ) . For p62 quantification ImageJ software was used to calculate the total amount of p62 staining per total number of cells using the following parameters: File = 8 bits , threshold = 40–255 , size = 0–infinity , and circularity = 0 . 00–1 . 00 . CD34+ cells were isolated from cord blood using the immunomagnetic technique ( Miltenyi Biotec , Paris , France ) . CD34+ cells were cultured for two days in presence of 10% FBS , 100 U/ml IL-3 , 10 ng/ml IL-6 , 25 ng/ml SCF , 10 ng/ml TPO , and 10 ng/ml Flt3-L . A first infection with shRNA-RPS19 compared to non-infected cells or shRNA-Scr infection ( all the shRNA generation and lentiviral production is previously described [67] ) was then performed at a 50 MOI . For shRNA-RPS19C , a second infection was performed 6 hours after the first infection . Cells were cultured for two more days and GFP-positive cells were sorted using the FACSDIVA Cell Sorter . Sorted cells were switched to the same IMDM medium with SCF ( 50 ng/ml ) , IL3 ( 1 U/ml ) and EPO at 1 U/ml till D7 when the FBS concentration was increased to 30% ( from D7 to D14 ) . Pellets of 100 , 000 cells were frozen until lysing . O-dianisidine staining was performed on 2 dpf embryos as previously described [68] . The genotypes of embryos were verified in a 3-primer PCR reaction: 5′-CTCTTGGATGGCTTGGACATGC-3′ , 5′- CACTATTTTCGCGCTGGTACTGAAC-3′ ( which paired give a 570 bp band with wild type gDNA ) and a primer that anneals to the viral insert nLTR3 , 5′-CTGTTCCATCTGTTCCTGAC-3′ which yields a 275 bp band in the presence of the viral insertion in the rpS7 gene . Statistics in all experiments were performed using the Student's two-tailed t-test . | Diseases linked to mutations affecting the ribosome , ribosomopathies , have an exceptionally wide range of phenotypes . However , many ribosomopathies have some features in common including cytopenia and growth defects . Our study aims to clarify the mechanisms behind these common phenotypes . We find that mutations in ribosomal protein genes result in a series of aberrant signaling events that cause cells to start recycling and consuming their own intracellular contents . This basic mechanism of catabolism is activated when cells are starving for nutrients , and also during the tightly regulated process of blood cell maturation . The deregulation of this mechanism provides an explanation as to why blood cells are so acutely affected by mutations in genes that impair the ribosome . Moreover , we find that the signals activating this catabolism are coupled to impairment of the highly conserved insulin-signaling pathway that is essential for growth . Taken together , our in-depth description of the pathways involved as the result of mutations affecting the ribosome increases our understanding about the etiology of these diseases and opens up previously unknown avenues of potential treatment . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"genetics",
"of",
"disease",
"insulin-dependent",
"signal",
"transduction",
"signal",
"transduction",
"developmental",
"biology",
"cell",
"biology",
"molecular",
"development",
"genetics",
"tor",
"signaling",
"biology",
"and",
"life",
"sciences",
"molecular",
"cell",
"biology",
"cell",
"signaling",
"gene",
"function"
] | 2014 | Ribosomal Protein Mutations Induce Autophagy through S6 Kinase Inhibition of the Insulin Pathway |
Active adult stem cells maintain a bipotential state with progeny able to either self-renew or initiate differentiation depending on extrinsic signals from the surrounding microenvironment . However , the intrinsic gene regulatory networks and chromatin states that allow adult stem cells to make these cell fate choices are not entirely understood . Here we show that the transcription factor DNA Replication-related Element Factor ( DREF ) regulates adult stem cell maintenance in the Drosophila male germline . A temperature-sensitive allele of DREF described in this study genetically separated a role for DREF in germline stem cell self-renewal from the general roles of DREF in cell proliferation . The DREF temperature-sensitive allele caused defects in germline stem cell self-renewal but allowed viability and division of germline stem cells as well as cell viability , growth and division of somatic cyst stem cells in the testes and cells in the Drosophila eye . Germline stem cells mutant for the temperature sensitive DREF allele exhibited lower activation of a TGF-beta reporter , and their progeny turned on expression of the differentiation factor Bam prematurely . Results of genetic interaction analyses revealed that Mi-2 and Caf1/p55 , components of the Nucleosome Remodeling and Deacetylase ( NuRD ) complex , genetically antagonize the role of DREF in germline stem cell maintenance . Taken together , these data suggest that DREF contributes to intrinsic components of the germline stem cell regulatory network that maintains competence to self-renew .
Adult stem cells maintain tissues during the lifetime of an organism by replenishing short-lived differentiated cells such as in the skin , intestinal epithelium and blood . Adult stem cells also give rise to differentiated cells upon injury in tissues such as skeletal muscle and lung . To maintain tissue homeostasis , daughter cells produced by adult stem cell divisions must make the critical cell fate decision between self-renewal and the onset of differentiation . Deviation from the tightly regulated balance between these alternate fates may result in poor tissue maintenance or cancerous growth of poorly differentiated precursor cells[1] . Adult stem cells are thus in a bipotential state , able to self-renew or to initiate differentiation in response to extrinsic signals from the surrounding microenvironment[2 , 3] . This bipotential state relies on intrinsic transcriptional and chromatin programs that dictate how stem cells respond to external signals from the niche . Here we show that in Drosophila male germline adult stem cells , the transcription factor DNA Replication-Related Element Factor ( DREF ) and members of the Nucleosome Remodeling and Deacetylase ( NuRD ) complex , Mi-2 and Chromatin Assembly Factor 1 ( Caf1 , also known as p55 ) , act antagonistically to regulate the balance between germline stem cell self-renewal and differentiation . In the adult testis , two populations of stem cells , germline stem cells ( GSCs ) and somatic cyst stem cells ( CySCs ) , reside adjacent to a group of post-mitotic somatic cells called the hub . The hub cells and the CySCs provide a microenvironment for GSC maintenance[4] . Both the germline and the somatic cyst stem cells divide asymmetrically: after division , one daughter remains in contact with the hub and self-renews while the other daughter is displaced away from the hub and initiates differentiation[5 , 6] . In the germline stem cell lineage , the differentiating daughter , termed the gonialblast , initiates four rounds of transit amplifying mitotic divisions with incomplete cytokinesis . The resulting 16 interconnected germ cells undergo premeiotic S phase in synchrony , become spermatocytes , and commit to terminal differentiation[7] . In the somatic cyst cell lineage , the differentiating daughter cell becomes a post-mitotic cyst cell , two of which enclose each gonialblast and its progeny , providing a supportive microenvironment necessary for the proper differentiation of the germ cells[8–10] . We identified a missense allele of DREF ( DREFts ) that revealed a role for DREF in the maintenance of Drosophila male GSCs . DREF is known to function in cell growth , cell division , and DNA replication[11–13] . However , its role in these housekeeping processes has masked identification of other biological functions of DREF . The protein encoded by DREFts is able to function in cell division and cell survival but is defective for maintenance of GSCs in the testes . Analysis of mutant germline stem cells demonstrated defects in downstream targets of TGF-beta signaling . Genetic interactions with this allele of DREF suggested that DREF functions antagonistically to the chromatin regulators Caf1/p55 and Mi-2 to maintain GSCs . We propose that DREF may promote expression of self-renewal genes by overcoming transcriptional repression by a Mi-2 containing chromatin-remodeling complex .
A temperature-sensitive allele of DREF ( DREFts ) was discovered in an EMS-mutagenesis screen to identify genes required cell-autonomously for GSC maintenance in the Drosophila male germline . When GSCs homozygous for DREFts and negatively marked for GFP were generated using the FLP/FRT system [14] and grown at 25°C for 3 days post clone induction , 84 . 2±3 . 0% of the testes scored ( n = 49 ) had at least one marked DREFts GSC adjacent to the hub . This is comparable to 95±3 . 0% of control testes ( n = 38 ) with marked GSCs wild-type for DREF adjacent to the hub . However , the percentage of testes with at least one marked homozygous DREFts mutant GSC steadily decreased over time , so that by day 12 post clone induction , only 16 . 7±4 . 7% of testes ( n = 91 ) contained one or more marked homozygous DREFts mutant GSCs , compared to 82 . 0±9 . 1% of testes ( n = 64 ) in control flies ( Fig 1A ) ( p<0 . 001 ) . Surprisingly , DREFts did not have the same effect on CySC maintenance . The percentage of testes with marked CySCs in DREFts/+ and wild-type controls were comparable at day 3 post clone induction ( 75 . 6% of DREFts/+ ( n = 38 ) to 78 . 3% ( n = 49 ) of control testes ) and also at day 12 post clone induction ( 39 . 2±14 . 5% of DREFts/+ ( n = 41 ) to 44 . 3±6 . 6% ( n = 62 ) of control testes ) ( Fig 1B ) . Complementation analysis indicated that the mutation responsible for the failure to maintain GSCs mapped to the DREF locus . The DREFts allele failed to complement Df ( 2L ) BSC17 [Df ( DREF ) , which deletes the DREF locus] , as well as two independently-generated alleles , DREFKG09294 ( referred to here as DREFnull ) and DREFNP4719 for early germ cell loss . Analysis of testes by phase contrast microscopy revealed that an average of 94 . 6± 5 . 6% testes ( n = 58 , p<0 . 0001 ) from newly eclosed DREFts/Df ( DREF ) males grown at 25°C had late-stage germ cells ( elongating spermatids ) but lacked early germ cells ( spermatogonia and primary spermatocytes ) , while 0% of testes from sibling DREF mutant/+ males lacked early germ cells ( Fig 1C and 1D ) . The presence of elongating spermatids in testes from newly eclosed DREFts/Df ( DREF ) males suggests that GSCs and early germ cells were initially present during development but were eventually lost due to differentiation . Immunofluorescence analysis of mutant testes confirmed the absence of GSCs and early germ cells; testes from newly eclosed DREFts/Df ( DREF ) males grown at 25°C lacked Vasa-positive germ cells at the apical tip ( Fig 1F ) , while testes from sibling controls had an abundance of Vasa-positive germ cells at the apical tip ( Fig 1G ) . DREF function is also required for maintenance of female GSCs/early germ cells . In ovaries isolated from DREFts/Df ( DREF ) mutant adult females grown at 25°C and examined three days after eclosion , 26 . 9% ( n = 26 ) of ovarioles contained developing germ cell cysts and egg chambers but empty germaria ( S1 Fig ) . The loss of early germ cells in DREFts/Df ( DREF ) males reared at 25°C was rescued by expression of a UAS-DREF cDNA transgene in early germ cells , confirming that the mutation causing GSC loss is in the DREF locus and that DREF function is required in germ cells for GSC maintenance . As assessed by phase contrast microscopy , only 5 . 37± 5 . 56% ( n = 58 ) of testes from DREFts/Df ( DREF ) males grown at 25°C contained early germ cells , while 100% ( n = 57 ) of testes from DREFts/Df ( DREF ) males expressing UAS-DREF cDNA under control of nanos-Gal4-VP16 ( Fig 1E ) and 93 . 3 ± 8 . 80% ( n = 59 ) of testes from DREFts/Df ( DREF ) males expressing UAS-DREF cDNA under control of Vasa-Gal4 ( Fig 1E and Table 1 ) contained early germ cells . Immunofluorescence analysis revealed the presence of GSCs adjacent to the hub in DREFts/Df ( DREF ) mutant testes expressing UAS-DREF cDNA under control of nanos-Gal4-VP16 ( Fig 1H ) . The GSC loss phenotype of DREFts/DREFnull flies was temperature-dependent . As assayed by phase contrast microscopy , 100 ± 0% ( n = 82 ) of testes isolated from DREFts/DREFnull transheterozygous flies grown at 22°C had early germ cells present on the day of eclosion . However , when DREFts/DREFnull flies were grown at 25°C , only 20 . 2 ± 12 . 9% ( n = 375 ) of testes scored contained early germ cells on the day of eclosion . DREFts/DREFnull flies grown at 30°C failed to survive to adulthood and died in late third instar larval and/or pupal stages ( Table 2 ) . Sequencing of the DREF coding region revealed that the DREFts allele had two amino acid substitutions ( Methionine 651 to Leucine and Glycine 652 to Alanine ) , which occur in a domain that has been shown to be responsible for co-factor binding in Drosophila DREF[15 , 16] . Germ cells homozygous mutant for DREFts still expressed DREF protein as assayed by immunostaining 3 days and 6 days post clone induction . In contrast , at day 3 post clone induction , DREF protein expression was not detected in DREFnull mutant germ cells by immunofluorescence using anti-DREF antibodies ( S2 Fig ) . Furthermore , GSCs in testes from DREFts/DREFnull versus DREFnull/+ flies grown at 22°C until eclosion then shifted to 30°C for 2 days failed to show significantly different levels of DREF protein ( 0 . 63±0 . 15 n = 18 GSCs versus 0 . 72±0 . 21 n = 24 GSCs , Materials and Methods ) . The DREFts allele allowed separation of the role of DREF in stem cell maintenance from a general role of DREF in cell survival and proliferation . Male GSCs homozygous mutant for the DREFts allele appeared to be lost because they differentiate more often than self-renew . While homozygous DREFts mutant GSCs were rapidly lost from the apical hub region of the testis at 25°C , their marked clonal progeny appeared to differentiate normally to at least the spermatocyte stage , as assessed by phase contrast microscopy . Cysts of spermatocyte clones homozygous for the DREFts allele examined on day 4 post clone induction contained 16 germ cells per spermatocyte cyst , as expected after 4 rounds of transit amplification divisions , and were comparable in size to neighboring wild-type spermatocytes ( Fig 2A , 2A’ and 2C ) . In contrast , loss of DREF function in germline clones homozygous for the DREFnull allele or due to expression in early gem cells of an RNAi hairpin directed against DREF mRNA under control of nanos-Gal4-VP16 resulted in defects in germ cell survival , growth , proliferation , and/or differentiation . At day 4 post induction of marked clones homozygous for DREFnull , 32 . 6% of the testes scored contained no DREFnull clones , 24 . 5% had marked GSCs and/or spermatogonia but no late spermatocyte cysts homozygous for DREFnull , and only 42 . 9% contained cysts with recognizable spermatocytes at 25°C ( Fig 2D ) . The cysts of DREFnull/DREFnull mutant spermatocytes in the clones appeared smaller in size compared to neighboring DREFnull/+ germ cell cysts ( Fig 2B and 2B’ ) . In addition , about 25% of the germ cell cysts that made it to the spermatocyte stage exhibited fewer than 16 cells per cyst ( Fig 2C ) . Together , these observations suggest that lack of DREF function may cause developing germ cell cysts to either die , grow slowly , or fail to initiate the differentiation program . Similarly , knockdown of DREF function in late spermatogonia and spermatocytes by expressing UAS-DREF-RNAi under the control of a Bam-Gal4 driver resulted in extensive cell death and absence of meiotic and post-meiotic stages ( S3 Fig ) . Consistent with the finding that homozygous DREFts germ cells proliferate and differentiate normally , cells homozygous for DREFts in the eye were able to proliferate and differentiate , unlike cells homozygous for DREFnull . Flies with eyes entirely composed of cells homozygous mutant for DREFnull generated using the EGUF-hid method [17] and reared at 25°C had very small eyes ( Fig 2G ) . In contrast , eyes entirely homozygous mutant for DREFts appeared similar in size and morphology to wild type controls , whether grown at 25°C ( Fig 2E ) or 30°C ( Fig 2F ) . Additionally , eyes in DREFts/ DREFnull transheterozygous flies grown at 25°C were wild type in size and appearance ( Fig 2H ) . The temperature-sensitivity of the DREFts allele allowed analysis of how GSCs are lost when shifted from permissive to non-permissive temperature . GSCs present when DREFts/DREFnull males were grown at 22°C until eclosion were lost within 2–3 days upon shifting the males to 30°C ( Fig 1I ) . On the day of eclosion , DREFts/DREFnull males grown at 22°C had an average of 6 . 6± 0 . 9 GSCs around the hub ( n = 22 testes ) , compared to 8 . 2± 0 . 6 GSCs in heterozygous control testes ( n = 27 testes ) . In contrast , testes from DREFts/DREFnull males grown at 22°C then shifted to 30°C for 3 days after eclosion had an average of 0 . 8± 0 . 7 GSCs ( n = 69 testes ) , while sibling control testes from DREF mutant/+ males had an average of 8 . 19±1 . 4 GSCs per testis ( n = 59 testes ) ( p<0 . 0001 ) ( Fig 1I ) . The loss of DREF function when DREFts/DREFnull males were shifted to 30°C did not appear to affect the rate of GSC division or survival . The percentage of GSCs in mitosis scored by immunostaining for phosphorylated-Threonine 3 of histone H3 ( PH3 ) was similar in testes isolated from DREFts/ DREFnull transheterozygotes ( 3 . 90 ± 0 . 65% , n = 154 GSCs ) and sibling control flies ( 4 . 49 ±2 . 84% , n = 178 GSCs ) that were grown at 22°C , shifted to 30°C at eclosion and then held at 30°C for two days ( Fig 3A ) . Additionally , TUNEL assays failed to detect dying GSCs in testes from either DREFts/ DREFnull transheterozygotes or control flies ( n = 83 and n = 94 GSCs , respectively ) , while in both cases some dying spermatogonial cysts were detected . The loss of GSCs in DREFts/ DREFnull testes did not appear to be due to loss of hub-GSC adhesion . Expression of UAS-E-Cadherin-GFP specifically in early germ cells using the nanos-Gal4-VP16 driver resulted in localization of E-Cadherin-GFP protein to the hub-GSC interface in both DREFts/ DREFnull ( Fig 3C–3C’’’ ) and DREFts/+ ( Fig 3D–3D’’’ ) GSCs in testes isolated from males grown at 22°C until eclosion then shifted to 30°C for two days . Expression of E-Cadherin-GFP using the UAS-GAL4 system in early germ cells did not rescue the loss of GSCs in DREFts/ DREFnull temperature-shifted flies , as testes from DREFts/ DREFnull males shifted to 30°C for two days contained similar numbers of GSCs per testis whether or not they expressed UAS-E-Cadherin-GFP under control of nanos-Gal4-VP16 ( average of 0 . 9±0 . 2 GSCs per testis ( n = 52 testes ) and 0 . 8±0 . 7 GSCs per testis ( n = 59 testes ) , respectively ) . DREFts/ DREFnull germ cells adjacent to the hub also oriented their centrosomes as in wild-type GSCs , where one centrosome is positioned adjacent to the hub-GSC interface throughout the cell cycle , resulting in oriented GSC division[5] . In testes isolated from DREFts/ DREFnull males grown at 22°C then shifted to 30°C for two days after eclosion , an average of 92 . 78 ± 3 . 76% ( n = 201 from 62 testes ) of GSCs that remained next to the hub and contained two centrosomes had one centrosome adjacent to the hub-GSC interface , similar to control DREFts/+ GSCs ( 94 . 1 ± 1 . 07% , n = 239 GSCs from 30 testes ) ( Fig 3B ) . Hub-GSC adhesion and oriented centrosome positioning are two features of GSCs that depend on activation of the JAK-STAT pathway in response to Unpaired ( Upd ) ligand secreted from the hub[10 , 18] . Consistent with intact hub-GSC attachment and correct positioning of centrosomes , GSCs in testes from DREFts/ DREFnull males grown at 22°C then shifted to 30°C for two days post-eclosion expressed STAT92E protein , a downstream target of JAK-STAT signaling , at levels comparable to sibling control GSCs ( Fig 3E–3E’ and 3F–3F’ ) . Many GSCs homozygous mutant for DREFts showed reduced expression of a reporter of TGF-beta signaling , Dad-LacZ , compared to their DREFts/+ neighbors next to the hub . In wildtype , Dad-LacZ is primarily expressed in GSCs , the gonialblast , and in later stage somatic cyst cells[19] . In many germ cells next to the hub made homozygous for DREFts by FLP induced mitotic recombination , LacZ staining was often drastically reduced compared to neighboring DREFts/+ GSCs ( Fig 4A–4A’’’ ) . The effect was variable , however , with some GSCs homozygous for DREFts appearing to have normal levels of LacZ , likely leading to the observed gradual loss of DREFts mutant GSCs . As a population , mutant GSCs showed a 30 . 3% reduction ( P<0 . 05 , n = 14 GSCs in 11 testes , Fig 4E ) in Dad-LacZ staining intensity relative to control GSCs ( n = 17 GSCs in 11 testes ) within the same testes as quantified by ImageJ ( Materials and Methods ) . Consistent with a reduced response to TGF-beta signaling , the differentiation marker Bag of Marbles ( Bam ) was expressed earlier than normal in germ cells in DREFts/DREFnull testes . Bam protein is normally detected by antibody staining in 4- to 16-cell spermatogonial cysts[20] . Likewise , using a reporter line driving GFP-tagged Bam protein expressed from its own promoter and regulatory elements[21] , Bam-GFP expression , as detected by anti-GFP immunostaining , was first detected in 4 cell cysts , with very few 2 cell cysts scoring positive . GSCs from DREFts/ DREFnull males grown at 22°C until eclosion then shifted to 30°C for two days did not express Bam-GFP in GSCs adjacent to the hub ( n = 36 GSCs from 14 testes ) , similar to GSCs from sibling controls that were either DREFts/+ or DREFnull /+ ( n = 100 GSCs from 12 testes ) ( Fig 4B–4B’’’ ) . Likewise , gonialblasts , defined as single germ cells away from the hub containing a dot-fusome , from DREFts/DREFnull temperature shifted males also did not express Bam-GFP ( n = 23 gonialblasts from 14 testes ) , similar to the gonialblasts in heterozygous controls , in which Bam-GFP was not detected ( n = 31 gonialblasts from 12 testes ) . However , there was a marked increase in the percentage of Bam-GFP positive two-cell cysts in DREFts/DREFnull mutant testes , where 35 . 1% of the two-cell cysts analyzed ( n = 37 counted in 14 testes ) were positive for Bam-GFP , compared with only 4 . 6% of two-cell cysts scored as positive for Bam-GFP in testes from DREFts/+ or DREFnull /+ sibling controls ( n = 44 cysts counted in 12 testes ) ( Fig 4C’–4C’’’ and 4D ) . Function of DREF appeared to be required for maintenance of germline stem cell state even under the condition of forced ectopic expression of the ligand Upd , which activates the JAK-STAT signal transduction pathway . In control DREFts/+ or Df ( DREF ) /+ flies , ectopic expression of UAS-Upd in early germ cells under the control of nanos-Gal4-VP16 at 25°C resulted in 100% ( n = 68 testes ) of testes examined containing an overabundance of GSC-like Vasa-positive cells with dot spectrosomes , as well as many CySC-like cells positive for Zfh-1 ( Fig 5A–5A’’’ ) . Under the same conditions , in contrast , only 4 . 98±5 . 64% ( n = 66 testes ) of DREFts/Df ( DREF ) testes ectopically expressing UAS-Upd under control of nanos-Gal4-VP16 exhibited an overabundance of GSC and CySC-like cells . Rather , the vast majority of testes ( 95 . 02% , n = 66 ) from DREFts/Df ( DREF ) males carrying nanos-Gal4-VP16; UAS-Upd grown at 25°C had an abundance of CySC-like cells but few or no germ cells as assayed by immunostaining for Vasa ( Fig 5B–5B’’’ ) . The abundance of Zfh-1 positive CySC-like cells in many of the DREFts/Df ( DREF ) testes suggested that sufficient Upd was expressed early , prior to the GSC loss ( Fig 5C–5C’’’ ) . The DREFts allele provided a sensitized background in which to screen for genetic interactors important for male GSC differentiation vs . maintenance . When DREFts/ DREFnull flies were raised at 25°C , 79 . 8±12 . 9% of testes ( n = 375 ) from newly eclosed males had few or no early germ cells and displayed elongating spermatid bundles close to the apical tip ( Fig 6A ) . Strikingly , reducing the gene dosage of either Chromatin assembly factor 1 , p55 subunit ( Caf-1/p55 ) , a subunit of multiple chromatin-modifying complexes including the NuRD complex , or Mi-2 , a subunit of the Nucleosome Remodeling and Deacetylase ( NuRD ) complex and the dMEC complex , rescued the DREF early germ cell loss phenotype . While only 20 . 2±12 . 9% ( n = 375 ) of testes from newly eclosed DREFts/DREFnull males had early germ cells ( Fig 6A ) , 93 . 9±5 . 35% ( n = 62 ) of testes from DREFts/DREFnull; Caf19-2/+ males contained both plentiful early germ cells and abundant spermatocytes as assessed by phase contrast microscopy ( Fig 6B ) . Similarly , 91 . 8±7 . 5% ( n = 76 ) of testes from DREFts/DREFnull; Df ( Mi-2 ) /+ and 55 . 6±4 . 0% of testes from DREFts/DREFnull; Mi-24/+ males contained many early germ cells and spermatocytes ( Fig 6C ) . Immunofluorescence analysis revealed that reducing Caf1 or Mi-2 dosage by half in a DREFts/DREFnull background restored the presence of GSCs next to the hub ( Fig 6D–6F ) . The suppression of the DREFts/DREFnull GSC loss phenotype by lowering the dosage of either Mi-2 or Caf1 , both of which are components of the NuRD complex , suggests that DREF and the NuRD complex may act antagonistically to influence GSC maintenance . However , lowering the dose of Rpd3 or Mbd-like , two other subunits of the NuRD complex , did not suppress the early germ cell loss phenotype in DREF mutant testes as assessed by phase contrast microscopy ( Fig 6G ) . Similarly , lowering the dose of XNP , a member of the DREF-containing XNP/dATRX repression complex , or Putzig , a member of the DREF-TRF2 complex , did not affect the germ cell loss phenotype , with only 29 . 1±10 . 9% ( n = 74 ) and 23 . 3±10 . 4% ( n = 41 ) , respectively , of testes scored by phase contrast microscopy containing visible early germ cells ( Fig 6G ) . TRF2 , which is male lethal , could not be tested in the same manner . Mi-2 is required cell-autonomously for GSC maintenance in the testis . GSCs made homozygous mutant for either the Mi-24 ( frameshift ) or the Mi-26 ( premature stop codon ) allele using the Flp-FRT system were lost over time . At day 3 post clone induction , 65 . 9±16 . 1% of testes scored ( n = 33 ) for Mi-24 and 70 . 7±11 . 5% of testes scored ( n = 66 ) for Mi-26 contained marked homozygous mutant GSCs , compared to control clones , for which 82 . 5±5 . 9% of testes scored ( n = 55 ) contained marked GSCs . By day 8 post clone induction , however , only 14 . 3±15 . 2% ( n = 44 ) for Mi-24 and 17 . 2±7 . 9% ( n = 70 ) for Mi-26 of testes scored contained marked homozygous mutant GSCs , while the percentage of control testes with marked GSCs was significantly higher ( 72 . 3±9 . 1% , n = 76 , p<0 . 001 ) ( Fig 7A ) . Germ cells homozygous mutant for Mi-2 were able to differentiate into spermatocytes . Expression of RNAi directed against Mi-2 using the nanos-Gal4-VP16 driver also resulted in early germ cell loss , confirming a role for Mi-2 in GSC maintenance ( Fig 7B and 7C ) . Similarly , RNAi targeting Caf1 expressed using the nanos-Gal4-VP16 driver also resulted in GSC loss , demonstrating a requirement for Caf1 for GSC maintenance ( S4 Fig ) . Homozygous Mi-24 or Mi-26 mutant GSCs remaining at day 5 post clone induction expressed DREF protein at levels comparable to neighboring Mi-2/+ heterozygous GSCs , as detected by immunostaining ( Fig 7D ) . However , later stage germ cells lacking Mi-2 function demonstrated prolonged DREF expression , visible as persistent high levels of DREF protein in spermatocyte cysts in germline clones homozygous mutant for Mi-2 , compared to neighboring Mi-2/+ spermatocytes ( Fig 7E–7E’’’ ) .
Drosophila DREF was initially isolated based on its ability to bind the DNA-Replication Related Element ( DRE ) , an 8bp sequence 5’TATCGATA’3 located upstream of many genes related to DNA replication . Binding of DREF protein to the DRE has been shown to activate expression of genes regulating cell division , including DNA polymerase , E2F , and Cyclin A[22–24] . In addition , Drosophila DREF has been shown to act downstream of the TOR [13] signaling pathway and Drosophila DREF and its human homolog , hDREF , have also been shown to control cell growth by regulating the expression of ribosomal genes and histone H1[13 , 25 , 26] . The DRE bound by DREF is known to be a key cis-regulatory component of a class of core promoters different from the canonical TATA box containing promoters[27] . Binding of DREF protein to the DRE recruits TRF2 , a transcription factor related to TATA-box-binding protein ( TBP ) , directing recognition of these alternate core promoters , regulating , for example expression of the proliferating cell nuclear antigen ( PCNA ) [28] . DREF Protein binds to and potentially regulates 1 , 961 distinct loci in the genome[29] . For example , DREF has been shown to regulate chromatin by: 1 ) by activating transcriptional expression of chromatin regulators such as brahma , moira and osa[22 , 30] , Mes4[31] , and HP6[32] , 2 ) physically interacting with XNP/dATRX and potentially targeting them to regions in the genome[16] , 3 ) competing with the chromatin insulator Boundary Element Association Factor 32 ( BEAF32 ) for a mutual binding site[29 , 33] , and 4 ) regulating the HET-A , TART , TAHRE array ( HTT ) array in Drosophila[34] . Previous reports , and our work here , showed that null alleles of DREF have defects in cell division and cell growth in tissues ranging from the eye[12 , 22] , salivary glands[12] , the imaginal wing discs[35] , and the testis ( this report ) . The DREFts allele we identified a role for DREF in self-renewal of germ line stem cells , genetically separable from the previously-defined role of DREF in cell proliferation . The DREFts allele did not strongly affect cell viability and division in the eye , and somatic cyst stem cells ( CySCs ) homozygous mutant for the DREFts allele were able to self-renew and differentiate normally . However , in flies homozygous mutant for the DREFts allele , GSCs were not maintained , although the progeny of mutant GSCs were able to differentiate into spermatocytes . It is possible that the DREFts allele might affect specific physical interactions between DREF and particular binding partner ( s ) required in GSCs but not in many other cell types , resulting in stem cell loss . Intriguingly , previous research suggests that the domain mutated in DREFts ( the CR3 domain ) does not contribute to the DNA binding or dimerization functions of DREF , but may play a role in the ability of DREF to bind and interact with different cofactors[36] . Overexpression of the CR3 domain of DREF has been shown to have a dominant-negative effect on DREF function , possibly by competing for normal binding partners of endogenous DREF[36] . In support of this view , DREF has been shown to bind to the chromatin remodeling factor XNP through its CR3 domain[16] . When bound to XNP , DREF can function as a transcriptional repressor , in contrast to its typical role as a transcriptional activator when bound to TRF2[16] . However , XNP showed no genetic interaction with the DREFts/DREFnull mutant phenotype , suggesting that either XNP is not dosage sensitive or that the role of DREF in GSC maintenance is not mediated through the DREF-XNP complex . Although an RNAi screen by Yan et al . [37] showed a requirement for DREF in the female germline , it is important to note that in our hands complete loss of function of DREF ( through either RNAi or null alleles ) causes severe phenotypes likely due to the role of DREF in housekeeping functions required for cell growth and division , which parallels the multiple , complex defects Yan et al . [37] notes in knockdown of DREF in the female germline . The JAK-STAT signaling pathway plays pivotal roles in regulating the two adult stem cell populations in the testis . In male GSCs , the JAK-STAT signal transduction pathway is required cell autonomously for adhesion to the hub and oriented divisions , but not for self-renewal[18] . Many mutants that result in GSC loss , such as NURF301 [38] have reduced JAK-STAT signaling , possibly resulting in loss of GSC adhesion to the hub and subsequent differentiation . In contrast , DREFts/DREFnull GSCs had normal levels of STAT protein and did not appear to be defective in hub-GSC adhesion , as evidenced by localization of E-Cadherin and proper centrosome orientation , suggesting that they are not likely lost due to defects in JAK-STAT signaling . JAK-STAT activation is also required in CySCs for self-renewal[18 , 39] . Forced activation of JAK-STAT signaling in the testes , either by expressing constitutively active JAK in CySCs or by forced ectopic expression of the activating ligand Upd in germ cells , results in an apparent failure of CySCs to differentiate[18] . As a consequence of the early CySC-like state , the neighboring germ cells fail to differentiate and the testis is filled with GSC-like and CySC-like cells . This overproliferation of GSC-like cells due to forced activation of the JAK-STAT pathway can mask or override the GSC-loss phenotype in his2Av or GEF26 mutants[40 , 41] . Although , like his2Av and GEF26 mutant GSCs , DREFts GSCs are lost to differentiation , the outcome of combining DREFts with forced expression of Upd ligand in germ cells was strikingly different , with the DREFts/Df ( DREF ) germ line stem cell loss phenotype predominating even with forced activation of JAK-STAT signaling due to ectopic expression of Upd . Thus , while his2av and GEF26 may be important for fine-tuning the balance between self-renewal and differentiation , the function of DREF altered by the temperature-sensitive mutation may be intrinsically required for maintaining the GSC state . Genetic interaction studies uncovered a novel role for Caf1 and Mi-2 , components of the Nucleosome Remodeling and Deacetylase ( NuRD ) chromatin-modifying complex in repressing DREF-mediated self-renewal . Reducing the gene dosage of either Caf1 or Mi-2 function by half was able to rescue the GSC-loss phenotype in a DREFts/DREFnull mutant background . Our results are consistent with previous studies indicating an antagonistic relationship between DREF and Mi-2 . Reduction of Mi-2 gene dosage by half had been shown previously to enhance defects caused by DREF overexpression in the eye , consistent with Mi-2 antagonizing DREF[42] . Yeast two-hybrid screening identified the human homolog of Mi-2 , CHD4 , as a binding partner of human DREF and pull-down assays confirmed this association in Drosophila , showing that Mi-2 physically associates with DREF[42] . Mi-2 has been shown to interact with the DNA-binding domain of DREF , thereby inhibiting the ability of DREF to bind DNA in vitro[42] . Additionally , recent work with hDREF has shown that the reciprocal regulation as well: hDREF has been shown to increase SUMOylation of Mi-2 protein , thereby increasing dissociation of Mi-2 from chromatin[43] . Our data also indicated an inhibitory interaction between DREF and CAF1 , raising the possibility that Mi-2 might act as a part of the NuRD complex to inhibit DREF function in male germline stem cells . The genetic interactions between Mi-2 or Caf1 and the DREFts allele suggest that Mi-2 and Caf1 can act as repressors of GSC self-renewal . Interestingly , we also found that Mi-2 and Caf1 are required for GSC maintenance in a genetic background wild-type for DREF function . The Mi-2/NuRD complex is known to play broad roles in reorganizing chromatin architecture to promote silencing[44] , and Mi-2/NuRD has been shown to localize to hundreds of regions in the genome of Drosophila[42] . One possibility is that complete loss of Mi-2 function may cause general de-repression of genes important for cell-identity , self-renewal , and differentiation . Indeed , loss of Mi-2 in spermatocytes causes activation cryptic promoters at many sites in the genome , leading to massive misexpression[45] . It may be that Mi-2 plays a similar role in restraining misexpression during the early stages of germ cell development as well . Previous studies suggest that Drosophila male GSCs are sensitive to changes in the transcript levels of genes required for self-renewal versus differentiation[41] . One role of DREF in Drosophila male GSCs may be to exclude Mi-2 from promoters of self-renewal genes , thereby allowing higher levels of expression at these loci . Under conditions of reduced DREF function in the DREFts mutant flies , Mi-2 may gain abnormal access to self-renewal genes , dialing down their expression and tilting the balance of GSC fate towards differentiation . We propose that in the context of DREFts/DREFnull background , a half-dose reduction of Mi-2 or CAF1 function is sufficient to allow the partially functional DREF protein expressed in the DREFts mutant to overcome Mi-2-mediated repression of self-renewal genes and tilt the balance back towards GSC maintenance .
All crosses were grown at 25°C on standard molasses media unless otherwise stated . DREF mutant alleles used in this study include 1 ) al1 , dpov1 , DREFts , b1 , pr1 , FRT40A/SM6a ( derived from an EMS mutagenesis screen ) , 2 ) DREFKG09294 ( from BDSC ) ; this allele is a P-element insertion into the 5’UTR of DREF and has been previously reported to be a null allele that expresses little to no DREF protein[13] , 3 ) DREFKG09294 , FRT40A ( from DGRC ) , 4 ) Df ( 2L ) BSC17 ( referred to as Df ( DREF ) , from BDSC ) is a deletion that spans from pelota to DREF , and 5 ) DREFNP4719 ( from DGRC ) , a P-element insertion in the 5’ UTR of DREF . Other mutant alleles used in this study include 1 ) Caf19-2 , a deletion allele of Caf1 ( gift from Joseph Lipsick , Stanford University[46] ) , 2 ) Df ( 3R ) BSC471 ( referred to as Df ( Caf1 ) , from BDSC ) , 3 ) Mi-24 and Mi-26 alleles ( from BDSC and gift from J . Müller[47] ) , 4 ) Df ( 3L ) BSC445 ( referred to as Df ( Mi-2 ) , from BDSC ) , 5 ) Df ( 3R ) BSC471 ( referred to as Df ( Putzig ) , from BDSC ) , 6 ) Df ( 3R ) XNP1 ( referred to as Df ( XNP1 ) , from BDSC ) , 7 ) Df ( 3L ) Exel7208 ( referred to as Df ( Rpd3 ) , from BDSC ) , 8 ) Df ( 3R ) Exel6153 ( referred to as Df ( Mbdl ) , from BDSC ) and 9 ) Nurf3014 , a deletion allele of Nurf301 ( from BDSC ) . Other fly stocks used include 1 ) al1 , dpov1 , b1 , pr1 , FRT40A ( an isogenized version of FRT40A from BDSC ) , 2 ) yw , hs-flp122; FRT40A , Ubi-GFP , 3 ) eyeless-Gal4 , UAS-Flp; GMR-hid , 2LCL FRT40A ( 17 ) , 3 ) UAS-DEFL #6–1 [48] 4 ) bam::bam-GFP , a transgenic line driving Bam-GFP under the control of bam promoter ( a gift from D . McKearin ( 21 ) ) 5 ) UAS-Upd [49] , and 6 ) UAS-DREF[22] . al , dp , DREFts , b , pr , FRT40A , UAS-DREF/SM6a and DREFts , UAS-Upd were generated by recombining DREFts chromosome onto the UAS-DREF or UAS-Upd chromosome , respectively . The following Gal4 drivers were used to drive UAS transgress in a cell-type specific manner , 1 ) yw;;nanos-Gal4VP16 ( a gift from R . Lehmann[50] ) , 2 ) UAS-dicer2;;nanos-Gal4VP16 , 3 ) yw;;Vasa-Gal4 , 4 ) yw;;Bam-Gal4 , UAS-dicer2 , ( Bam-Gal4 lines was a gift from D . McKearin[21] ) and 5 ) C587-Gal4;tub-Gal80ts;UAS-dicer2 ( C587-Gal4 line was a gift from S . Hou ) . RNAi lines for DREF ( BDSC#35962 ) , Caf-1 ( BDSC#34069 and VDRC#105838 ) and Mi-2 ( VDRC#107204 and 10766 ) were used in this study . Temperature shift experiments for germ cell loss analysis in DREF transheterozygotes were performed by growing flies at 22°C until eclosion , and then shifting to 30°C on the day of eclosion . Dissections were performed on day of eclosion , or on one , two , or three days post eclosion . For analysis of DREF function in the eye , clones were generated by crossing DREF alleles on FRT40A chromosomes to eyeless-Gal4 , UAS-Flp; GMR-hid , 2LCL FRT40A/SM6a . These crosses were grown at either 25°C , or at 30°C one day after the cross was set up . Testes were dissected in 1X phosphate-buffered saline ( PBS ) and fixed in 4% paraformaldehyde diluted in PBS for 20 minutes at room temperature , washed twice in PBS with 0 . 1% TritonX-100 , permeabilized in PBS with 0 . 3% TritonX-100 and 0 . 6% sodium deoxycholate for 30 minutes and blocked in PBS with 0 . 1% TritonX-100 and 3% bovine serum albumin for 30 minutes . Testes were incubated overnight at 4°C in primary antibodies against DREF ( mouse , 1:100; gift from Dr . Andreas Hochheimer[28] ) , Vasa ( goat , 1:50; Santa Cruz Biotechnology ) , Zfh-1 ( rabbit , 1:5000; gift from R . Lehman ) , Traffic Jam ( guinea pig , 1:5000; gift from D . Godt ) LI ET AL NAT CELL BIOL 2003 , Armadillo ( mouse , 1:10; Developmental Studies Hybridoma Bank ( DSHB ) ) , DE-Cadherin ( rat , 1:10; DSHB ) , FasIII ( mouse , 1:10; DSHB ) , alpha-spectrin ( mouse , 1:10; DSHB ) , Green Fluorescent protein ( rabbit 1:400–1:1000; Invitrogen and Sheep 1:1000; Abd-Serotec ) , gamma-tubulin ( mouse , 1:50; Sigma ) , phosphor-Histone3 Threonine3 ( rabbit , 1:100; Upstate Biotechnology/Millipore ) , and STAT92E ( rabbit , 1:100; gift from E . Bach[51] ) . Testes were incubated in appropriate secondary antibodies were from the Alexa Fluor-conjugated series ( 1:500; Molecular Probes ) and mounted in VECTASHIELD medium containing DAPI to visualize DNA ( Vector Labs ) . Tunel Assays were performed using the In Situ Cell Death Detection Kit , TMR red by Roche . Immunofluorescence images were taken using the Leica SP2 Confocal Laser scanning microscope . Phase and clonal analysis images were captured using a Zeiss Axioskop microscope and SPOT RT3 camera by Diagnostic Instruments or CoolSNAePz camera by Phomometrics . Images were processed using Adobe Photoshop and Illustrator CS6 . Nuclear protein quantification was performed using ImageJ software and comparing the relative levels of the stained protein of interest ( standardized to DAPI to control for sample depth ) between experimental and control GSCs- . To perform RNAi knockdown in the germline , males from strains containing the RNAi hairpin of interest were crossed to UAS-dicer2;; nanos-Gal4VP16 virgins and grown at 25°C for 4 days after which the progeny of the cross were shifted to 30°C . Testes were isolated from males of the cross on the day of eclosion and 7 days post eclosion . In other cases , to follow the time course of stem cell loss , the progeny of crosses were grown at 18°C until eclosion , shifted to 30°C on the day of eclosion and males from the cross were dissected at different days post-shift to 30°C . Somatic RNAi knockdown was performed by crossing the RNAi hairpin strains to virgin females containing the somatic lineage driver C587-Gal4;tub-Gal80ts;UAS-dicer2 . Crosses were performed at 18°C until eclosion , when adults were shifted to 30°C and dissected 3- and 7-days post temperature shift . Knockdown in transit amplifying cells and later stages was performed by crossing the RNAi hairpin to;; BamGal4 , UAS-dicer2 . Crosses were grown at 25°C for 4 days and then shifted to 29°C until eclosion and then they were dissected . | Many adult tissues are maintained throughout life by the dual ability of adult stem cells to produce progeny that either self-renew or differentiate to replace specialized cells lost to turnover or damage . Although signals from the surrounding microenvironment have been shown to regulate the choice between self-renewal and onset of differentiation , the intrinsic gene regulatory programs that set up and maintain this bipotential state are not well understood . In this report we describe antagonistic components of an intrinsic stem cell program important for maintaining the balance between self-renewal and differentiation in Drosophila male germline adult stem cell lineage . We identified a temperature-sensitive mutant in the transcription factor DNA Replication-related Element Factor ( DREF ) gene that disrupts the ability of germline stem cells to self-renew , but not stem cell viability , ability to divide or differentiate under the same conditions . DREF mutant germline stem cells showed defects in the TGF-beta signaling pathway , a pathway that is critical for maintaining the stem cell population . Genetic interaction analyses revealed that Mi-2 and Caf1/p55 , components of the Nucleosome Remodeling and Deacetylase complex genetically antagonize the role of DREF in germline stem cell maintenance . We propose that DREF contributes to a transcriptional environment necessary for maintaining a bi-potential stem cell state able to properly respond to extrinsic niche signals . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] | [
"medicine",
"and",
"health",
"sciences",
"reproductive",
"system",
"spermatocytes",
"rna",
"interference",
"cloning",
"cell",
"differentiation",
"germ",
"cells",
"developmental",
"biology",
"stem",
"cells",
"molecular",
"biology",
"techniques",
"epigenetics",
"sperm",
"research",
"and",
"analysis",
"methods",
"adult",
"stem",
"cells",
"animal",
"cells",
"genetic",
"interference",
"gene",
"expression",
"molecular",
"biology",
"testes",
"biochemistry",
"rna",
"anatomy",
"cell",
"biology",
"nucleic",
"acids",
"genetics",
"biology",
"and",
"life",
"sciences",
"cellular",
"types",
"genital",
"anatomy"
] | 2019 | DREF Genetically Counteracts Mi-2 and Caf1 to Regulate Adult Stem Cell Maintenance |
Calcium through NMDA receptors ( NMDARs ) is necessary for the long-term potentiation ( LTP ) of synaptic strength; however , NMDARs differ in several properties that can influence the amount of calcium influx into the spine . These properties , such as sensitivity to magnesium block and conductance decay kinetics , change the receptor's response to spike timing dependent plasticity ( STDP ) protocols , and thereby shape synaptic integration and information processing . This study investigates the role of GluN2 subunit differences on spine calcium concentration during several STDP protocols in a model of a striatal medium spiny projection neuron ( MSPN ) . The multi-compartment , multi-channel model exhibits firing frequency , spike width , and latency to first spike similar to current clamp data from mouse dorsal striatum MSPN . We find that NMDAR-mediated calcium is dependent on GluN2 subunit type , action potential timing , duration of somatic depolarization , and number of action potentials . Furthermore , the model demonstrates that in MSPNs , GluN2A and GluN2B control which STDP intervals allow for substantial calcium elevation in spines . The model predicts that blocking GluN2B subunits would modulate the range of intervals that cause long term potentiation . We confirmed this prediction experimentally , demonstrating that blocking GluN2B in the striatum , narrows the range of STDP intervals that cause long term potentiation . This ability of the GluN2 subunit to modulate the shape of the STDP curve could underlie the role that GluN2 subunits play in learning and development .
The striatum is the main input structure of the basal ganglia , which is necessary for proper motor function and habit formation . The medium spiny projection neurons ( MSPNs ) , which comprise ∼95% of striatal neurons , undergo changes in synaptic strength during the learning of a motor task [1] . This synaptic plasticity is thought to be the cellular basis of motor learning and habit formation , and it is disrupted in animal models of Parkinson's Disease [2] and Huntington's Disease [3] . One of the critical mechanisms for inducing synaptic plasticity in neurons is calcium elevation in the spine . The sources of calcium are quite diverse , and depend on brain region and direction of plasticity . In particular , LTD often requires release of calcium from internal stores [4] or voltage dependent calcium channels [4] , [5] . In contrast , the source of spine calcium that contributes to long-term potentiation ( LTP ) is the NMDA receptor ( NMDAR ) in the hippocampus [6] , cortex [7] , and striatum [8] . Because NMDARs permit calcium influx in response to the coincidence of pre-synaptic glutamate release and post-synaptic depolarization , they are well situated to modulate spike timing dependent plasticity ( STDP ) . In STDP protocols , an action potential ( AP ) is caused by depolarizing the soma of a neuron and is paired in time with a pre-synaptic stimulation . However , NMDARs differ in several properties that may be critical for timing-dependent synaptic plasticity . They contain various combinations of GluN1 , 2 , and 3 subunits which can change their maximal conductance , current decay time , and sensitivity to magnesium block [9] . While the GluN1 splice variant has some control over the kinetic properties of the NMDAR , the four GluN2 subunits ( A , B , C , and D ) strongly control them when the GluN1 splice variant is kept the same [9] . The GluN2 subunit can thereby alter the calcium influx through the NMDAR . Because the specific differences between GluN2 subunits are the ones that would affect the NMDARs dependence on AP timing , and because calcium through the NMDAR plays an essential role in striatal timing-dependent long term potentiation ( tLTP ) [10]–[12] we hypothesized that changes in GluN2 subunit would modulate STDP in the striatum . The MSPNs of the striatum contain both GluN2A and GluN2B subunits in abundance [13] , and it has been suggested that GluN2D subunits may be present in low concentrations [14] . In animal models of Parkinson's disease , the NMDAR subunit composition is altered in the striatum [2] and subunit-specific NMDAR antagonists have been shown to alleviate Parkinson's like symptoms [15] . However , the intracellular consequences of such altered NMDAR subunit composition has not yet been made clear . In this study , we investigate the effects of altering NMDAR subunit composition on tLTP in the striatum . Using a multi-compartmental model of a MSPN , we examine NMDAR-mediated calcium influx through receptors containing different GluN2 subunits and under different STDP conditions . We find that calcium elevation depends on which GluN2 subunit the NMDAR contains , the relative timing of the AP , the duration of somatic depolarization , and the number of consecutive APs . More significantly , model predictions about the effect of GluN2 subunit on the shape of the STDP curve are confirmed experimentally .
To evaluate the role of the GluN2 subunit in STDP , we used a model of a dorsal striatal MSPN , which was modified from a nucleus accumbens neuron model [16] , [17] . The model of the dorsal striatum MSPN had the same general morphology as the nucleus accumbens model ( Figure 1A ) , with explicit spines that include synaptic glutamate receptors ( AMPA and NMDA ) and voltage dependent calcium channels . Kinetics and maximal conductances of voltage dependent channels ( See Tables 1 and 2 ) were tuned to match the characteristics of neurons in the dorsal striatum , such as smallest current that evokes a single AP , long latency to first spike during somatic depolarization ( Figure 1B ) , AP width of ∼1 ms , and distinct inward rectification ( Figures 1C and see Text S1 ) . Once the model was tuned to match these parameters , it also closely matched the average current-voltage relationship of 25 MSPNs from the mouse dorsal striatum undergoing a series of hyperpolarizing and depolarizing currents incrementing by 50 pA from −500 pA to 200 pA ( Figure 1D and see Text S1 ) . This model did not contain GABAergic synapses and represents experimental conditions in which GABAergic channels are blocked [10] , [11] , [18] . Because NMDAR channel opening requires both glutamate to activate the channel and membrane depolarization to relieve the voltage-dependent magnesium block , the occurrence of a somatic AP close in time to glutamatergic stimulation can drastically increase the calcium elevation due to the NMDAR . Previous studies have shown that different somatic depolarization patterns back-propagate into the dendrites with varying strength [19]–[21] . Consequently , STDP protocols that have been used on striatal cells vary in AP number and length of somatic depolarization . To investigate the influence that different protocols have on the NMDAR-mediated calcium elevation in the spine , simulations were performed for each STDP protocol that has been used in the striatum: a 30 ms depolarization to 1 AP [10] , [22] , a 5 ms depolarization to 1 AP [11] , and an AP triplet at a frequency of 50 Hz [11] , [18] ( Figure 2A ) . Each simulation involved the stimulation of two adjacent spines on the secondary dendrite of the model cell and a somatic depolarization to AP initiation , with the two events separated by a specific temporal interval ( Δt ) . For each STDP protocol , the peak calcium elevation due to the NMDAR in a single spine was recorded for both positive and negative Δt and was compared to the peak NMDAR-mediated calcium elevation due to control stimulation with no somatic AP ( Figure 2B ) . When repeated over a range of Δt , simulations reveal that a 30 ms depolarization to 1 AP significantly increased the NMDAR-mediated calcium in the spine when it occurred +2 to +20 ms after pre-synaptic stimulation ( ∼250% increase ) , while a 5 ms depolarization to AP caused much reduced elevations in NMDAR-mediated calcium at these intervals ( ∼175% increase ) . The NMDAR-mediated calcium elevation observed with negative Δt was reduced compared to positive Δt for all three protocols , consistent with the experimental observation that in the presence of GABA inhibitors , negative Δt does not produce tLTP [10] . Because in reality , the AP does not always occur at the exact same time within the depolarization , STDP simulations were repeated with added random synaptic input to cause noise and jitter to the timing of the AP within the 30 ms depolarization . Simulations were repeated 6 times using different input frequencies and random seed values . The averages of these simulations overlay the control conditions , demonstrating that this model and protocol are robust to noise ( Figure S1 ) . Because calcium through the NMDAR is essential for tLTP [10] , [12] , the model predicts that the 30 ms depolarization protocol , which causes a strong increase in NMDAR-mediated calcium , will more readily induce tLTP than the 5 ms depolarization protocol which causes a weak increase in NMDAR-mediated calcium . This model prediction was confirmed experimentally with whole-cell patch-clamp recordings in rat striatal slices . STDP experiments show that using Δt of +5 to +30 ms , robust tLTP is induced with the 30 ms depolarization protocol , but tLTP is not induced with the 5 ms depolarization protocol ( Figures 2D &E ) . Indeed , with the 30 ms depolarization protocol , the mean EPSC amplitude value , recorded 60 minutes after the induction protocol , was 186 . 0±11 . 6% , ( Δt = +18 . 6±1 . 5 ms , n = 22 ) , a value significantly different ( p<0 . 01 ) from baseline ( Figure 2E ) . However , 5 ms postsynaptic suprathreshold depolarizations paired with presynaptic stimulations ( Δt = +10 . 3±0 . 4 ms , n = 5 ) were not able to induce significant plasticity . The mean EPSC amplitude recorded 60 minutes after the induction protocol ( 97 . 5±11 . 0% ) did not significantly differ from baseline ( Figure 2E ) . The type of GluN2 subunit ( A , B , C , and D ) strongly determines the maximal conductance , decay time , and sensitivity to magnesium block for the NMDAR [9] . Therefore , differences in GluN2 subunit are predicted to modulate synaptic plasticity by controlling the calcium influx through the NMDAR . To evaluate this hypothesis , simulations were repeated using the 30 ms depolarization STDP protocol that caused tLTP for four types of NMDAR each representing a receptor containing two GluN1 subunits and two GluN2 subunits of the same type ( either A , B , C , or D ) . The current decay time , maximal conductance , and magnesium block parameters of these NMDARs were adjusted to reflect experimental measurements [23] , [24] ( Table 3 , Figures S2A & B ) . Calcium elevation due to NMDAR was evaluated for each GluN2 subunit condition over all Δt ( from −100 to +100 ms ) . The resulting calcium curves ( Figure 3 ) show clear differences in shape depending on GluN2 subunit . GluN2A and GluN2B conditions show the highest normalized calcium peaks , reaching greater than 250% of control , while GluN2C demonstrates a much reduced increase in calcium due to AP timing , just reaching 200% at the narrowest intervals . Similarly , GluN2D exhibits reduced calcium peaks and shows very little change in calcium concentration based on the AP timing ( Figure 3 ) . These simulations suggest that tLTP would be readily induced in GluN2A and GluN2B-containing synapses , but would not be as easily induced in GluN2C and GluN2D-containing synapses . Although GluN2A and GluN2B-containing receptors both reach calcium peaks greater than 250% , the shapes of the calcium curves are strikingly different . Because of the faster decay time of GluN2A , the range of Δt that causes strong calcium elevation is much narrower than for the more slowly decaying GluN2B-containing receptors . Calcium elevation due to influx through GluN2A only reaches levels >250% for narrow Δt . However , the GluN2B-containing receptors maintain elevations in calcium >250% even at wide Δt ( Figure 3 ) . Because the predicted intervals that allow significant amounts of calcium through NMDARs differ between GluN2A and GluN2B and because striatal MSPNs contain both GluN2B and GluN2A subunits [25] , the model predicts that blocking the GluN2B subunit in an MSPN would narrow the range of Δt that produces tLTP . To test this prediction , STDP experiments were conducted in either normal aCSF , or in the presence of 10 µM Ifenprodil , a potent and selective GluN2B antagonist . In the control condition , Δt between +5 ms and +30 ms elicits tLTP ( 186 . 0±11 . 6% , n = 22 ) ( Figure 4A–D ) . This Δt range includes both narrow timing intervals ( 5 ms<Δt<12 ms ) , and wide timing intervals ( 12 ms<Δt<30 ms ) . Under control aCSF conditions , the tLTP induced using narrow Δt ( 207±26% , n = 5 ) does not significantly differ from the tLTP induced using wide Δt ( 179 . 9±13 . 4% , n = 17 ) ( P = 0 . 72 ) . In contrast , when GluN2B-containing NMDARs are blocked with Ifenprodil , tLTP at intervals wider than +12 ms is abolished ( 97 . 7±6% , n = 7 ) , while tLTP between +5 ms and +12 ms was maintained ( 142±12% , n = 12 ) . This difference between wide and narrow timing intervals is significant for the Ifenprodil condition ( p<0 . 05; Figure 4D ) . These results demonstrate that GluN2 subunits not only control whether potentiation occurs , as previous studies have shown [26]–[28] , but also that they hone plasticity , making it sensitive to wider or narrower time intervals between pre-synaptic neurotransmitter release and post-synaptic firing . To reflect these experiments in the model cell , control aCSF simulations were run using a combined GluN2A and B decay time constant ( 25% GluN2B , 75% GluN2A [25] ) , and compared to the GluN2A only condition . For both cases the calcium response was normalized by the control aCSF no AP condition . The calcium curves generated by the model MSPN reflect the experimental STDP curves . When GluN2A and GluN2B are present , calcium elevation due to NMDAR stays above 250% of control for both narrow ( 2 ms<Δt<12 ms ) and wide ( 12 ms<Δt<30 ms ) Δt , whereas when GluN2A is present alone , calcium elevation due to NMDAR drops below 250% of control for wide timing intervals ( Figures 4E&F ) . Interestingly , when GluN2A and GluN2B are present , the interval at which NMDAR-mediated calcium drops below 250% in the model ( +40 ms Δt ) is the same interval that no longer gives tLTP in the experiments ( Figures 4C &E ) . During the narrow timing intervals ( 5 ms<Δt<12 ms ) , Ifenprodil significantly reduced the amount of tLTP induced for the experiments , and reduced the NMDAR mediated calcium in the model simulations . This result is not surprising considering that the synapses in the Ifenprodil condition have fewer total NMDARs available than those in the control condition . To test whether a general 75% reduction in NMDAR conductance would produce the same calcium effects as the GluN2A only condition , simulations were run using 75% of the NMDAR conductance and the combined GluN2A+B kinetics ( Figures 4E&F ) . This general reduction reduced the effect of AP pairing , but did not reproduce the timing difference seen with the GluN2A only case . Our model predicts that if NMDAR conductance is reduced by 75% , LTP would be reduced for both narrow and wide Δt . Previous studies in other neuronal types have shown that distance from the soma alters the strength [29] and direction of STDP [30] . Because the AP decays as it back-propagates in MSPNs [31] , we hypothesized that the timing dependence seen in secondary dendrites of these neurons would shift with distance from the soma . To investigate these interactions , we evaluated the peak calcium elevation during the 30 ms depolarization STDP protocol while stimulating spines located on the third segment of the tertiary dendrite , 150 µm away from the soma , compared to the secondary dendrites at 40 µm . Because a smaller branch diameter increases the input resistance at the tertiary dendrite , stimulating two adjacent tertiary spines resulted in a larger post-synaptic potential ( PSP ) than stimulating the two adjacent secondary spines , when seen at the spine head ( Figures 5A & B ) . However , the greater electrotonic distance for the tertiary stimulation causes the same stimulation to be smaller when seen at the soma ( Figures 5A&C ) . As predicted by cable theory [32] , the depolarization seen in the spine due to the back propagating AP is smaller in the tertiary dendrites than in the secondary dendrites ( Figure 5D ) . A reduced sensitivity to AP timing for positive Δt is observed under every GluN2 subunit condition ( Figure 5E ) , due to both the decrement in back-propagating AP , which decreases the amplitude of the NMDAR-mediated calcium elevation , and the larger PSP , which decreases the need for depolarization by the AP . In contrast , virtually no change in calcium elevation or timing dependence appears for negative Δt ( Figure 5E ) . The most drastic decreases in NMDAR-mediated calcium elevation occurs in the subunit conditions most sensitive to the magnesium block , i . e . GluN2A and GluN2B , while the GluN2C and GluN2D subunit conditions showed a smaller decrease in NMDAR-mediated calcium elevation due to distance . Because there is no change for negative Δt and a decrease in peak calcium for positive Δt , the difference between positive and negative Δt is strongly reduced when synapses are on the tertiary dendrite . For example , when GluN2A is stimulated on the tertiary spine , positive Δt causes nearly the same elevation in calcium as negative Δt ( Figure 5E inset ) . Under these distal dendrite conditions , positive Δt is unlikely to produce tLTP , and the dependence on AP timing is drastically reduced . The MSPNs of the striatum receive inputs from almost all areas of the cortex [33] and from the thalamus [34] . Using a recently developed slice preparation which preserves both cortico-striatal and cortico-thalamic fibers [35] , Ding et al . [25] and Smeal et al . [36] have characterized the two major glutamatergic inputs to the striatum in mouse and rat respectively . Ding et al . found that thalamo-striatal synapses had a lower NMDAR/AMPAR ratio than cortico-striatal synapses in the mouse , while Smeal et al . found the opposite , that the NMDAR/AMPAR ratio was higher for thalamo-striatal synapses in the rat . Measuring the decay time constant , Smeal et al . found that thalamo-striatal synapses had lower GluN2B/NMDAR ratios than cortico-striatal synapses , while using bath application of Ifenprodil Ding et al . found the opposite , that thalamo-striatal synapses had a higher GluN2B/NMDAR ratio . Smeal et al . also found that the thalamo-striatal synapses were electrotonically more distant from the soma than the cortico-striatal synapses , Ding et al . did not measure this ( Table 4 ) . To investigate the implications of differences in the NMDAR/AMPAR ratio , the GluN2B/NMDAR ratio , and the electrotonic distance from the soma , we modeled synapses with specific cortico-striatal characteristics and specific thalamo-striatal characteristics . Because the two studies report contrasting results [25] , [36] , we ran separate simulations using each set of data . We found that , based on different electrotonic properties , thalamo-striatal and cortico-striatal synapses differentially affect the NMDAR mediated calcium elevations for positive Δt , but not negative Δt . The distal thalamo-striatal synapses had lower peak calcium , and a weaker dependence on AP timing than did the cortico-striatal synapses for both mouse and rat ( Figure 6A ) . This result suggests that NMDAR-dependent tLTP would be more readily induced in cortico-striatal synapses than in thalamo-striatal . However , this result assumes that thalamo-striatal synapses are more distal than the cortico-striatal as has been experimentally suggested [36] . To distinguish the effect of distance from the effect of NMDAR/AMPAR and GluN2B/NMDAR ratios , simulations were repeated with thalamo-striatal synapses in the same location ( secondary dendrite ) as the cortico-striatal synapses . When the synapses were located at the same distance from the soma , the effect of Δt on peak calcium concentration did not differ strongly between thalamo-striatal and cortico-striatal synapses ( Figure 6A ) . Therefore , the lower calcium peaks in the thalamo-striatal simulations are entirely due to its location on the tertiary dendrite . Interestingly , this result was independent of which dataset ( Ding et al . or Smeal et al . ) was used . In conclusion , we predict that the cortico-striatal and thalamo-striatal synapses onto an MSPN will respond similarly to STDP protocols if they are located a similar distance away from the soma . However , if as Smeal et al . [36] have shown , the thalamo-striatal synapses are more distal , then we predict it will be more difficult to induce NMDAR dependent tLTP in those synapses than in cortico-striatal synapses .
In the model , NMDAR-mediated calcium elevations show strong sensitivity to AP timing in proximal dendrites , but this sensitivity as well as the maximum change in calcium is diminished at distal synapses , independent of STDP protocols and GluN2 subunit . Given the diminished effect of AP timing in tertiary dendrites , LTP at distal synapses might be achieved through entirely different plasticity mechanisms not requiring a somatic AP . A recent study suggests that a limited number of inputs on distal dendritic branches can induce upstates in MSPNs due to the high input resistance of the branch and non-linearity of the NMDA receptor [43] . Thus , the conjunction of many synaptic inputs may be more important than somatic depolarization for distal synaptic plasticity . However , further work is needed to investigate whether this upstate induction can induce LTP at synapses distal to the soma . We have used the recently characterized thalamo-striatal and cortico-striatal inputs to MSPNs to model synapses of each type . Simulations show that distance from the soma was the only factor that created a differential response to STDP protocols between these two types of synapse . Thus , our results predict that any disparity in plasticity expression between these synapses will not be NMDAR related , and will come from other neuromodulatory mechanisms , such as differences in pre-synaptic endocannabinoid receptor expression [44] . Our model shows that the time course and voltage dependence of GluN2 subunit can influence the way synapses respond to STDP protocols . GluN2A and GluN2B produce a stronger sensitivity to AP timing than GluN2C and GluN2D . Although the striatum contains mostly GluN2A and GluN2B subunits , our findings predict that neuron types with predominantly GluN2C or GluN2D will not exhibit NMDAR dependent tLTP . In addition , GluN2A , because of its fast decay time , results in a narrowing of the STDP curve ( compared to GluN2B ) by decreasing the Δt that permits sufficient calcium influx . While previous studies have focused on whether a particular GluN2 subunit is necessary for plasticity [26]–[28] , we have shown that the relationship between GluN2 subunit and plasticity is more complex than simply allowing or preventing LTP . In addition to the basic subunit characteristics modeled here , there are other , less well understood parameters that differ between the four GluN2 subunits . For example , differing amino acid patterns in the c-terminal tail allow differential phosphorylation by kinases and differential binding to scaffolding proteins [45] . Some of these differences are known to change the calcium permeability of the channel [46] , and could further influence the receptor's response to STDP . Similarly , the intracellular location of the GluN2 subunits is not well known in the striatum . A few studies have looked at the synaptic versus extrasynaptic location of GluN2A and GluN2B [47] , [48] in the striatum , but the techniques used were comparative rather than quantitative; thus we cannot be certain that we are not stimulating extrasynaptic NMDARs , but the low frequency stimulation ( always 1 Hz or lower ) used here is unlikely to cause significant glutamate spillover [49] . It is possible that NMDA triheteromers ( containing both GluN2A and B ) are present in the striatum ( [50] , but see [25] ) . Our experimental protocol does not distinguish between di or triheteromers and therefore we cannot specifically determine whether the effect is due to Ifenprodil's full blockade of GluN2B diheteromers or its weak blockade of GluN2A/B triheteromers [51] . The balance between GluN2A and GluN2B , and thus the shape of the STDP curve may be modulated dynamically . Studies in the hippocampus have demonstrated a widening of the STDP curve in response to dopamine D1 receptor stimulation [52] , and β-adrenergic receptor stimulation [53] . Interestingly , both D1 and β-adrenergic receptors activate the cAMP-dependent protein kinase ( PKA ) which is essential for striatal LTP [54] . PKA phosphorylation is known to increase the calcium permeability of NMDARs and in particular , those containing GluN2B subunits [46] . Therefore , PKA activation may widen the STDP curve by increasing the calcium permeability of GluN2B-containing NMDARs . The role for NMDAR subunits in neurological disorders has recently been suggested . Studies conducted in rodent models of Parkinson's Disease have shown that dopamine depletion results in the reconfiguration of the NMDARs , specifically in a reduction of GluN2B subunits [47] . Other studies have found that the administration of GluN2B antagonists reduces dyskinesia in animal models of Parkinson's Disease while GluN2A antagonists may increase it [55] . Similarly , GluN2B containing receptors are selectively potentiated by mutant huntingtin [56] , suggesting abnormal GluN2B subunit activity in Huntington's Disease . An imbalance of GluN2B containing NMDARs may allow non-specific potentiation that could lead to excessive and uncontrolled movement . Our results contribute to the emerging picture of GluN2 subunit antagonists as treatments for neurological disorders of the striatum by elucidating a possible mechanism for GluN2 subunit alterations to alter striatal plasticity and therefore motor behavior . Our findings suggest a novel role for the GluN2A NMDAR subunit in striatal synaptic plasticity . Instead of allowing or preventing LTP , this subunit hones plasticity , narrowing the STDP curve and allowing for the fine-tuning of neuronal pathway strength . Previous work has shown that the medial striatum undergoes plastic changes during the early , more coarse , phase of skill learning , while the lateral striatum undergoes plasticity during the late , fine-tuning , phase of skill learning [1] . Interestingly , the lateral striatum contains a higher ratio of GluN2A to GluN2B subunits than the medial striatum [13] . A high density of the GluN2A subunit may functionally underlie the fine-tuning phase of skill learning , allowing potentiation of only the most closely-timed connections . While the higher ratio of GluN2B subunits in the medial striatum would be useful for less specific , but possibly faster acquisition of a skill . This role for GluN2A may also underlie the experience-dependent developmental shift from GluN2B to GluN2A in the visual cortex [57] , and may be responsible for the increase in spatial learning ability that coincides with the developmental switch from GluN2B to GluN2A at hippocampal synapses [58] .
All plasticity experiments were performed in accordance with local animal welfare committee ( Center for Interdisciplinary Research in Biology and College de France ) and EU guidelines ( directive 86/609/EEC ) . For electrophysiology used to tune the computational model , animal handling and procedures were approved by the George Mason University IACUC committee ( Text S1 ) . Every precaution was taken to minimize stress and the number of animals used in each series of experiments . Animals , OFA rats ( Charles River , L'Arbresle , France ) ( postnatal days 17–25 ) were sacrificed by decapitation and brains were immediately removed . Patch-clamp recordings of MSPNs were performed in horizontal brain slices ( 330 µm ) from OFA rats . These horizontal slices included the somatosensory cortical area and the corresponding cortico-striatal projection field [22] and were prepared with a vibrating blade microtome ( VT1000S and VT1200S , Leica Micosystems , Nussloch , Germany ) . Patch-clamp whole-cell recordings were performed in the somatosensory area of the dorsal striatum and made as previously described [10] , [22] . Briefly , borosilicate glass pipettes of 5–7 MΩ resistance contained ( mM ) : 105 K-gluconate , 30 KCl , 10 HEPES , 10 phosphocreatine , 4 ATP-Mg , 0 . 3 GTP-Na , 0 . 3 EGTA ( adjusted to pH 7 . 35 with KOH ) . The composition of the extracellular solution was ( mM ) : 125 NaCl , 2 . 5 KCl , 25 glucose , 25 NaHCO3 , 1 . 25 NaH2PO4 , 2 CaCl2 , 1 MgCl2 , 10 µM pyruvic acid bubbled with 95% O2 and 5% CO2 . Picrotoxin ( 50 µM ) ( Sigma , Saint Quentin , France ) was dissolved in ethanol and then added in the external solution for a final ethanol concentration of 0 . 01% . All recordings were performed at 34°C using a temperature control system ( Bioptechs ΔTC3 , Butler , PA , USA and Bath-controller , Luigs&Neumann , Ratingen , Germany ) and slices were continuously superfused at 2–3 ml/min with the extracellular solution . Individual neurons were identified using infrared-differential interference contrast microscopy with CCD camera ( Hamamatsu C2400-07; Hamamatsu , Japan ) . Signals were amplified using EPC10-2 amplifiers ( HEKA Elektronik , Lambrecht , Germany ) . Current-clamp recordings were filtered at 2 . 5 kHz and sampled at 5 kHz , and voltage-clamp recordings were filtered at 5 kHz and sampled at 10 kHz using the program Patchmaster v2x32 ( HEKA Elektronik ) . The series resistance was compensated at 75–80% . Electrical stimulation of the cerebral cortex was performed with a bipolar electrode ( Phymep , Paris , France ) placed in the layer 5 of the somatosensory cortex [22] . Electrical stimulations were monophasic at constant current ( Stimulator WPI , Stevenage , UK or ISO-Flex stimulator controlled by a Master-8 , A . M . P . I . , Jerusalem , Israel ) . Currents were adjusted in order to evoke striatal excitatory postsynaptic currents ( EPSCs ) ranging in amplitude from 50 to 200 pA . Repetitive control stimuli were applied at 0 . 1 Hz , a frequency for which neither short- nor long-term synaptic efficacy changes in EPSC amplitudes were induced [22] . STDP protocols consisted in pairings of pre- and post-synaptic stimulations with the two events separated by a specific temporal interval ( Δt ) repeated 100 times at 1 Hz . Pre-synaptic stimulations correspond to cortical stimulations and the post-synaptic stimulation to an AP evoked by a direct application of a depolarizing current step ( 5 or 30 ms duration ) in the MSPN . Neurons were recorded for 10 minutes during baseline and for at least 60 minutes after the cellular conditioning protocol; long-term synaptic efficacy changes were measured after approximately 60 minutes . Series resistance was monitored and calculated from the response to a hyperpolarizing potential ( −5 mV ) step during each sweep throughout the experiments and a variation above 20% led to the rejection of the experiment . Repetitive control stimuli were applied at a frequency of 0 . 1 Hz for 60 minutes . Drugs were applied in the bath , after recording 10 minutes of baseline and 10 minutes before cellular conditioning protocol , and were present continuously until the end of the recording . Ifenprodil was dissolved in water at 15 mM and then added to extracellular solution for a final concentration of 10 µM ( Tocris , Ellisville , MO , USA ) . Off-line analysis was performed using Igor-Pro 6 . 0 . 3 ( Wavemetrics , Lake Oswego , OR , USA ) . All results are expressed as mean±SEM and statistical significance was assessed using the Student's t-test or the non-parametric Wilcoxon signed-rank test when appropriate at the significance level ( p ) indicated . Statistical analysis was performed using Prism 5 . 0 software ( San Diego , CA , USA ) . A dorsal striatum MSPN model cell was created based on the nucleus accumbens neuron model by Wolf et al . ( 2005 ) [16] . The Wolf model was translated from NEURON into Genesis simulation software , and channel concentrations and kinetics were adjusted to closely match those of a mature ( >3 weeks old ) MSPN in the mouse dorsal striatum ( Text S1 ) . The MSPN morphology is the same as in Wolf et al . [16] , but with the addition of individual spine compartments on the primary dendrites to the third segment of the tertiary dendrites ( Figure 1A ) . Primary dendrites are 20 µm long , secondary dendrites 24 µm , and each of the 11 tertiary segments is 36 µm long . For all simulations , two adjacent spines were stimulated ( on the third segment of the tertiary dendrite , or on the secondary dendrite ) , and the NMDAR-mediated calcium was recorded from one of the two spines . This model is available on Model DB: http://senselab . med . yale . edu/modeldb/ All channels kinetics ( Table 1 ) were taken from published data , using dorsal striatum MSPNs when possible . The Na+ kinetics were obtained from dissociated dorsal striatum MSPNs in the guinea pig [59] . The fast A-type potassium channel ( Kv4 . 2 ) data was obtained from slice dorsal striatum MSPNs in rat [60] . The slow A-type potassium channel ( Kv1 . 2 ) data was obtained from dissociated and slice dorsal striatum MSPNs in rat [61] . The inwardly rectifying potassium channel ( Kir ) kinetics were extracted from the computational studies of Wolf et al . [16] and Steephen and Manchanda [62] . The KDr channel was from Migliore et al . [63] The BK channel [64] and SK channel [65] were activated by a specific pool of calcium from the N and R type calcium channels , but not the T or L type calcium channels [16] , [66] . L ( Cav1 . 2 and Cav1 . 3 ) , N , R , and T type voltage sensitive calcium channels use the same parameters as Wolf et al . [16] . Calcium channels were added to the soma , and dendritic shafts . L , R , and T-type channels were also added to spines [20] . MSPNs of the dorsal striatum display different characteristics from those of the ventral striatum [67] . Both hand tuning and the simulated annealing parameter optimization routine in Genesis were used to adjust channel maximal conductances ( Table 2 ) , and channel activation and inactivation time constants ( Table 1 , scaling factor ) . These parameters were adjusted to match spike frequency , spike width , and latency to first spike extracted from current clamp data obtained at 30–32°C from mouse dorsal striatum MSPN ( Figures 1B–D ) . The change in channel time constants of NaF , KDr , and KAf ( scale in table 1 ) was required to produce the correct spike width as faster time constants produced spikes that were too narrow . In contrast , varying the maximal conductances by ±10% did not significantly change spike width ( data not shown ) . Δt for both experiments and the model MSPN is defined as the time from pre-synaptic stimulation ( stimulus artifact in the experimental case ) to the peak of the AP . AMPA and NMDAR channels were added using the synchan object in Genesis to all spines in the model . The synchan object uses equation 1 to calculate the conductance of the channel from the activation and inactivation time constants ( τ1 and τ2 respectively ) , time t relative to the action potential , and the maximal conductance ( gmax ) . K is a normalization constant which is calculated from the time constants and allows Gsyn to reach a peak value of gmax . ( 1 ) GABA channels were not added to this model , thus all simulations should be interpreted as occurring in the presence of GABA receptor antagonists . The default AMPA maximal conductance is 342 pS , which agrees with data from Carter and Sabatini [20] . Default NMDA maximal conductance was 940 pS to maintain the NMDA/AMPA ratio of 2 . 75/1 in cortico-striatal terminals [25] . AMPARs have an activation time constant ( τ1 ) of 1 . 1 ms and an inactivation time constant ( τ2 ) of 5 . 75 ms [16] , [68] . NMDARs have an activation time constant ( τ1 ) 2 . 25 ms [13] , but inactivation ( τ2 ) depends on subunit . Magnesium sensitivity to the NMDAR was implemented by using the “Mg_block” object in Genesis which utilizes equation 2: ( 2 ) In which , parameter B = 1/62 , while parameter A depends on subunit ( Table 3 ) . Specific GluN2-subunit containing NMDARs were modeled by adjusting the decay time constant , the maximal conductance , and the sensitivity to magnesium block according to published data [23] , [24] . GluN2 subunits differ in open probability [69] and affinity for glutamate [70] . These differences , though not modeled explicitly , contribute to the maximal conductance and the decay time which are taken into account in this model . Single decay time constants , averaged from the fast and slow time constants in Vicini et al . [24] , were used for each GluN2 subunit ( Table 3 ) . All NMDA decay time constants are adjusted for temperature by dividing by a scaling factor of 2 . Tau decay in tables 3 and 4 are the scaled values . Maximal conductances were calculated from the slope of the magnesium-free data [23] The ratio between the maximal conductance of GluN2A+B , C , and D was maintained , but the conductances were universally reduced such that the value for GluN2A+B ( the predominate subunits in striatal MSPNs ) matched cortico-striatal NMDAR/AMPAR ratios [25] . Current-voltage curves for each subunit in the presence of 1 mM magnesium from Monyer et al . [23] were matched by adjusting parameter “A” in this equation ( Table 3 , Figure S2B ) . | The striatum of the basal ganglia plays a key role in fluent motor control; pathology in this structure causes the motor symptoms of Parkinson's Disease and Huntington's Chorea . A putative cellular mechanism underlying learning of motor control is synaptic plasticity , which is an activity dependent change in synaptic strength . A known mediator of synaptic potentiation is calcium influx through the NMDA-type glutamate receptor . The NMDA receptor is sensitive to the timing of neuronal activity , allowing calcium influx only when glutamate release and a post-synaptic depolarization coincide temporally . The NMDA receptor is comprised of specific subunits that modify its sensitivity to neuronal activity and these subunits are altered in animal models of Parkinson's disease . Here we use a multi-compartmental model of a striatal neuron to investigate the effect of different NMDA subunits on calcium influx through the NMDA receptor . Simulations show that the subunit composition changes the temporal intervals that allow coincidence detection and strong calcium influx . Our experiments manipulating the dominate subunit in brain slices show that the subunit effect on calcium influx predicted by our computational model is mirrored by a change in the amount of potentiation that occurs in our experimental preparation . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] | [
"learning",
"and",
"memory",
"developmental",
"neuroscience",
"cellular",
"neuroscience",
"ion",
"channels",
"synaptic",
"plasticity",
"synapses",
"computational",
"neuroscience",
"single",
"neuron",
"function",
"biology",
"neuroscience",
"neurophysiology"
] | 2012 | The Effects of NMDA Subunit Composition on Calcium Influx and Spike Timing-Dependent Plasticity in Striatal Medium Spiny Neurons |
Circadian pacemakers are essential to synchronize animal physiology and behavior with the day∶night cycle . They are self-sustained , but the phase of their oscillations is determined by environmental cues , particularly light intensity and temperature cycles . In Drosophila , light is primarily detected by a dedicated blue-light photoreceptor: CRYPTOCHROME ( CRY ) . Upon light activation , CRY binds to the pacemaker protein TIMELESS ( TIM ) and triggers its proteasomal degradation , thus resetting the circadian pacemaker . To understand further the CRY input pathway , we conducted a misexpression screen under constant light based on the observation that flies with a disruption in the CRY input pathway remain robustly rhythmic instead of becoming behaviorally arrhythmic . We report the identification of more than 20 potential regulators of CRY-dependent light responses . We demonstrate that one of them , the chromatin-remodeling enzyme KISMET ( KIS ) , is necessary for normal circadian photoresponses , but does not affect the circadian pacemaker . KIS genetically interacts with CRY and functions in PDF-negative circadian neurons , which play an important role in circadian light responses . It also affects daily CRY-dependent TIM oscillations in a peripheral tissue: the eyes . We therefore conclude that KIS is a key transcriptional regulator of genes that function in the CRY signaling cascade , and thus it plays an important role in the synchronization of circadian rhythms with the day∶night cycle .
The rotation of the Earth on its axis is responsible for the mild temperatures found in most regions of the globe , which allow for a complex biosphere to thrive . However , this rotation is accompanied by important variations in light intensity and temperature , which are challenges for most organisms . Since the day∶night cycle has a stable period , the physical and ecological changes it induces in the environment can be temporally predicted . This anticipation is made possible in most organisms by circadian clocks . In Drosophila , the molecular circadian pacemaker is a transcriptional feedback loop: two proteins - PERIOD ( PER ) and TIMELESS ( TIM ) - repress their own gene transcription by interfering with the activity of the transcription factors CLOCK ( CLK ) and CYCLE ( CYC ) [1]–[4] . PER and TIM stability , their translocation into the nucleus , and their repressive activity are tightly regulated by kinases ( DBT , CKII , and SGG ) and phosphatases ( PP2A and PP1 ) [5]–[14] . Importantly , while this molecular clock free-runs - i . e . its oscillations persist in absence of environmental cues - its period only approximates 24 hr and must thus be reset every day to be in phase with the day∶night cycle . This is the role of light and temperature input pathways . The blue-light photoreceptor CRYPTOCHROME ( CRY ) is the main Drosophila circadian photoreceptor [15]–[17] . After absorbing a photon , CRY undergoes a conformational change involving its C-terminal domain and binds to TIM [18] , [19] . TIM is then tagged for ubiquitination and proteasomal degradation [20] , [21] . The mechanisms by which CRY initiates the cascade of events that leads to TIM degradation remain unclear . However , JETLAG ( JET ) plays an important role in targeting TIM for proteasomal degradation [22] , [23] . JET is part of an SCF E3 ubiquitin ligase complex responsible for TIM ubiquitination . Interestingly , JET also regulates CRY's own light-dependent degradation [24] . The COP9 signalosome subunits CSN4 and CSN5 are also required for circadian behavioral photoresponses [25] . The CSN complex regulates the activity of SCF E3 ubiquitin ligases and might thus be functioning upstream of JET in circadian neurons . Another protein known to regulate the CRY input pathway is the kinase SGG [26] . SGG interacts with CRY , and its overexpression inhibits CRY activity , through as yet unclear mechanisms . There is also little known about the regulation of the expression and stability of the proteins involved in CRY photoreception . This is an important question , because the level of expression of these proteins needs to be tightly regulated so that circadian rhythms are tuned to the proper range of light intensities . They should be able to respond to subtle and progressive changes in light intensities at dawn or dusk , without being excessively sensitive to light and for example inappropriately respond to moonlight levels of illumination . We therefore undertook a misexpression screen , which identified more than 20 genes that might affect circadian photoreception . We focused on one gene in particular , which encodes the chromatin remodeling protein KISMET . Indeed , by downregulating its expression with RNA interference , we found that KISMET is essential for CRY-dependent light responses .
The circadian behavior of wild-type flies is dramatically affected by the presence of constant light ( LL ) at an intensity of 10 lux or more [27] . They become totally arrhythmic , while under constant darkness they would remain rhythmic for weeks . This circadian response to constant light is dependent on the circadian photoreceptor CRY . cryb flies , which carry a severely hypomorphic cry mutation ( a quasi-null mutation ) , remain behaviorally rhythmic under constant light , with a periodicity of 24 hours , as if they were under constant darkness [15] ( Figure 1A ) . To identify new components of the CRY light input pathway , we decided to screen the P . Rørth collection of EP lines under LL . These EP lines carry randomly inserted P-elements in their genome ( Figure 1B ) [28] , which contain UAS binding sites . By crossing these EP lines to flies expressing GAL4 under the control of the timeless promoter ( tim-GAL4 flies ) , we overexpressed or in rare cases downregulated [28] the genes targeted by the EP element specifically in tissues with circadian clocks . Whether the targeted genes were up- or down-regulated depended on the orientation of the EP element insertion . A sense RNA is usually produced , which results in overexpression of the targeted gene ( Figure 1B ) . Sometimes however , the EP element generates an antisense RNA . Thus , at least four mechanisms could explain how EP lines might be rhythmic in LL when crossed to tim-GAL4 . First , we might overexpress a negative regulator of the CRY input pathway , such as SGG . Indeed , when SGG is overexpressed , flies are robustly rhythmic in LL [26] . Second , we might downregulate , with the few EP lines that generate antisense RNAs , genes crucial to CRY signaling . Third , the overexpression of a gene crucial to CRY signaling might be toxic for the CRY input pathway . Fourth , we might affect PER regulation , since its overexpression results in LL rhythms [26] , [29] . Supporting our strategy , we found that overexpressing JET - which promotes light-dependent TIM degradation and hence circadian photoresponses [22] , [23] - also results in rhythmic behavior in LL ( Figure 1A ) , although the long period length we observed indicates that these flies are not entirely blind to constant light [27] , [29] . We screened ∼1800 EP lines located on the second and third chromosomes , which target ∼1350 genes [30] or about 10% of the genome . 30 of these lines showed consistent rhythms in LL ( >50% rhythmic flies , Table 1 ) . This is a hit rate of 1 . 7% , which is in the range of other misexpression screens [28] , [30] . In constant darkness ( DD ) , none of these lines showed any obvious defect ( Table S1 ) . They were all robustly rhythmic , with a period length close to that of control flies . In LL however , the period length was usually not that of control flies in DD , or of cryb flies in LL . Periods were in most cases long , with most lines ( 20/30 ) showing a period range that was centered around 26 . 5–27 hr ( Figure 1A and 1C and Table 1 ) . Power was also weaker than under DD conditions , and the variability of period within lines was higher . This shows that the robustness of circadian rhythmicity is affected; this is not surprising since LL is disruptive to circadian rhythms . In sum , the selected lines are clearly not insensitive to LL , while cryb are virtually blind to constant light [15] , [31] . Of note is that 10 lines did not fall into the 26 . 5–27 hr range . 3 lines with a very long period ( 27 . 5–28 . 5 hr ) affected the same gene , miR-282 . However and as mentioned above , these lines all showed normal period length in DD . Thus , these differences in LL period length are not due to pacemaker dysfunction , i . e . , overexpression of specific genes results in specific LL phenotypes . In some lines , we occasionally observed one or two flies with complex behavior . Most of them displayed two components in their circadian behavior , one with a periodicity of 24 hr , and the other with 26 . 5 hr . This complex behavior was very rare , and probably occurred randomly . However , one line ( EP ( 2 ) 2356/miR-310-311-312-313 ) was strikingly different: about one quarter to a third of rhythmic flies showed a complex behavior when crossed to tim-GAL4 ( Table S2 ) . The short component had a periodicity that varied between 18 and 22 hr , while the long component varied from 26 to 29 hr , or more rarely was around 24 hr . The other rhythmic flies from this line showed only one component , mostly 18–22 hr , with a few individuals around 24 hr or 26–29 hr ( Table S2 ) . This behavior is to our knowledge unique . Complex behavior has been observed in cry0 flies , as well as in cryb and wild-type flies under specific light conditions , but the short period component had a period of approximately 23 hours [32]–[34] . Virtually all EP insertions of the collection have been mapped to the genome ( see Flybase , http://www . flybase . org ) . We verified the insertion location for eight lines , which were all identical to the insertion sites given in Flybase . For two genes ( lk6 and morgue ) , we also verified that they were indeed overexpressed as predicted ( data not shown ) . We can therefore predict which genes are misexpressed in the selected lines ( Table 1 ) . Among the candidate genes identified in our screen , we found seven genes that regulate gene expression ( Table 2 ) . Among them were three transcription factors . elB is known for its role in trachea and appendage development [35] , [36] . kay is the Drosophila homologue of fos and is implicated in several signal transduction pathways . KAY is for example essential during embryonic development , for the differentiation of R3/R4 photoreceptors and in immune responses [37]–[42] . kis encodes a chromatin-remodeling protein of the Trithorax family [43] , [44] . cpo was also identified in our screen ( EP ( 3 ) 661 ) . A second EP line affecting cpo ( EP ( 3 ) 3611 ) was also detected initially , but its phenotype was weaker and thus not listed in Table 1 . CPO is an RNA binding protein that regulates different aspects of Drosophila behavior ( flight , phototaxism , negative geotactic behavior for example ) [45] . An interesting set of lines that resulted in LL rhythmicity affected microRNA genes ( EP ( 3 ) 3041/miR-282 , EP ( 3 ) 714/miR-282 , EP ( 3 ) 3718/miR-282 , EP ( 3 ) 3617/miR-282 , EP ( 2 ) 2402/miR-8 , EP ( 2 ) 2356/miR-310-311-312-313 ) . Strikingly , among the six lines with the strongest phenotype were three lines ( EP ( 3 ) 3041 , EP ( 3 ) 714 and EP ( 3 ) 3718 ) that are predicted to overexpress miR-282 . A fourth line ( EP ( 3 ) 3617 ) affects the same gene , and also showed a robust phenotype . A second category of interesting candidate genes regulates protein stability . Two genes are implicated in proteasomal degradation ( morgue and slimb ) . MORGUE has an E2 ubiquitin conjuguase domain [46] , and SLIMB is an E3 ubiquitin ligase implicated in the control of PER levels [47]–[49] . In addition , overexpressing the E3 ubiquitin ligase complex subunit JET also results in LL rhythms ( Figure 1A ) . Since these three proteins carry an F-box , we wondered whether the overexpression of any protein with an F-box or involved in proteasomal degradation could render flies resistant to LL . This was not the case: overexpressing the ubiquitin conjuguase UBCD1 or the F-box containing E3 ligase AGO did not result in LL rhythmicity ( Figure S1 ) . This demonstrates the specificity of the constant light phenotype to MORGUE , SLIMB and JETLAG . Two additional candidate genes encode proteases . CALPB is a calcium activated protease of the calpain family [50] , whereas another protease , SDA , is known to be important for normal nervous system function: sda mutants are prone to seizures [51] . A kinase ( lk6 ) involved in the control of growth through its action on the translation initiation factor 4E [52] , [53] and a putative protein phosphatase ( cg9801 ) were also isolated . For several lines , the identity of the targeted gene was unclear . Two lines were located in front or within a genomic region with two overlapping genes . One of them ( EP ( 3 ) 703 ) was among the lines with the strongest behavior . The two overlapping genes encode a putative metalloprotease ( cg8176 ) and a putative transcription factor of the jumanji family ( cg8165 ) . Finally , for five EP lines ( EP ( 2 ) 575 , EP ( 3 ) 902 , EP ( 2 ) 2345 , EP ( 2 ) 813 and EP ( 2 ) 2469 ) the targeted genes are not up-regulated , but probably down-regulated , since the P-element is predicted to produce an antisense RNA . For EP ( 2 ) 575 , EP ( 3 ) 902 , EP ( 2 ) 2345 and EP ( 2 ) 813 , the identity of the gene which might be down-regulated is uncertain , because these EP-elements are located near the very 5′end of the Rapgap1 , kay , dap and wech genes , respectively . They might thus affect the genes located 5′ of these three genes . EP ( 2 ) 2469 is inserted in the 12th intron of the kis gene . It could also potentially overexpress cg13693 , which is nested 10kb downstream of the EP-element , within kis 4th intron . Evidence presented below will demonstrate that kis downregulation is sufficient to explain the LL phenotype obtained with EP ( 2 ) 2469 , although we cannot entirely exclude a contribution of cg13693 overexpression to this phenotype . To determine how profoundly misexpression of the candidate genes disrupts the CRY input pathway , we tested the response to short light pulses , which is entirely CRY-dependent [16] . In wild-type flies , a 5-min light pulse induces a phase delay when administered at the beginning of the night and a phase advance when administered at the end of the night [54] , [55] . All EP lines selected from the primary screen were challenged with 5-min light pulses at ZT15 and ZT21 . Three EP lines responded poorly to light pulses at high intensity ( 1000 lux ) : EP ( 2 ) 2367/morgue , as described previously [29] , EP ( 3 ) 3041/miR-282 and EP ( 3 ) 714/miR-282 ( Figure 2 and data not shown ) . For EP ( 3 ) 3041/miR-282 we observed a clear reduction of the response to light at ZT15 compared to all controls . At ZT21 , we also observed a reduction , although the magnitude of the phase shift was not significantly different from that of one control ( Canton-S ) . A similar phenotype was observed with EP ( 3 ) 714/miR-282 ( data not shown ) . The other EP lines responded normally to short light pulses . Previous studies have shown that among CRY dependent photoresponses , the behavioral responses to constant light are far more sensitive to partial disruption of the CRY input pathway [18] , [23] . Thus , it appears that morgue and miR-282 overexpression have the most profound effect on CRY signaling , while overexpression of the other candidate genes affect the CRY input pathway more moderately . As mentioned above , most of the EP lines show a long period phenotype and a weak power , which demonstrate that their circadian clock is not insensitive to light . Strikingly , the three lines that disrupt phase responses to short light pulses are among those with the strongest phenotype in LL . To identify positive regulators of CRY signaling among our candidate genes , we screened loss-of-function mutants under constant light ( 200 lux ) . For 5 candidate genes , we tested severe loss-of-function alleles that were homozygous viable and healthy ( lk6 , morgue , akap200 , GstS1 and sda , see Materials and Methods ) . They were all completely arrhythmic in LL ( data not shown ) . For most candidate genes however , severe loss-of-function mutations were either lethal or not available . We therefore decided to use RNA interference , targeted specifically to circadian tissues , which should in most cases downregulate candidate gene expression and avoid lethality . We crossed tim-GAL4 flies to transgenic flies from the VDRC and NIG-Fly RNAi collections , which carry transgenes encoding gene-specific double-stranded RNAs ( dsRNAs ) . Lethality was only observed with three candidates: kay , cpo and with one of the RNAi line targeting elB . Most viable tim-GAL4/dsRNA flies were arrhythmic in LL . However , two lines targeting the chromatin remodeling gene kismet ( NIG-Fly3696R-1 and VDRC46685 ) were 50–60% rhythmic , with a power of about 35 ( Table 3 ) . These results are strikingly similar to those obtained with EP ( 2 ) 2469/kismet , which as mentioned above is inserted in an antisense orientation within kis' 12th intron , and is thus predicted to generate a kis antisense RNA . The only noticeable difference in behavior between the two RNAi lines was that NIG-Fly3696R-1 shows a period close to 24 hr , whereas VDRC46685 and the EP line have a ca . 26 hr period rhythm . This could indicate that the NIG-Fly line is more efficient at repressing KIS expression . In DD , both kis RNAi lines were rhythmic with a period close to that of control flies ( Table S3 ) . Since kis RNAi flies are clearly less robustly rhythmic than cryb flies , we co-expressed dicer2 along with the kis dsRNAs , to increase the RNAi effects [56] . This approach proved successful , as flies expressing the VDRC dsRNAs directed against kis were almost as robustly rhythmic as cryb flies ( Table 3 , Figure 3 ) . The period was also similar to that observed in DD ( Table S3 ) , further indicating that these flies are virtually insensitive to constant illumination and that KIS is essential for normal circadian photoresponses . As the period length and amplitude of DD rhythms are not affected by KIS downregulation ( Table S3 ) , KIS specifically regulates circadian photoresponses . A major potential caveat with expression of dsRNAs is off-target effects: dsRNAs sometimes target other genes than the one they were designed to downregulate . With kis , we have observed a similar phenotype with two independent RNAi transgenes that target two different regions of the kis RNA . In addition , EP ( 2 ) 2469/kismet targets a third region of the kis gene specific to the KIS-L isoform , which contains most domains necessary for KIS function in chromatin remodeling [43] . Thus , the risks of off-target effects are virtually non-existent . In addition , no potential off-targets are predicted for VDRC46685 ( http://stockcenter . vdrc . at/control/main ) . Nonetheless , we verified that the RNAi lines we used indeed downregulate KIS expression . We first measured KIS expression in whole heads by Western Blot using the tim-GAL4 driver to induce kis RNAi , and found KIS levels to be reduced by 30% with both RNAi lines ( Figure 4A and 4B ) . Since the dsRNAs were expressed with tim-GAL4 , this highly reproducible reduction in KIS expression should reflect the downregulation occurring specifically in circadian tissues , in particular the eyes which are by far the largest contributors of circadian tissues in whole heads [57] . KIS should not be down-regulated in non-circadian tissues . There are ∼150 circadian neurons in the Drosophila brains that are organized in 5 clusters in each hemisphere: the large and the small ventral lateral neurons ( l-LNvs and s-LNvs ) , which express the neuropeptide PDF , the dorsal lateral neurons ( LNds ) and three groups of dorsal neurons ( DN1s , 2s and 3s ) . We verified that KIS-L expression was indeed specifically down-regulated in circadian neurons by staining brains from flies expressing kis dsRNAs in clock tissues and control flies with an antibody directed against KIS-L [43] along with an anti-PDF antibody . KIS-L was strongly expressed in the LNvs of wild-type flies , as well as in non-circadian neurons . KIS-L levels were unaffected in non-circadian neurons of flies expressing the dsRNAs with tim-GAL4 , but KIS-L expression was severely downregulated in the LNvs ( Figure 4C , Figure S2A ) . In summary , our results identify KIS as a crucial regulator of CRY-dependent circadian photoresponses . This validates our screen strategy . Since KIS regulates constant light responses , it should be expressed in circadian neurons known to control these responses . To identify easily circadian neurons , we stained brains expressing a lacZ reporter gene in all clock neurons ( line R32 , [58] ) with antibodies directed against ß-GAL and KIS-L , because our results indicate that this isoform regulates light responses ( see above ) . We found KIS-L to be expressed in l- and s-LNvs , in the LNds , and in DN1s and 3s ( Figure 5A ) . Most of these neurons have been implicated in circadian behavioral photoresponses [26] , [29] , [33] , [59]–[65] . In particular , the LNds and DN1s have been shown to be able to generate LL rhythms when the CRY input pathway is inhibited in these cells . That KIS is expressed in these cells , and that KIS downregulation results in LL rhythms clearly support the idea that KIS is a crucial regulator of the CRY input pathway . We were therefore curious to verify that behavioral rhythms observed under constant illumination in kis RNAi flies were due to KIS downregulation in dorsally located circadian neurons , and not in the PDF positive LNvs . We thus used the pdf-GAL4 driver [66] to restrict kis RNAi expression to the LNvs , and a combination of tim-GAL4 and pdf-GAL80 to express kis dsRNAs in dorsal neurons only [67] . kismet down-regulation in the LNvs alone did not affect behavior in LL: all flies were arrhythmic . On the other hand , flies expressing kis dsRNAs in PDF negative circadian neurons were rhythmic in LL ( Figure 5B ) . The absence of LL rhythms when driving kis dsRNAs only in the LNvs is not due to inefficient KIS downregulation: pdf-GAL4 and tim-GAL4 are equally efficient at downregulating KIS expression ( Figure S2A ) . Together , these results demonstrate that KIS activity is required in PDF negative light-sensitive circadian neurons for normal CRY-dependent circadian photoresponses . CRY targets TIM to proteasomal degradation [16] , [20] , [21] . The prediction would therefore be that if KIS regulates the CRY input pathway , light-dependent TIM degradation would be inhibited by KIS downregulation . Unexpectedly , we did not observe any defect in TIM light-dependent protein cycling in head extracts of flies expressing kis dsRNAs ( data not shown ) . This is probably because RNA interference does not completely abolish KIS expression in the eyes , which are the main contributor of TIM protein in head extracts . Since CRY levels appear to vary significantly in different circadian tissues [68]–[70] , we reasoned that CRY ( or its signals to the pacemaker ) might be limiting in the PDF negative neurons generating LL rhythms ( LNds , DN1s ) , but not in the eyes . This would explain why LL behavior is particularly sensitive to KIS levels . We therefore overexpressed CRY in all clock neurons of flies expressing kis dsRNAs . As we anticipated , this genetic manipulation rescued the behavioral phenotype resulting from KIS knockdown: the double-manipulated flies behaved like wild-type flies as they became arrhythmic after their release in LL ( Figure 6A and Figure S3 ) . Thus , when CRY is not limiting , circadian photoresponses are much less sensitive to KIS levels . We therefore decided to express kis dsRNAs in crym mutant flies [18] . The crym allele encodes a truncated form of CRY lacking the last 19 amino acids of its C-terminal domain . This truncated CRY is very unstable but remains functional . In crym flies , TIM levels oscillate in LD but the amplitude of this molecular rhythm is significantly reduced relative to wild-type flies , because CRY levels are limiting [18] ( Figure 6B and 6C ) . Interestingly , when we expressed kis dsRNAs in crym flies , TIM cycling amplitude was further reduced to about half of what it normally is in crym flies . This effect was highly reproducible and statistically significant ( p<0 . 05 , t test ) . Notably , TIM levels remained higher during the day , even though during the night they were reduced . Note that TIM cycling is not sensitive to eye color: no difference was observed in amplitude between the white-eyed crym flies ( y w; crym ) and deep orange-eyed crym flies ( y w; tim-GAL4 UAS-dcr2/+; crym; Figure 6B and 6C ) . Thus , we are confident that the addition of the kis dsRNA transgene ( in y w; tim-GAL4 UAS-dcr2/VDRC46685; crym ) , which only very slightly darkens eye color compared to control flies ( y w; tim-GAL4 UAS-dcr2/+; crym ) , does not contribute to the reduction in TIM cycling amplitude . Moreover , in cry+ flies , CRY and TIM levels are unaffected by this slightly darker eye color ( data not shown ) . In summary our results demonstrate that kis interacts genetically with cry and regulates circadian photoresponses not only in brain pacemaker neurons , but also in peripheral oscillators such as the eyes .
Our constant light screen has identified a novel important regulator of circadian photoresponses: KISMET ( KIS ) . Our results show that KIS is essential for a well-characterized CRY-dependent circadian photoresponse: constant light ( LL ) induced behavioral arrhythmicity . This arrhythmic behavior is caused by the persistent activation of CRY by blue-light photons . CRY thus binds constantly TIM and tags it for JET-mediated proteasomal degradation . This leaves PER unprotected from being itself targeted to proteasomal degradation , and leads to the disruption of the molecular pacemaker in neurons controlling circadian behavior . Consistent with previous studies in which the CRY input pathway was partially disrupted [18] , [23] , we observed that among CRY-dependent light responses , LL arrhythmicity is much more sensitive to reduction in KIS expression . Our results further suggest that CRY levels are limiting in circadian neurons that can generate LL behavior . It is becoming clear that CRY level varies significantly between circadian neurons [69] , [70] , and it is therefore likely that those with lower CRY levels are more prone to become rhythmic in LL when the CRY input pathway is partially disrupted ( this study and [26] , [29] ) . We presume that this reflects the very fine photosensitive tuning needed for CRY photoresponses . The circadian input pathway has to be able to respond to photoperiod length and to progressive light intensity changes at dawn and dusk , but at the same time should not respond inappropriately to moonlight . KIS role as a regulator of the CRY input pathway is not limited to circadian neurons . It also influences circadian photoresponses in peripheral circadian tissues such as the eyes , since we observed a significant reduction in diurnal CRY-dependent TIM protein oscillations in crym flies expressing kis dsRNAs . Whether KIS is actually essential for light responses in every circadian tissue is not yet clear . Indeed , we were limited to use RNA interference - which usually does not completely abolish gene expression - to study KIS function in adult flies , because kis null mutants are lethal . Incomplete KIS knockdown might explain why we could not detect any defects in the phase response to short light pulses ( data not shown ) . The neurons controlling this circadian photoresponse are at least partially distinct from those controlling LL behavioral responses [59]; they may have retained sufficient residual KIS or having high CRY levels . Alternatively , KIS might not be essential in these circadian neurons . KIS is a chromatin-remodeling enzyme that was initially discovered in a screen for extragenic suppressor of Polycomb mutations [71] . It was thus categorized as a Trithorax protein , a group of transcriptional activators of homeotic genes that counteracts Polycomb negative regulators . Recent evidence obtained with larval salivary glands has suggested that KIS might be a general regulator of transcription [43] . Indeed , KIS is associated with most , but not all , transcriptionally active sites of larval salivary gland polytenic chromosomes . In kis mutants , RNA Polymerase II is associated with these sites , but remains hypophosphorylated , which indicates that it is unable to initiate mRNA elongation . In addition , elongation factors are not recruited at transcriptionally active sites . It was therefore proposed that KIS is necessary for the recruitment of these elongation factors and for reorganizing chromatin downstream of the transcriptional start site [43] , [72] . However , while we observed strong effects on circadian light responses , we did not detect any significant effects on the period of the circadian oscillator . Our results therefore indicate that KIS is specifically involved in the control of circadian light input genes in neurons controlling circadian light responses . Since we have used RNA interference to disrupt KIS expression , we cannot entirely exclude that residual KIS expression is sufficient for maintaining normal transcription of pacemaker and non-circadian genes . We also cannot exclude the possibility that other chromatin remodeling enzymes could substitute for KIS in the control of these genes . There is however very clear experimental evidence that supports the idea that KIS regulates the expression of specific genes . kismet loss-of-function results in specific segmentation defects and homeotic transformation during development [44] . Moreover , a recent study demonstrates that KIS plays a central and specific role in the regulation of atonal , a pro-neural gene , in the fly retina [73] . We therefore strongly favor the hypothesis that KIS is dedicated to the control of circadian light input genes in Drosophila circadian clock neurons . We do not know yet the identity of these genes . We have measured by Real-Time quantitative PCR the expression of the known components of the CRY input pathway ( cry , sgg , jet , tim , csn4 , csn5 ) but did not detect any significant change in their mRNA levels when KIS is knocked-down ( data not shown ) . This indicates that important elements of the CRY input pathway remain to be identified . These proteins either function downstream of CRY or regulate CRY activity , but they apparently do not affect CRY abundance . Indeed , we could not detect any changes in CRY levels by Western Blots or brain immunostainings in flies expressing kis dsRNAs ( data not shown ) . The demonstration that KIS is essential for circadian light responses validates our screen for circadian light input genes , which has identified over 20 additional genes that might regulate circadian light responses . As most of these genes were overexpressed in the screen , a significant fraction of them might be negative regulators of the CRY input pathway . It is thus entirely possible that a reduction in their expression levels or a complete loss-of-function would result in an increase in light sensitivity , rather than a loss of CRY responses . Our loss-of-function subscreen was aimed at genes essential for the CRY input pathway and would not have detected genes that increase light sensitivity . It is therefore not surprising that we confirmed a gene ( kis ) that was predicted to be downregulated in our initial screen as essential for circadian photoresponses . Future studies will determine whether other candidate genes are negative regulators of CRY signaling . They will also be aimed at determining whether some of the candidate genes might be part of the circadian pacemaker , rather than regulators of CRY signaling . This is entirely possible , since overexpression of circadian pacemaker genes such as PER or TIM results in LL rhythms [26] , [29] . Actually , we identified one other pacemaker gene in our screen: slimb [48] , [49] . The isolation of this gene is unexpected however . slimb overexpression would be predicted to reduce PER levels , since it promotes PER ubiquitination and proteasomal degradation . A possibility is that slimb overexpression is toxic to the PER degradation pathway , and thus results in an increase , rather than a decrease , in PER levels . This idea is supported by the fact that strong slimb overexpression results in the same circadian phenotype as slimb loss-of-function mutations: arrhythmic behavior in DD [48] , [49] . Moreover , we observed that overexpressing jet - which is involved in proteasome-dependent protein turnover , like slimb [23] - disrupts circadian photoresponses in LL . This could also be explained by a dominant-negative effect of jet overexpression on TIM proteasomal degradation . However , recent results demonstrating that JET also promotes CRY proteasomal degradation [24] point at another potential explanation: CRY levels might be reduced when jet is overexpressed . In any case , the negative effect of jet overexpression on circadian light responses might explain why using the GAL4/UAS system to try to correct the photoreceptive defects of jet mutants resulted only in a partial rescue [23] . In addition to slimb , several other genes have been connected to circadian rhythms: lk6 , akap200 , calpB and morgue . The mRNAs of the last three genes were shown to cycle in the fly head in a DNA microarray study [74] , while lk6 was shown to oscillate in fly bodies [75] . RNase protection and Northern Blot assays revealed that lk6 expression also undergoes circadian oscillations in heads , with a cycling phase and amplitude of oscillation similar to those of cry ( Figure S4 ) . Thus lk6 and cry may be co-regulated . It might however be the presence of microRNAs in our screen that is most intriguing . MicroRNAs are known to play an important regulatory role in a variety of biological processes , which include development and neuronal function . In mouse , two miRNAs , miR-219 and miR-132 , have recently been shown to be under circadian regulation in the suprachiasmatic nucleus , and miR-132 may be important in the regulation of photic responses [76] . In Drosophila , a role for miRNAs in the control of circadian rhythms has recently been demonstrated . In particular , bantam was shown to regulate CLK expression and thus to affect the amplitude of circadian rhythms [77] . In addition , a few miRNAs have been shown to be under circadian regulation [78] , although their importance for circadian rhythms is currently unclear . None of the miRNAs we isolated are described to cycle in fly heads . However , two of them are expressed in circadian tissues: miR-282 and miR-8 [77] . This strongly supports the idea that these miRNAs are important for the regulation of circadian rhythms . miR-282 appears particularly likely to be an important regulator of circadian photoresponses , since its overexpression affects profoundly both the behavior of the flies in LL and their response to short light pulse . Moreover , a predicted target of miR-282 is jetlag ( TargetScanFly , release4 . 2; Yong and Emery , unpublished observations ) , which is crucial for CRY signaling and TIM degradation . We are currently determining whether miR-282 is indeed a regulator of the CRY input pathway . In summary , our work has identified new candidate circadian genes . They might control or modulate circadian light responses and photosensitivity , or they might regulate circadian pacemaker function . Importantly , we have assigned a function to the chromatin-remodeling factor KISMET in adult flies: KIS control the photosensitivity of the circadian clock . The function of most chromatin-remodeling proteins is well documented during Drosophila development , but their function in the adult fly is not well studied , because null mutants for these genes are frequently lethal . The adult function of KIS was completely unknown , although we show here its expression in both circadian and non-circadian fly brain neurons . CHD7 is a human KIS homolog associated with CHARGE syndrome , a genetic disorder characterized by developmental retardation and complex abnormalities affecting several organs , including the brain and sensory systems [79] , [80] . The partial loss of CRY signaling should be a powerful tool for a genetic screen aimed at finding KIS-interacting genes that contribute to transcriptional regulation . This might in turn reveal how CHD7 functions , and help illuminate the causes of CHARGE syndrome .
The EP line collection was previously described [28] . The following strains were used: y w , Canton-S , y w; tim-GAL4/CyO [17] , y w; pdf-GAL4 [66]; y w; cryb ss [16]; y w; crym [18]; y w ; tim-GAL4 pdf-GAL80/CyO; pdf-Gal80/TM6B [29]; y w; UAS-cry #12 [17] . The second chromosomes bearing tim-GAL4 and pdf-GAL4 insertions were independently and meiotically recombined with a chromosome containing the UAS-dcr2 insertion [56] . The presence of the two transgenes on the same chromosome was verified by PCR . To localize the circadian neurons in Drosophila brain , the previously described enhancer trap line R32 was used [58] . We also used the following mutant strains: sdaiso7 . 8 [51] , Akap200Δ7 [81] , lk61 and lk62 [52] , Df ( 3R ) exel9019 , GstS1M38-3 and GstS1M29-13 ( Benes et al . , unpublished GstS1 null mutants ) , morgueΔ457 ( Schreader and Nambu , unpublished deletion strain ) . To induce RNAi of the candidate genes , the following lines were used from the NIG-Fly stock and the VDRC . NIG-Fly lines: 30152R-1 ( cg30152 ) , 30152R-2 ( cg30152 ) , 10459R-1 ( cg10459 ) , 3412R-1 ( slimb ) , 3412R-2 ( slimb ) , 15437R-1 ( morgue ) , 15437R-2 ( morgue ) , 15507R-2 ( kay ) , 15507R-4 ( kay ) . VDRC lines: 4024 ( cg8735 ) , 4025 ( cg8735 ) , 5647 ( akap200 ) , 11763 ( elB ) , 42813 ( elB ) , 14385 ( cpo ) , 14691 ( cg1621 ) , 14692 ( cg1621 ) , 22144 ( sda ) , 22145 ( sda ) , 23037 ( calpB ) , 46241 ( calpB ) , 25033 ( cg31123 ) , 25034 ( cg31123 ) , 31674 ( cg1273 ) , 31676 ( cg1273 ) , 32885 ( lk6 ) , 38326 ( cg10082 ) , 38327 ( cg10082 ) , 38848 ( cg30152 ) , 41451 ( cg10459 ) , 41623 ( wech ) , 35794 ( HSPC300 ) . kismet RNAi was induced using either the line 3696R-1 ( NIG-Fly ) or 46685 ( VDRC ) . UAS-ubcd1 was described in [82] , and AGO overexpression was obtained with EP ( 3 ) 1135 [49] . Two alleles of tim can be found in lab stocks: ls-tim and s-tim [22] . s-tim is more sensitive to light than ls-tim . Our y w stock carries the s-tim allele , while w1118 and Canton-S carry the less sensitive ls-tim . The 2nd chromosome with the tim-GAL4 insertion carries ls-tim , as well as the 2nd chromosomes of UAS-cry ( line #12 ) , UAS-dcr2 , EP ( 3 ) 3041 , EP ( 3 ) 714 , VDRC46685 , y w; cryb , y w; crym stocks and y w; pdf-GAL4 UAS-dcr2 ( rec4 ) /CyO . NIG-Fly3696R-1 and the tim-GAL4 UAS-myccry recombinant chromosome carries s-tim . We also generated a VDRC46685 stock with a s-tim allele , and a recombinant tim-GAL4 VDRC46685 chromosome carrying the ls-tim allele . All these lines are wild-type for jet . Appropriate controls were included in all tables and figures to verify that the decrease in light sensitivity observed with kis downregulation or with miR-282 overexpression was not due to tim variants . 20 fly heads were homogenized in extraction buffer ( 20mM HEPES pH7 . 9 , 100mM KCl , 5% Glycerol , 0 . 1% Triton X100 , 0 . 1mM DTT and 1× protease inhibitor [Roche] ) , 5× SDS-PAGE loading buffer was then added , and samples denatured at 100°C for 10 minutes . After centrifugation , protein extracts were loaded on 5% and 6% 29 . 6∶0 . 4 acrylamide∶bisacrylamide gels for KIS and TIM , respectively , and a 9% 29∶1 acrylamide∶bisacrylamide gel ( for Tubulin ) . Proteins were transferred on nitrocellulose and blots were incubated with primary antibody ( 1∶1000 for anti-KIS-L [43] , 1∶5000 for anti-TIM [57] and 1∶10 , 000 for anti-Tubulin [Sigma] ) and HRP-conjugated secondary antibodies ( Jackson Immuno Research ) . Films were imaged with the Fujifilm LAS-1000 and band intensities were quantified using the ImageGaugev4 . 22 software . KIS or TIM protein levels were normalized with Tubulin . Brains from adult flies were dissected in 1× PBS , 0 . 1% Triton ( PBT ) , fixed in 4% paraformaldehyde in PBT and blocked in 10% Normal Goat Serum ( Jackson Immuno Research ) in PBT . They were then washed and incubated overnight at 4°C with primary antibodies . After several washes in PBT , they were incubated with secondary antibodies ( Jackson Immuno Research ) coupled to FITC , Rhodamine , Cy3 or Cy5 for 2–3hours at room temperature at a concentration of 1∶200 . Brains were mounted in antifade reagent ( Biorad or Vectashield ) . The anti-KIS-L ( generous gift from J . Tamkun ) and anti-PDF antibodies were previously described [43] , [85] and both used at a 1∶400 dilution . For ß-Gal immunostaining , we used a mouse anti-ß-Gal antibody ( Promega ) at a concentration of 1∶1000 as previously described in [58] and CRY antibody was used at a concentration of 1∶200 . Levels of KIS protein expression in circadian neurons were quantified using ImageJ v1 . 42q ( http://rsb . info . nih . gov/ij ) . For each genotype , 5 to 12 neurons from at least 3 different brains were quantified . For each circadian neuron , the nuclear fluorescence corresponding to KIS staining was measured and normalized , after subtraction of background signal , to the fluorescence of two neighbor neurons on the same focal slice . An average of the two normalized values was then calculated for each circadian neuron . Total RNAs from about 60 fly heads collected at ZT 4 and ZT16 were prepared using Trizol ( Invitrogen ) according to the manufacturer's instructions . 2 µg of total RNAs were then treated with RQ1 DNAse ( Promega ) for 2 hours and subsequently reverse transcribed using random hexamer primers ( Promega ) and Superscript II ( Invitrogen ) , following manufacturer's instructions . Real-time PCR analysis was performed using SYBR Green fluorescent dye ( Biorad ) in an ABI SDS 7000 instrument ( Applied Biosystems ) . For each set of primers , we generated a standard and a melting curve , using RNAs extracted from wild-type fly heads , to verify amplification efficiency and specificity , respectively . For each transcript , data were normalized to rp32 using the 2−ΔΔCt method . The concentration of per , tim , cry , csn4 , csn5 , sgg and jet were measured in flies expressing kis dsRNAs ( tim-GAL4 UAS-dcr2/VDRC46685 ) and in two control genotypes ( tim-GAL4 UAS-dcr2/+ and VDRC46685/+ ) . Primers used: rp32-forward ATGCTAAGCTGTCGCACAAA; rp32-reverse GTTCGATCCGTAACCGATGT; tim-forward TGAACGAGGACGACAAAGCC; tim-reverse GATTGAAACGCCTCAGCAGAAG; per-forward TCATCCAGAACGGTTGCTACG; per-reverse CCTGAAAGACGCGATGGTGT; cry-forward CCGCTGACCTACCAAATGTT; cry-reverse GGTGGCGTCTTCTAGTCGAG; csn4-forward AGCAAGTTGCCTGACGATCT; csn4-reverse GAAACGTATGCCAGCCACTT; csn5-forward ACCCAGATGCTCAACCAGAC; csn5-reverse CTTTTGGATACGTGCGGAAT; sgg-forward TGCTGCTCGAGTATACGCCC; sgg-reverse TCCATGCGTAGCTCATCGAAG; jet-forward CTGCTGCAGTCACTGATGGT; jet-reverse ATGTTGCACAGTTGGCATGT . RNase protections were performed and quantified with an rp49 loading control as described in [17] . The lk6 probe covered nucleotide 919–1055 of transcript lk6-RA . It generated two major bands . The larger one corresponds to transcript lk6-RA , and the smaller one most likely to lk6-RB . Quantification for the RA band is shown on Figure S4 . The smaller band cycled with a similar phase and amplitude ( data not shown ) . | In most organisms , intracellular molecular pacemakers called circadian clocks coordinate metabolic , physiological , and behavioral processes during the course of the day . For example , they determine when animals are active or resting . Circadian clocks are self-sustained oscillators , but their free-running period does not exactly match day length . Thus , they have to be reset by environmental inputs to stay properly phased with the day∶night cycle . The fruit fly Drosophila melanogaster relies primarily on CRYPTOCHROME ( CRY ) —a cell-autonomous blue-light photoreceptor—to synchronize its circadian clocks with the light∶dark cycle . With a genetic screen , we identified over 20 candidate genes that might regulate CRY function . kismet ( kis ) is among them: it encodes a chromatin remodeling factor essential for the development of Drosophila . We show that , in adult flies , KIS is expressed and functions in brain neurons that control daily behavioral rhythms . KIS determines how Drosophila circadian behavior responds to light , but not its free-running period . Moreover , manipulating simultaneously kis and cry activity demonstrates that these two genes interact to control molecular and behavioral circadian photoresponses . Our work therefore reveals that KIS regulates CRY signaling and thus determines how circadian clocks respond to light input . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"neuroscience/behavioral",
"neuroscience",
"genetics",
"and",
"genomics/gene",
"function"
] | 2009 | A Constant Light-Genetic Screen Identifies KISMET as a Regulator of Circadian Photoresponses |
Tsetse flies are the sole vectors of Trypanosoma brucei parasites that cause sleeping sickness . Our knowledge on the early interface between the infective metacyclic forms and the mammalian host skin is currently highly limited . Glossina morsitans flies infected with fluorescently tagged T . brucei parasites were used in this study to initiate natural infections in mice . Metacyclic trypanosomes were found to be highly infectious through the intradermal route in sharp contrast with blood stream form trypanosomes . Parasite emigration from the dermal inoculation site resulted in detectable parasite levels in the draining lymph nodes within 18 hours and in the peripheral blood within 42 h . A subset of parasites remained and actively proliferated in the dermis . By initiating mixed infections with differentially labeled parasites , dermal parasites were unequivocally shown to arise from the initial inoculum and not from a re-invasion from the blood circulation . Scanning electron microscopy demonstrated intricate interactions of these skin-residing parasites with adipocytes in the connective tissue , entanglement by reticular fibers of the periadipocytic baskets and embedment between collagen bundles . Experimental transmission experiments combined with molecular parasite detection in blood fed flies provided evidence that dermal trypanosomes can be acquired from the inoculation site immediately after the initial transmission . High resolution thermographic imaging also revealed that intradermal parasite expansion induces elevated skin surface temperatures . Collectively , the dermis represents a delivery site of the highly infective metacyclic trypanosomes from which the host is systemically colonized and where a proliferative subpopulation remains that is physically constrained by intricate interactions with adipocytes and collagen fibrous structures .
Human African trypanosomiasis , also known as sleeping sickness , is indigenous for the African continent and is caused by two subspecies of Trypanosoma brucei , namely T . brucei rhodesiense and T . brucei gambiense . A range of other trypanosome species including T . congolense , T . vivax and T . b . brucei , is responsible for the majority of infections in livestock and wild animals . The current decline in number of sleeping sickness cases holds promise of moving into the eradication phase for this disease [1 , 2] whereas for animal trypanosomiasis with many problems of reservoir management and drug resistance [3–5] this goal is not yet in sight . In tsetse flies , T . brucei parasites go through a complex developmental cycle in the alimentary tract and salivary glands ending with the cellular differentiation into the metacyclic form that can infect a new mammalian host [6 , 7] . Once this trypanosome population has been established in the salivary glands , it is maintained throughout the entire life span of the tsetse fly . The presence of high densities of trypanosomes in the salivary glands was shown to dramatically affect the tsetse saliva protein composition with a severe reduction of the major anti-hemostatic activities resulting in a hampered blood feeding process [8] . These infection-induced changes in the feeding physiology are believed to favor increased vector/host contact and enhance the chance of parasite transmission to a range of hosts . Only few studies have so far addressed the early parasitological features of a naturally transmitted trypanosome infection in the mammalian host . It is known that variant surface glycoproteins ( VSGs ) play a crucial role in the escape from immune elimination by the host adaptive immune response . Various glycoproteins [e . g . procyclin and BARP ( brucei alanine-rich protein ) ] are displayed on the T . brucei surface in the tsetse fly vector [9] and VSGs are for the first time expressed at the metacyclic stage in the tsetse fly salivary glands . A range of metacyclic VSG antigen types ( M-VATs ) are present on the coats of the trypanosome population in tsetse saliva , which is considered beneficial for an infection establishment in mammals . Together with M-VAT expression , a set of predominant bloodstream VATs appears early during infection in the lymph and in the blood [10 , 11] . Experimental infections with purified bloodstream form ( BSF ) T . brucei and T . congolense revealed the stringent nature of the intradermal route that compromises the ability of these parasites to achieve host infection [12] . Tumor necrosis factor α ( TNF-α ) and inducible nitric oxide synthase ( iNOS ) were shown to contribute to the innate resistance to intradermal infection with low numbers of bloodstream form T . congolense parasites [12] . Addition of saliva to the inoculation mixture was also documented to enhance parasitemia onset following intrapinna BSF T . brucei injection , linked to a local immunosuppressive effect of saliva in the murine dermis [13] . Recently , a peptide with immunosuppressive properties has been identified from G . morsitans saliva with the ability to inhibit LPS induced MAPK signaling in splenocytes [14] . Together , these observations clearly indicate that the natural intradermal inoculation of metacyclic trypanosomes by tsetse flies adds a significant layer of complexity as compared to needle injection of bloodstream form trypanosomes in the peritoneal cavity used routinely in experimental infection models [15] . In humans , bites from infected tsetse flies regularly result in the formation of a skin ulceration or chancre [16] . This skin ulceration has also been documented during early T . b . rhodesiense infections in vervet monkeys [17] and following T . congolense infections in rabbits , goats , calves and sheep [18–21] . Histological and electron microscopy studies have documented the presence of parasites in the skin reactions [20 , 22–24] . Fluorescent trypanosomes or bioluminescent trypanosome versions have been used following needle injection to study tissue tropism and responses to drug treatment in the mammalian host [25–27] and to illustrate parasite genetic exchange in the tsetse fly vector [28] . In this study we have established tsetse mediated intradermal infections ( i . e . a natural infection model ) using fluorescently tagged trypanosomes to assess the kinetics of mammalian host colonization through the intradermal route using a combination of flow cytometry and molecular parasite quantification . This study unequivocally describes that metacyclic trypanosomes are highly infective and that a subpopulation of parasites is retained and actively proliferates in the dermis in close proximity of the initial inoculation site . This parasite retention was studied by scanning electron microscopy documenting a number of intricate interactions between parasites , adipocytes with characteristic periadipocyte collagenic baskets [29] and collagen bundles . These intradermal features and the induced temperature changes are compatible with the re-acquisition of parasites by the tsetse fly from this site . This study is the first to document in detail the presence of a residing intradermal trypanosome population at the tsetse biting site from which the host is systemically infected . These parasites may play a role as an early trypanosome reservoir in the mammalian host and could be picked up by the tsetse vector , demonstrated here for flies that fed soon after the primary inoculation .
Mouse care and experimental procedures were performed under approval of the Animal Ethical Committee of the Institute of Tropical Medicine ( Ethical clearance nr . VPU2014-1 ) and of the Vrije Universiteit Brussel ( Ethical clearance nr . 13-220-1 and 15-220-14 ) . Tsetse fly maintenance and experimental work was approved by the Scientific Institute Public Health department Biosafety and Biotechnology ( SBB 219 . 2007/1410 ) . The experiments , maintenance and care of animals complied with the guidelines of the European Convention for the Protection of Vertebrate Animals used for Experimental and other Scientific Purposes ( CETS n° 123 ) . C57BL/6JRj mice ( Janvier ) were used as animal model for natural trypanosome infections transmitted by tsetse flies . Tsetse flies ( Glossina morsitans morsitans ) were available at the Institute of Tropical Medicine , Antwerp . The flies originated from puparia collected in Handeni ( Tanzania ) and Kariba ( Zimbabwe ) . The tsetse fly colony is maintained on an in vitro bovine blood feeding system at 26°C and a relative humidity of 65%-75% . Three Trypanosoma brucei stocks were used for infection of tsetse flies and mice , the AnTAR1 strain and two transgenic lines of the pleiomorphic AnTat1 . 1E . T . b . brucei AnTAR1 is a post-tsetse fly strain derived from the EATRO 1125 stabilate that was originally isolated from a bushbuck in Uganda in 1966 and from which the AnTat1 . 1E strain was derived [30] . Two different transgenes were integrated in the tubulin array of AnTat1 . 1E using the constitutive trypanosomal expression vector pHD309: the genes encoding red fluorescent protein DsRed of a coral of the Discosoma genus ( DsRED , Clontech ) and a mutant of the Aequorea macrodactyla GFP-like protein ( TagGFP2 , Evrogen ) . Transfection and selection with 10 μg/ml hygromycine were performed as described in [25] . Epifluorescence microscopy was used to select the most fluorescent clones . The in vitro growth curves of the wild-type and AnTat1 . 1EDsRED or AnTat1 . 1ETagGFP2 stocks were generated by seeding cells at 1 x 104 cells ml−1 in 500 μl of HMI-9 with 15% fetal bovine serum in three replicate wells that were counted every 24 h . Teneral tsetse flies were fed 24–48 hours after emergence with a T . b . brucei AnTAR1 , AnTat1 . 1EDsRED or AnTat1 . 1ETagGFP2 infected blood meal supplemented with 10 mM reduced L-glutathione to increase trypanosome infection establishment [31] . Parasitized blood was harvested with heparin from cyclophosphamide-immune suppressed mice ( Endoxan , Baxter ) at 6–7 days post-infection and mixed with defibrinated horse blood ( E&O Laboratories Limited ) to obtain > 106 BSF trypanosomes/ml in the initial blood meal . After this trypanosome-infected blood meal , flies were fed every 2–3 days on uninfected defibrinated horse blood . Salivary gland infected flies were selected by induced probing on pre-warmed glass slides that were microscopically examined for the presence of metacyclic trypanosomes . Flies with a mature , metacyclic salivary gland infection ( SG+ ) were selected for infecting mice . Some SG+ flies were dissected to isolate the salivary glands and to evaluate by flow cytometry the spontaneous parasitic outflow from non-disrupted glands in phosphate saline glucose buffer ( PSG; PBS pH 7 . 4 supplemented with 1% glucose ) . Mice were anesthetized by ketamine ( 100 mg/kg ) and xylazine ( 20 mg/kg ) and infected with T . b . brucei AnTAR1 , AnTat1 . 1EDsRED or AnTat1 . 1ETagGFP2 parasites by allowing individual infected tsetse flies ( 1 fly for each mouse ) to probe on the mouse ear pinnae . In order to assess the parasitemia development in the peripheral blood , 1:200 diluted tail vein blood samples were analysed using improved Neubauer counting chambers and Uriglass disposable slides ( Menarini Diagnostics ) for parasitema levels < 107/mL . When using fluorescently tagged trypanosomes , parasite appearance in the bloodstream was also monitored by flow cytometry using the FACSVerse and BD FACSsuite software . At regular time points after infection ( 4 . 5 , 18 , 42 , 66 and 90 hpi ) , mice were euthanized and blood , cervical and mandibular lymph nodes , and ears were collected . Tissues were either preserved in RNAlater according to the manufacturer’s recommendations for RT-qPCR , fixed in 2 . 5% EM-grade glutaraldehyde 0 . 1M cacodylate pH7 . 2 for scanning electron microscopy or collected in phenol red-free RPMI 1640 medium ( Life Technologies ) for further processing for flow cytometry analysis . Blood samples collected by a cardiac puncture were subjected to erythrocyte lysis ( 1:10 addition of 8% ammonium chloride , 0 . 84% sodium bicarbonate 1mM of EDTA pH 7 . 3 ) . Lysis was conducted for 10 minutes on ice followed by centrifugation at 870 × g for 8 minutes at 4°C . The pellet was rinsed with RPMI and centrifuged again at 870 × g at 4°C for 8 minutes . The pellet was resuspended in RNAlater , kept at 4°C overnight and , after removal of RNAlater , snap-frozen in liquid nitrogen and stored at -80°C . Harvested ears and cervical and mandibular lymph nodes were incubated overnight in RNAlater at 4°C , snap-frozen and stored at -80°C . Various doses of purified bloodstream form ( BSF ) and metacyclic form ( MCF ) trypanosomes were used for experimental infection experiments using an intradermal injection method . BSF parasites were expanded in C57Bl/6JRj mice and purified from their heparinized blood , using diethylaminoethyl cellulose ( DEAE52 ) anion exchange chromatography . Parasites were collected in PSG buffer ( PBS pH 7 . 4 supplemented with 1% glucose ) , centrifuged at 870 × g and resuspended just prior to injection in sterile LPS-free PBS and kept on ice . MCF parasites were collected as the outflow from SG+ glands , washed to remove tsetse saliva components and resuspended in PBS . To exclude differences as a result of the anion exchange procedure , MCF parasites were also subjected to an anion exchange chromatography in one experiment . Motile/viable parasites were counted using the Uriglass disposable slides and live parasite concentration adjusted in order to initiate experimental trypanosome infections intradermally with various parasite doses . In one experiment , viability of DEAE52 purified MCF and BSF parasite was verified by flow cytometry using 7AAD ( Via-Probe , BD ) . The inoculation consisted of a single injection of 20 μl buffer containing 500 , 200 , 100 , 50 , 20 or 5 BSF or MCF trypanosomes between the ventral and dorsal ear dermal layer using a 30 gauge insulin microsyringe . High resolution thermographic imaging of mice exposed to a T . b . brucei AnTAR1 infective tsetse bite ( on the left mouse ear pinnae ) was achieved using the Testo 890 thermal camera equipped with a 42° lens . Acquisition was at a 640 × 480 resolution with a thermal sensitivity ( NETD ) of 40 mK and an instantaneous field of view ( IFOV ) of 0 . 11 mm . The emissivity was fixed at 0 . 95 . The reflected temperature ( RTC ) was set to 20°C . Images were analysed using the IRsoft software provided by the camera manufacturer . Surface temperatures of dorsum and ear extremities were recorded to evaluate the effect of infection on the total body temperature and to assess the impact of dermal ( left ear ) parasite presence on local skin temperature . The potential effect of parasite-induced thermal changes in the skin on tsetse feeding was simulated by an in vitro feeding system . Tsetse flies were fed in obscurity on defibrinated horse blood through a silicone membrane equilibrated at either 25 . 1°C or 24 . 0°C corresponding to the measured average surface temperatures of respectively a heavily parasitized ear dermal site and the contralateral control site . The effect of this thermal difference was also evaluated at higher temperatures , i . e . by comparing the feeding response at 29 . 0°C and 30 . 2°C . In order to evaluate the availability of skin-residing trypanosomes for acquisition by tsetse flies , a total of 6 mice in two independent experiments were exposed on the venter to the bites of 7 T . b . brucei AnTAR1 SG+ infected flies . This primary exposure results in the deposition of a skin-residing parasite population . Eighteen hours after this primary exposure , mice were challenged with a new set of flies on the venter ( n = 15/mouse ) and the dorsum ( opposite to primary exposure , n = 15/mouse ) . For each mouse and exposure side , 10 flies that engorged a full blood meal were dissected 1 hour after the blood feeding and abdomens were collected in 50 μl PSG . Tissues were next homogenized in Trizol reagent ( Invitrogen ) and stored at -80°C until further processing . In order to evaluate that the 18 hpi dermal parasites can establish infections in tsetse flies , sets of 40 teneral male tsetse flies were allowed to feed on dorsal and ventral sides . Seven days later , tsetse flies were dissected to microscopically determine the midgut infection status . Blood samples for flow cytometry analysis were obtained from the tail vein and diluted 1/100 in RPMI 1640 without phenol red ( Life Technologies ) supplemented with penicillin ( 10U/μl ) , streptomycin ( 10U/μl ) , L-glutamine , 10 μg/ml anti-Fc antibody and 25 U/ml heparin . Dissected lymph nodes and ears were placed in a petri dish with 2 ml RPMI 1640 without phenol red ( Life Technologies ) and with penicillin , streptomycin and L-glutamine , 100μg/ml Liberase TL ( Roche ) and 50μg/ml DNase ( Sigma ) similar to the procedure described elsewhere [32] . The ears were separated in ventral and dorsal dermal sheets . The different tissues were crushed using a syringe plunger and the samples were incubated at 37°C , ears for one hour and lymph nodes for thirty minutes to maximally liberate cells from these tissues . After incubation , ice cold PBS supplemented with 2mM EDTA and 10% of fetal calf serum was added to stop the enzymatic digestion . This solution with the tissue debris was filtered through a 70 μm cell strainer ( BD ) and cells were collected by centrifugation at 870 × g for 8 minutes at 4°C . Cells were resuspended in phenol red-free RPMI 1640 with 10 μg/ml anti-Fc antibody and 25 U/ml heparin and stained with 7-amino-actinomycin D ( 7AAD , BD Via-Probe Cell Viability Solution ) and anti-mouse CD45 APC-Cy7 ( clone 30-F11 , eBioscience ) for exclusion of potentially autofluorescent white blood cells . At least 2x105 events were acquired for each sample in a volumetric analysis on the FACSVerse flow cytometer ( BD ) and analysed with BD FACSSuite Software . To identify T . b . brucei AnTat1 . 1EDsRED and AnTat1 . 1ETagGFP2 parasites , a parasite gate was set in the FSC/SSC density plot based on the cell characteristics of an in vitro trypanosome culture . CD45+ and 7AAD+ ( dead ) cells were excluded . Parasites were also visualized in saliva deposits on glass slides by conventional light and epifluorescence microscopy . Confocal microscopy in mounted ear dermal sheets was performed by using the Zeiss LSM700 microscope equipped with the ECPlan-Neofluar 40×/1 . 30 Oil DIC M27 objective lens . Videos of 3 s were obtained by combining 10 consecutive images made in a 14 . 1 s time window . Glutaraldehyde-fixed ear samples were embedded in 1% agarose and cut into 500 μm sections with a Vibratome ( Leica ) . Sections were then incubated overnight at 4°C in 2 . 5% glutaraldehyde , 0 . 1 M cacodylate buffer ( pH 7 . 2 ) , and post-fixed in 2% OsO4 in the same buffer . After serial dehydration , samples were dried at critical point and coated with platinum by standard procedures . Observations were made in a Tecnai FEG ESEM QUANTA 200 ( FEI ) and images processed by the SIS iTEM software ( Olympus ) . RNAlater-treated samples stored at -80°C were thawed and weighted . Ears ( separated in ventral and dorsal dermal sheets ) and lymph node tissue samples were transferred into lysing matrix D homogenization vessels ( MP Biomedicals LLC ) , containing Lysis/Binding Solution ( 10–12 μL/mg tissue ) of the RNAqueous Kit ( Ambion ) followed by homogenization with FastPrep ( MP Biomedicals LLC ) . This lysing matrix was identified as the most appropriate based on a pilot experiment that compared lysing matrixes A , D , F and S . Ear samples were homogenized twice at 6 . 0 m/s for 40 seconds and lymph node samples once . After homogenization , the lysates were centrifuged at top speed ( 21130 × g ) for 2 minutes in order to remove tissue debris that may be present in the lysate . RNA isolation was carried out with the RNAqueous Kit ( Ambion , Life technologies ) according to the manufacturer’s recommendations . Potential DNA contamination was enzymatically removed with 2U RNAse-free DNase I ( Ambion ) followed by inactivation using the DNase inactivation reagent ( Ambion ) and heat treatment at 75°C for 10 minutes . RNA from the tsetse fly abdomens was extracted using the Trizol reagent followed by 2 consecutive chloroform extractions , RNA precipitation in the presence of Glycoblue and two 75% ethanol washes of the RNA pellet . RNA was resuspended in 20 μl DEPC water ( Ambion ) . Samples were subjected to DNAse I treatment ( Biolabs ) for 10 minutes at 37°C . The reaction was stopped by the addition of 5 mM EDTA and heat inactivation for 10 minutes at 75°C . The concentration and purity of the RNA extracts was measured by using a NanoDrop spectrophotometer ( Isogen , Life Science ND-1000 ) . Absorbance ratios at 260/280nm and 260/230nm were determined as measures for RNA-purity . Samples were stored at -80°C until further use . Parasite presence in the various tissues was quantified using the spliced-leader ( SL ) RNA specific RT-qPCR as described elsewhere [33] . Briefly , RNA was reverse transcribed and cDNA amplified using the SensiFAST SYBR No-ROX One-Step Kit ( Bioline ) in a reaction volume of 20 μl with 1x SensiFast containing 400 nM of reverse and forward primers ( 5’-CAATATAGTACAGAAACTG-3’ and 5’-AACTAACGCTATTATTAGAA-3’ ) , 0 . 2 μL Reverse Transcriptase and 0 . 4 μL RNase inhibitor . Thermal cycling conditions were 10 minutes incubation at 45°C , 2 minutes at 95°C , followed by 40 cycles of 95°C for 5 seconds , 50°C for 10 seconds , 60°C for 5 seconds . Post-amplification melting curves were recorded from 45°C to 95°C , with increments of 0 . 1°C and continuous acquisition . Cp values were obtained from the amplification curves by the second derivative approach and Tm calling was used to confirm the presence of a single amplicon . Flow cytometry analyses were performed in a volumetric mode , allowing for the accurate calculation of the parasite concentrations in the sample and the actual number of recovered parasites from the tissues using the BD FACSSuite Software . RT-qPCR based calculation of parasite presence in the tissues was based on linear regression of Cp values obtained for a standard curve ( containing 2×104 , 104 , 5×103 , 2 . 5×103 , 103 , 5×102 , 102 , 10 , 1 parasites ) . In vitro parasite doubling times ( Td ) and median infectious parasite doses were calculated by using non-linear regression fitted with an exponential equation in GraphPad Prism 6 . 0 . The same software was used for preparing the graphs and for the statistical analysis ( two-tailed unpaired Mann-Whitney-t test , one-way ANOVA ) of the data . Data were represented as means ± standard error of the mean . P values ≤ 0 . 05 were considered to be statistically significant .
Lines of T . b . brucei AnTat1 . 1E transfected to express the dsred or taggfp2 transgene were used to infect G . m . morsitans tsetse flies . These two fluorescent strains displayed the same in vitro doubling times as the wildtype ( respectively 7 . 0 ± 0 . 5 h and 7 . 5 ± 0 . 2 h versus 6 . 9 ± 0 . 5 h ) . In comparison with the T . b . brucei AnTAR1 parental strain , lower salivary gland infection rates were recorded in flies infected with the T . b . brucei AnTat1 . 1EDsRED and AnTat1 . 1ETagGFP2 strains ( S1 Table ) . This was linked to a lower maturation index of the transgenic parasites in the fly given that midgut infection rates following feeding on a reduced L-glutathione supplemented parasitized blood meal exceeded 95% for all tested strains . Following salivary gland colonization ( SG+ ) , it was observed that parasite densities in the saliva deposits on pre-warmed glass slides were higher in the T . b . brucei AnTAR1 as compared to the fluorescently tagged T . b . brucei AnTat1 . 1E-infected flies ( Fig 1A ) . Parasite outflows from individually dissected salivary gland pairs were volumetrically analysed by flow cytometry ( Fig 1B ) and shown to contain about 8-fold less metacyclic parasites ( p < 0 . 05 ) as compared to AnTAR1 SG+ flies ( Fig 1C ) . Fluorescently tagged trypanosomes were used for analysing parasite kinetics from the infection initiation site following an infective tsetse fly bite . In order to simulate a natural transmission in a mouse model , T . b . brucei AnTat1 . 1EDsRED infected G . m . morsitans flies were used to initiate intradermal trypanosome infections in mouse ear pinna followed by analysis of host colonization by the parasites . Early kinetics of the emigration of fluorescently labeled parasites from the dermal inoculation site to the lymphoid tissue and blood stream were examined by flow cytometry and confocal microscopy . Fluorescently tagged Trypanosoma parasites were detected in the ear dermis at the various time points analysed in this setup ( 4 . 5 , 18 and 66hpi; Fig 2A ) . Parasite presence was also quantified by an SL-RNA specific RT-qPCR ( Fig 2B ) confirming the presence and expansion of parasites in the ear dermis during the early course of infection ( Fig 2B ) . Analysis of the kinetics of trypanosome emigration revealed that parasites could be detected by 18hpi in the lymph nodes draining the dermal site of infection ( Fig 2A ) . Then , around 42–44 hpi , parasites were detected in the peripheral blood using flow cytometry and qPCR ( Fig 2A and 2B ) . In order to compare parasitemia onset following natural inoculation of different parasite doses , infected flies with a high or a low parasite density in the saliva ( i . e . infected with either the T . b . brucei AnTAR1 or the AnTat1 . 1EdsRED strain ) were used to initiate intradermal infections in the ears of mice followed by conventional parasitological detection in the peripheral blood . This suggested that infection with a lower parasite dose resulted in a delayed appearance of parasites in the blood circulation , whereas peak parasitemia levels were not significantly altered ( Fig 3A ) . Excluding that this observation results from intrinsic differences between the wildtype and dsRed transgenic strain , experimental needle inoculation of various doses of the T . b . brucei AnTAR1 metacyclic parasites confirmed this inverse correlation between inoculated parasite dose and parasitemia onset in the peripheral blood ( Fig 3B ) . In addition , metacyclic ( MCF ) trypanosomes showed to be highly capable of host colonization through this intradermal route in contrast to purified bloodstream forms ( BSF , Fig 3C versus 3D ) . Indeed , for the intradermal needle injection of metacyclic parasites , 7 parasites was calculated as the 50% infectious dose with doses as little as 5 metacyclic trypanosomes often resulting in no establishment of infection under the used experimental conditions . Purification of metacyclic trypanosomes by anion exchange chromatography was found not to reduce the infectivity ( Fig 3C , red squares ) . A dose as high as 200 bloodstream form parasites was not infective through the same intradermal route . A cell viability assay was performed on the prepared DEAE52-purified BSF and MCF inoculums , revealing very comparable viability ( ≥ 95% ) just prior to injection . Assessment of post-injection viability approximately 3 hours after preparing the samples revealed a slightly elevated cell death in the BSF ( 16% ) as compared to the MCF sample ( 5% ) ( S2 Fig ) . Some variation was observed in the independent BSF infection experiments ( Fig 3D ) , but it can be stated that the critical threshold for infection of mice with DEAE52-purified T . b . brucei AnTAR1 is situated around 200–300 BSF parasites in contrast to a > 10-fold lower critical threshold observed for the MCF parasites . Intradermal inoculation of T . b . brucei AnTat1 . 1EDsRED or AnTat1 . 1ETagGFP2 by the bites of SG+ tsetse flies resulted in a trypanosome population residing inside the dermis ( left ear ) whereas this population could not be detected at the contralateral side of the mouse body ( right ear ) ( S1 Fig ) . Scanning electron microscopic analysis of vibratome ear sections revealed the entanglement of skin residing parasites through different apparent mechanisms including ( i ) intricate interactions with adipocytes that are present in the connective tissue in close proximity to the cartilage layer of the ear ( Fig 4A–4G ) ( ii ) entanglement by reticular fibers ( Rf ) of the periadipocytic baskets ( Fig 4E–4I ) and those in the interstitial spaces ( Fig 4J–4K ) and ( iii ) embedding in collagen bundles ( Cb , Fig 4J–4M ) . Interaction with the adipocytes mainly involved burying of the anterior part of the parasites into the fat cells , leaving the flagellar pocket accessible for nutrient uptake ( Fig 4D–4G ) . Consistent with these various modes of entanglement , trypanosomes were characterized by a restricted anterior-posterior movement inside the dermis as evidenced by a confocal microscopy analysis inside mounted ear dermal sheets ( S1 Video ) . Despite these structural interaction features , numerous parasites were observed to be proliferative as indicated by the presence of a double flagellum ( Fig 4H–4K ) . Trypanosomes with a double flagellum were also observed on the surface of adipocytes which strongly suggests parasite multiplication ( Fig 4H–4I ) . Moreover , evaluation of the viability by 7-amino-actinomycin D ( 7AAD ) staining demonstrated very limited numbers ( < 1% ) of non-viable trypanosomes during the expansion phase in the dermis indicating that the observed interactions do not have a detrimental impact on the parasite . The absence of significant numbers of non-viable parasites in the dermal trypanosome population is also demonstrated in the confocal microscopy video ( S1 Video ) . In order to evaluate whether the expanding parasite population is derived from the initial inoculum or from parasites of the peripheral blood that re-colonized the dermal site , mice were naturally infected with two differentially labelled parasites ( respectively the dsRed and TagGFP2 expressing T . b . brucei AnTat1 . 1E ) on remote sites of the mouse body ( left ear pinna and venter , Fig 5A ) . This infection resulted in a mixed trypanosome population in the peripheral blood ( Fig 5B ) whereas the expanding population reaching > 105 at the ear inoculation site expresses a single fluorescent marker ( dsRed ) corresponding to the initial inoculum . No expanding parasite population could be detected at the contralateral side of the body ( right ear pinna ) . The dermal parasite population seems to reduce to low numbers of remaining dsRed-expressing T . b . brucei AnTat1 . 1E detected by flow cytometry at 18 dpi despite parasite presence in the peripheral blood ( S3 Fig ) . Following exposure of the left ear pinna to SG+ tsetse fly bites , thermal changes in the infected mouse were measured at various time points ( 4 , 8 and 11 days post-infection ) with an emphasis on the differences between the temperatures of the inoculated left ears and those of the non-exposed right ears ( Fig 6A and 6B ) . An overall lethargy with significant hypothermia was observed from 8 dpi with a more than 3°C reduction in recorded surface body temperatures by 11 dpi as compared to non-infected littermates ( p < 0 . 0001 ) . Parasite exposed ear pinnae were characterized by a significantly increased temperature ( 25 . 1 ± 0 . 1°C ) relative to the non-exposed ear ( 24 . 0 ± 0 . 2°C ) with an average elevation of 1 . 1 ± 0 . 2°C at 8 dpi ( p < 0 . 0001; Fig 6B ) . This difference in temperature over the course of infection closely correlated with the ear dermal parasite burden as determined by SL-RNA specific RT-qPCR in the same mice used for the thermographic analyses ( Fig 6C ) . To evaluate whether the thermal difference could potentially be of any physiological importance for the tsetse fly feeding responses , 3-day starved flies were offered an artificial horse blood meal through a silicone membrane with surface temperatures representing the mean left and right ear temperatures measured at 8 dpi ( 25 . 1°C versus 24 . 0°C respectively ) . Significantly more flies ( 59 . 1% , n = 176 ) obtained a blood meal from the 25 . 1°C surface compared to only 39 . 9% ( n = 176 ) at 24 . 0°C ( p = 0 . 0003 , S2 Table ) . The tsetse engorgement rate was not significantly altered upon temperature increase from 29 . 0°C ( 84 . 3% , n = 70 ) to 30 . 2°C ( 79 . 7% , n = 68 ) . This suggests that the increased temperature induced by the dermis-residing trypanosome infection could represent an attractive cue for tsetse fly feeding in a specific , low temperature range . In order to assess whether this skin residing parasite population can be acquired by these feeding tsetse flies , freshly emerged tsetse flies were fed on the ventral and dorsal side of mice that were prior exposed at the ventral side to the bites of multiple T . b . brucei SG+ tsetse flies , 18 hours earlier ( Fig 7A ) . After feeding , ingested parasites were quantified through SL-RNA RT-qPCR on RNA extracted from the abdomens of flies that engorged a full blood meal . On average , higher parasite numbers were recovered from tsetse flies that were fed on the ventral side which corresponds to the primary T . b . brucei SG+ exposure site ( 75 parasites versus 39 from the dorsal side , p < 0 . 0001; Fig 7B and 7C ) . Parasite acquisition by tsetse from the non-exposed dorsal side indicated that exposure to the multiple tsetse fly bites at the ventral side already resulted in parasite presence in the peripheral blood within 18h after exposure . In two independent experiments using a total of 160 flies , blood fed flies were evaluated for the establishment of a midgut trypanosome infection by dissection and microscopic evaluation at 7 days post feeding . We could not detect the establishment of infections in the tsetse fly midgut following uptake of parasite populations from the dorsal or ventral side at this early time point of infection .
A significant body of research on African trypanosomes is conducted by making use of needle injection of bloodstream parasites in the peritoneal cavity of the murine host . Although many breakthroughs in understanding the biology and immunology of trypanosomes have been made through such experimental models [15] , very few information is yet available on the parasitological features of naturally transmitted trypanosome infections . In this study we have used T . b . brucei AnTAR1 , AnTat1 . 1EDsRED and AnTat1 . 1ETagGFP2 for natural transmission studies using the savannah tsetse fly , Glossina morsitans that represents a proficient vector for T . brucei . Although these parasites display the same in vitro replication times , the two fluorescent transgenic strains showed lower metacyclic infection frequencies and lower parasite densities in the salivary glands . This may be explained by a non-neutral characteristic of the dsred and taggfp2 transgenes or effects related to the integration site ( tubulin array ) resulting in a less virulent phenotype in the fly . Exploiting these differences between the dsRed-tagged AnTat1 . 1E and AnTAR1 strain in terms of numbers of parasites in the tsetse salivary glands , this study was able to compare the early parasitemia progression following a naturally inoculated low and an average 8-fold higher parasite dose , respectively . Combination with experimental needle injections of varying doses of the same parasite strain ( AnTAR1 ) , to exclude intrinsic strain differences in the host colonization process , revealed that parasite dose largely determines the time of appearance in the peripheral blood , without affecting peak parasitemia levels . A study by Wei et al has shown that the intradermal route of T . brucei and T . congolense parasite transmission differs considerably from the experimental peritoneal route and constitutes a stringent bottleneck for infection establishment [12] . These observations were made following intradermal inoculation of purified blood stream form trypanosomes . In this study , these experimental conditions were replicated for the infectious metacyclic trypanosomes purified from the tsetse fly salivary glands , demonstrating that the natural intradermal transmission route does not impose a significant constraint for infection establishment by these salivary gland derived forms . The 50% infectious doses were calculated to be 7 metacyclic ( MCF ) trypanosomes whereas even 200 bloodstream ( BSF ) trypanosomes were unable to establish an infection upon intrapinna injection in mice . This clearly suggests that the metacyclic forms of T . brucei are far better prepared than bloodstream forms to survive and develop at the intradermal inoculation site in the mammalian host . The infective dose of metacyclic trypanosomes in mice that we determined in our study is > 50-fold lower than the one reported for a T . brucei rhodesiense strain in humans . In this human infection study , it was concluded that an average man would require a subcutaneous dose of 300–450 metacyclic parasites to develop a systemic infection . However , a high variability was observed with a lowest infective dose of 170 and the highest non-infective dose of 1067 parasites [35] . All these experimental observations suggest that the MCF infective dose delivered through tsetse intradermal inoculation is variable and host/parasite-dependent . Metacyclic parasites are known to be in a quiescent non-proliferative stage . Recent comparisons between MCF trypanosomes in the tsetse salivary gland and BSF trypanosomes indicated some transcriptional differences between these different life cycle stages including metabolic genes ( glycolysis , phosphorus metabolism ) , surface molecules , transporters ( ion , glucose , amino acid and nucleoside transporters ) , transcriptional regulators and translation machinery components [36] but it is not clear what could be the basis for the higher infectivity of metacyclic trypanosomes . The higher infectivity of metacyclic trypanosomes did not relate to the previously described immunomodulatory activity of soluble salivary factors [13] given that parasites were washed prior to inoculation . As TNF and iNOS were found implicated in determining the dermal infection bottleneck for T . congolense [12] , susceptibility of metacyclic T . brucei parasites to locally produced levels of these effectors would be of interest for further exploration . Interestingly , an ultrastructural study suggested some specific characteristics of the intracutaneous forms of T . brucei with smaller mitochondria and rough endoplasmatic reticulum as compared to the MCF and BSF trypanosomes [20] suggesting a specific trypanosome subpopulation ( stage ) at this site . This is strengthened by our clear observations that a parasite subpopulation remained and proliferated in proximity of the initial inoculation site following a tsetse-mediated infection . The presence of significant numbers of parasites has been previously documented in local skin reactions induced by various trypanosome species ( T . brucei and T . congolense ) [20] . Tsetse fly mediated T . b . rhodesiense infections induced chancres in > 50% of vervet monkeys within 4 to 8 days post-infection [17] . In goats , the inoculation of a single metacyclic T . brucei parasite was reported to be sufficient for the typical ulceration [19] . Cannulation of the lymphatics in goats revealed that parasites can be detected in the lymph within 1–2 days after the infective tsetse fly bite with the formation of a chancre at the infection initiation site by the third day [10] . Our studies in the murine model confirm this rapid migration to the lymph , with detectable levels of parasites in the draining lymph nodes within 18 hours post-infection . Despite this rapid migration to the lymph , a proportion of trypanosomes was found to remain and proliferate in the mouse ear dermis where parasite levels larger than 105 were detected by SL-RNA RT-qPCR and flow cytometry . Using mixed infections with differentially labeled trypanosomes , the expanding intradermal trypanosome population ( expressing the dsRed transgene ) was unequivocally shown not to result from a re-invasion from the bloodstream . The kinetics of differentiation of the inoculated metacyclic into blood stream trypanosomes in the skin remains to be further explored but will require the identification of suitable surface exposed discriminative markers . A transmission electron microscopy ( TEM ) study in goat skin chancres between 3 and 11 dpi demonstrated a constantly high percentage ( 50–75% ) of aberrant trypanosome forms with strong vacuolization which were therefore considered to be degenerating and non-dividing forms [20] . Another study in sheep [24] reported the same cell death phenomenon for T . congolense in the dermis coinciding with the chancre formation . However , a study in New Zealand White rabbits only observed parasite degeneration by 11 dpi [22] . Our cell viability staining ( 7AAD-based ) of fluorescently tagged intradermal T . brucei trypanosomes in naturally infected mice did not reveal significant cell death in the parasite population during the early progressive phase of parasite expansion as determined up to 7 dpi . These apparent host differences might correlate to the fact that beyond a certain degree of edema no chancres were found to develop in this mouse intrapinna model . RNA-based parasite quantification demonstrated an abrupt reduction in dermal parasite numbers by 12 dpi corresponding to the observations made in rabbits and was further confirmed by a strong reduction in the dermal parasite population as observed by flow cytometry at 18 dpi . Confocal microscopy analyses on separated ear sheets illustrated a clear motility of the bulk of parasites during the expansion phase ( see S1 Video ) . Scanning electron microscopy ( SEM ) on ear tissue vibratome sections confirmed the proliferation of the dermal parasites as exemplified by the presence of double flagella . A previous study also described the occurrence of proliferative as well as “giant” forms of T . brucei with multiple nuclei and several axonemes per flagellum [20] . Interestingly , our SEM data additionally illustrate intricate interactions of the anterior end of trypanosomes with adipocytes . This interaction seems to be very tight as indicated by the twisting and folding of the flagellar membrane ( Fig 4G ) . Intriguingly , the entanglement did not implicate the posterior end , leaving free the flagellar pocket , the site for endocytosis . Similar interactions have also been observed in SEM analyses of T . cruzi cocultures with in vitro cultured adipocytes [37] . Of importance , T . cruzi infection was shown to influence the secretion by adipocytes of insulin-response regulatory adipokines [37] and skin adipocytes also represent a source of antimicrobial peptides during bacterial infections [38] . Possibly , these interactions might provide an advantage to the parasite in terms of metabolic requirements or exposure to cellular or humoral components of the immune system . In view of the observed intricate interactions with T . brucei parasites , adipocytes remain an understudied cell type in the regulation of metabolism and immunity during African trypanosomiasis . In this context , recent findings demonstrated that adipose tissue is a functional parasite reservoir for T . brucei during systemic development in the mammalian host where parasite gene expression is remodeled to metabolically adapt to the utilization of lipids as a carbon source [39] . Collagen fibers of the periadipocyte baskets acted synergistically in supporting the interaction of trypanosomes with these skin adipocytes . Parasite cell bodies were discovered in several images to be heavily entangled by collagen microfibers while other parasites were found embedded between collagen bundles in the connective tissue . Although the presence of trypanosomes in close proximity to the collagen has been described in early histology studies [23] and TEM studies [22 , 24] , the degree of entanglement observed in the present SEM images had not yet been fully appreciated . It is known that extracellular matrix proteins are used by pathogens for adhesion and invasion ( reviewed in [40] ) . T . brucei was found to secrete an active prolyl oligopeptidase that cleaves collagen and thereby might regulate its interactions with collagen in the extracellular matrix [41] . The interactions with collagen and adipocytes could be responsible for retention of parasites in the dermis but could also play an important role in creating a physiologically or immunologically privileged site that is beneficial for the inoculated parasites allowing them to proliferate and establish a locally adapted subpopulation as suggested by our study . The observation that in vitro culture of T . brucei was more efficient in the presence of collagen-producing fibroblasts [42] is strengthening this hypothesis . In these cultures aberrant multinucleate and giant forms similar to those described inside the chancre [20] were also observed upon surpassing the maximal growth phase . The support of T . b . gambiense culture by primary murine bone marrow derived cells was found to require the indispensable presence of adipocyte clusters [43] , which also points to a positive influence of adipocytes and collagen on parasite expansion . High resolution thermographic imaging revealed significant changes in the skin surface temperatures correlating with the intradermal parasite burden . We have interrogated whether the recorded temperature increase could represent a thermal cue for tsetse flies in favour of parasite acquisition . Unexpectedly , nearly 60% of 3-day starved tsetse flies acquired a blood meal when the surface temperature of the artificial membrane was equilibrated at 25 . 1°C corresponding with the ear temperature at a high parasite burden . In contrast , less than 40% of flies blood fed at a temperature ( 24 . 0°C ) corresponding to the non-inflamed ear tissue . Previous electrophysiological studies demonstrated that the firing frequency of thermosensitive cells on the tarsi of tsetse flies increased linearly in a temperature range from 24°C to 34°C which could underlie the observed difference in feeding response [44] . At a temperature around 30°C , around 80% of the flies acquired a blood meal , but no impact of a 1 . 2°C difference ( between 29 . 0°C and 30 . 2°C ) could be noted in our feeding experiments . At temperatures from 37°C up to 42°C , tsetse flies were documented to acquire blood meals more efficiently with shorter mean feeding times at elevated temperatures [45] . Collectively , the trypanosome carrying dermal site with an elevated local temperature will be a good location for tsetse fly blood feeding in a recently infected mammalian host . Indeed , once landed on the host skin , local temperature sensed by antennal and tarsal thermosensors will be an important determinant for tsetse probing activity [44 , 46] . Tsetse flies also seem to have preferential landing sites and feed at concentrated feeding sites , particularly on the lower front legs of cattle [47] , which therefore represents a site prone to trypanosome deposition and acquisition . Our results indicate that parasite acquisition by tsetse flies can occur during the first time period from the primary exposure site where the infection was initiated although parasites at 18 hpi seemed not yet able to colonize the tsetse midgut . Nevertheless , parasites that are present in the skin could potentially play a role as a transient reservoir for early uptake by tsetse flies soon after their deposition into the skin of animals with low to undetectable parasite concentrations in the peripheral blood . This type of role has been recently described for Leishmania infantum parasites that persisted for several months in lesions at the primary inoculation sites in dogs and from where sand flies could efficiently acquire the parasite [48] . The contribution of the dermal trypanosome population in chronic , low parasitemic models as a reservoir for tsetse fly infection , for recrudescence and its susceptibility to drug treatment remain to be further established . Collectively , this study illustrates the high intradermal infectivity of metacyclic trypanosomes and retention of a parasite subpopulation proximal to the initial inoculation site . The expanding dermal trypanosome population is engaged in a number of remarkable interactions with skin adipocytes and the extracellular matrix . In addition , parasites at this site induced changes in the skin surface temperature . Further unraveling of the basis of the skin-resident trypanosome phenotype will shed new light on processes underlying host colonization and parasite acquisition by the tsetse fly vector . | Sleeping sickness is caused by trypanosomes that are transmitted by the blood feeding tsetse flies . The present study has established an experimental transmission model with fluorescently labeled parasites in mice that allows us to study their fate following natural transmission by a tsetse fly bite . Parasites that arise in the tsetse salivary glands were found to be highly infective following inoculation in the mammalian skin in contrast with previous observations made for trypanosomes purified from the blood stream . This study unveiled that a proportion of parasites is retained in the skin and actively proliferates close to the initial inoculation site resulting in significantly elevated skin temperatures . This retention was linked to interaction with fat cells and collagen fibrous structures . Experimental transmission experiments were able to demonstrate that parasites can be acquired from the inoculation site immediately after the initial transmission . | [
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] | [
"invertebrates",
"medicine",
"and",
"health",
"sciences",
"body",
"fluids",
"pathology",
"and",
"laboratory",
"medicine",
"ears",
"parasitic",
"diseases",
"animals",
"glossina",
"otology",
"ear",
"infections",
"tsetse",
"fly",
"insect",
"vectors",
"digestive",
"system",
"epidemiology",
"exocrine",
"glands",
"head",
"otorhinolaryngology",
"pathogenesis",
"disease",
"vectors",
"insects",
"hematology",
"arthropoda",
"blood",
"anatomy",
"bloodstream",
"infections",
"host-pathogen",
"interactions",
"physiology",
"salivary",
"glands",
"biology",
"and",
"life",
"sciences",
"organisms"
] | 2016 | The Dermis as a Delivery Site of Trypanosoma brucei for Tsetse Flies |
Two observations about the cortex have puzzled neuroscientists for a long time . First , neural responses are highly variable . Second , the level of excitation and inhibition received by each neuron is tightly balanced at all times . Here , we demonstrate that both properties are necessary consequences of neural networks that represent information efficiently in their spikes . We illustrate this insight with spiking networks that represent dynamical variables . Our approach is based on two assumptions: We assume that information about dynamical variables can be read out linearly from neural spike trains , and we assume that neurons only fire a spike if that improves the representation of the dynamical variables . Based on these assumptions , we derive a network of leaky integrate-and-fire neurons that is able to implement arbitrary linear dynamical systems . We show that the membrane voltage of the neurons is equivalent to a prediction error about a common population-level signal . Among other things , our approach allows us to construct an integrator network of spiking neurons that is robust against many perturbations . Most importantly , neural variability in our networks cannot be equated to noise . Despite exhibiting the same single unit properties as widely used population code models ( e . g . tuning curves , Poisson distributed spike trains ) , balanced networks are orders of magnitudes more reliable . Our approach suggests that spikes do matter when considering how the brain computes , and that the reliability of cortical representations could have been strongly underestimated .
Neural systems need to integrate , store , and manipulate sensory information before acting upon it . Various neurophysiological and psychophysical experiments have provided examples of how these feats are accomplished in the brain , from the integration of sensory stimuli to decision-making [1] , from the short-term storage of information [2] to the generation of movement sequences [3] . At the same time , it has been far more difficult to pin down the precise mechanisms underlying these functions . A lot of research on neural mechanisms has focused on studying neural networks in the framework of attractor dynamics [4]–[6] . These models generally assume that the system's state variables are represented by the instantaneous firing rates of neurons . While quite successful in reproducing some features of electrophysiological data , these models have had a hard time reproducing the irregular , Poisson-like statistics of cortical spike trains . A common assumption is that the random nature of spike times is averaged out over larger populations of neurons or longer periods of time [7]–[10] . However , the biophysical sources of noise in individual neurons are insufficient to explain such variability [11]–[13] . Several researchers have therefore suggested that irregular spike timing arises as a consequence of network dynamics [8] , [14] . Indeed , large networks of leaky integrate-and-fire ( LIF ) neurons with balanced excitation and inhibition can be “chaotic” and generate asynchronous and Poisson-like firing statistics [15]–[18] . While these studies explain how relatively deterministic single units can generate similar statistical properties as random spike generators in rate models , they generally do not clarify how particular computations can be carried out , nor do they fundamentally answer why the brain would be operating in such a regime . Here we show that the properties of balanced networks can be derived from a single efficiency principle , which in turn allows us to design balanced networks that perform a wide variety of computations . We start from the assumption that dynamical variables are encoded such that they can be extracted from output spike trains by simple synaptic integration . We then specify a loss function that measures the system's performance with respect to an idealized dynamical system . We prescribe that neurons should only fire a spike if that decreases the loss function . From these assumptions , we derive a recurrent network of LIF neurons that is able to implement any linear dynamical system . We show that neurons in our network track a prediction error in their membrane potential and only fire a spike if that prediction error exceeds a certain value , a form of predictive coding . Our work shows how the ideas of predictive coding with spikes , first laid out within a Bayesian framework [19] , 20 , can be generalized to design spiking neural networks that implement arbitrary linear dynamical systems . Such multivariate dynamical systems are quite powerful and have remained a mainstay of control-engineering for real-world systems [21] . Importantly , the networks maintain a tight balance between the excitatory and inhibitory currents received by each unit , as has been reported at several levels of cortical processing [22]–[26] . The spike trains are asynchronous and irregular . However , this variability is not noise: The neural population essentially acts as a deterministic “super-unit” , tracking the variable with quasi-perfect accuracy while each individual neuron appears to behave stochastically . We illustrate our approach and its usefulness with several biologically relevant examples .
Our basic model strategy is represented in Fig . 1 A . Let us consider a linear dynamical system describing the temporal evolution of a vector of dynamical variables , : ( 1 ) Here is the state transition matrix , and are time-varying , external inputs or command variables . We want to build a neural network composed of neurons , taking initial state and commands as inputs , and reproducing the temporal trajectory of . Specifically , we want to be able to read an estimate of the dynamical variable from the network's spike trains . These output spike trains are given by , where is the time of the spike in neuron . Our first assumption is that the estimate is obtained by a weighted , leaky integration of the spike trains , ( 2 ) where the matrix contains the decoding or output weights of all neurons , and is the read-out's decay rate . Whenever neuron fires , a -function is added to its spike train , . The integration of the respective delta-function contributes a decaying exponential kernel , , weighted by , to each dynamical variable , . This contribution can be interpreted as a simplified postsynaptical potential ( PSP ) . The effect of a neuron's spike can be summarized by its weights , , which we call the output kernel of neuron . Note that these weights correspond to the columns of the matrix . The estimate can also be written as a weighted linear summation of the neuron's firing rates , , if we define the time-varying firing rates of the neurons , , as ( 3 ) Our second assumption is that the network minimizes the distance between and by optimizing over the spike times , and not by changing the fixed output weight matrix . This approach differs from the “liquid computing” approach in which recurrent networks have fixed , random connectivities while the decoding weights are learnt [27] . In our case , the decoding weights are chosen a-priori . In order to track the temporal evolution of as closely as possible , the network minimizes the cumulative mean-squared error between the variable and its estimate , while limiting the cost in spiking . Thus , it minimizes the following cost function , ( 4 ) where denotes the Eucledian distance ( or L2 norm ) , and the Manhattan distance ( or L1 norm ) , which here is simply the sum over all firing rates , i . e . , . Parameters and control the cost-accuracy tradeoff . The linear cost term , controlled by , forces the network to perform the task with as few spikes as possible , whereas the quadratic cost term , controlled by , forces the network to distribute spikes more equally among neurons , as explained in Material and Methods . To derive the network dynamics , we assume that the firing mechanism of the neurons performs a greedy minimization of the cost function . More specifically , neuron fires a spike whenever this results in a decrease of , i . e . , whenever . As explained in Material and Methods , this prescription gives rise to the firing rule ( 5 ) with ( 6 ) ( 7 ) Since is a time-varying variable , whereas is a constant , we identify the former with the -th neuron's membrane potential , and the latter with its firing threshold . In the limit , the membrane potential of the -th neuron can be understood as the projection of the prediction error onto the output kernel . Whenever this projected prediction error exceeds a threshold , a new spike is fired , ensuring that precisely tracks . For finite , the membrane voltage measures a penalized prediction error . If the neuron is already firing at a high rate , only a correspondingly larger error will be able to exceed the threshold and lead to a spike . To connect this firing rule with the desired network dynamics , Eqn . ( 1 ) , we take the derivative of each neuron's membrane potential , Eqn . ( 6 ) , and consider the limit of large networks ( see Material and Methods ) to obtain the differential equation ( 8 ) where is a leak term , is a weight matrix of connectivity filters , explained below , and corresponds to a white “background noise” with unit-variance . The leak-term does not strictly follow from the derivation , but has been included for biological realism . A similar rationale holds for the noise term which we add to capture unavoidable sources of stochasticity in biological neurons due to channel noise , background synaptic input , etc . The differential equation then corresponds to a standard LIF neuron with leak term , external , feedforward synaptic inputs , recurrent synaptic inputs mediated through the weight matrix , and a firing threshold , as specified in Eqn . ( 7 ) . The weight matrix of connectivity filters is defined as ( 9 ) and contains both “fast” and “slow” lateral connections , given by the matrices ( 10 ) ( 11 ) where corresponds to the identity matrix . Accordingly , the connectivity of the network is entirely derived from the output weight matrix , the desired dynamics , and the penalty parameter . Note that the diagonal elements of implement a reset in membrane potential after each spike by . With this self-reset , individual neurons become formally equivalent to LIF neurons . Whereas the linear penalty , , influences only the thresholds of the LIF neurons , the quadratic penalty , , influences both the thresholds , resets , and dynamics of the individual neurons , through its impact on the diagonal elements of the connectivity matrix . Slow and fast lateral connections have typically opposite effects on postsynaptic neurons , and thereby different roles to play . The fast connections , or off-diagonal elements of the matrix , implement a competition among neurons with similar selectivity . If neuron fires , the corresponding decreases in prediction errors ( ) are conveyed to all other neurons . Neurons with similar kernels ( inhibit each other , while neurons with opposite kernels ( ) excite each other . This is schematized by the blue and red connections in Fig . 1 A . In contrast , the slow connections , , implement a cooperation among neurons with similar selectivity . These connections predict the future trajectory of ( term “” ) but also compensate for the loss of information due to the decoder leak ( term “” ) . For example , when the variable is static ( , ) , these connections maintain persistent activity in the network , preventing the variable from decaying back to zero ( see below ) . Note that when the internal dynamics of change on a slower time scale than the decoder ( i . e . ) , and if we neglect the cost term , slow and fast connections have the same profile , ( i . e . ) , but opposite signs . The combined effect of fast and slow connections yields the effective PSPs in our network , , with , which can be obtained by integrating Eqn . ( 8 ) for a single spike . Two example PSPs are shown in Fig . 1 B . We note that our network model may contain neurons that both inhibit and excite different targets , depending on the kernel sign , a violation of Dale's law . This problem can be addressed by creating separate cost functions for excitatory and inhibitory neurons , as laid out in full detail in Text S1 . Here , we simply interpret the resulting connectivity as the effective or functional connectivity of a network , akin to the types of connectivities arising in generalized linear models ( GLMs ) of neural networks [28] . We now briefly consider how the above equations can be mapped onto realistic physical units . This consideration has the additional benefit that it clarifies how the network parameters scale with the number of neurons ( see also Material and Methods ) . In order to express the network dynamics in biophysically relevant units , the membrane potential , Eqn . ( 6 ) , and threshold , Eqn . ( 7 ) , have to be rescaled accordingly . We can obtain proper membrane potential units in mV if we apply the simple transformations and . In turn , we obtain the modified equations ( 12 ) ( 13 ) and the modified dynamics ( 14 ) with a resting potential of . Note that both the feedforward and recurrent connectivities change in this case . Specifically , we obtain and , and a similar expression for the noise , . In turn , we can freely choose and to find realistic units . For instance , we can fix the threshold at , and the reset potential at , which uniquely determines both and for each neuron . In the absence of linear costs ( ) , the reset potential becomes simply . When we increase the network size while keeping the average firing rates and the read-out constant , we need to change the decoding kernels . Specifically , the decoding kernels need to scale with . If we assume that the relative importance of the cost terms is held fixed for each neuron , then the original threshold scales with , and the original connectivities similarly scale with , compare Eqns . ( 9–11 ) . As a consequence , the rescaled synaptic weights , do not scale in size when the network becomes larger or smaller . When considering the summation over the different input spike trains , we therefore see that all synaptic inputs into the network scale with : the feedforward inputs , the slow recurrent input , and the fast recurrent inputs ( the latter two are both contained in the matrix ) . The equal scaling of all inputs maintains the detailed balance of excitation and inhibition in the network . An instructive case is given if we neglect the cost terms for a moment ( ) in which case we obtain the following ( rescaled ) feedforward weights and connectivities: ( 15 ) ( 16 ) Accordingly , the strength of the lateral connections is independent of the kernel norm . In contrast , the strength of the feed-forward connections scales with the inverse of the kernel norm . Since smaller kernels provide a more precise representation , the precision of the rescaled network , and its firing rates , are controlled entirely by its input gain . Once the dynamics and the decoders are chosen , Eqn . ( 1 ) and Eqn . ( 2 ) , the only free parameters of the model are , , , and . The model presented previously can in principle implement any linear dynamical system . We will first illustrate the approach with the simplest linear dynamical system possible , a leaky integration of noisy sensory inputs where can be interpreted as the sensory stimulus while represents shared sensory noise . The corresponding dynamical system , Eqn . ( 1 ) , is then given by ( 17 ) The integrated sensory signal is a scalar ( ) and represents the leak of the sensory integrator . For a completely homogeneous network , in which the output kernels of all neurons are the same , we can solve the equations analytically which is shown in Text S1 . A slightly more interesting case is shown in Fig . 1 C , D , which illustrate network dynamics for two different choices of . Here we used neurons , half of them with positive kernels ( ) , and the other half with negative kernels ( ) . Neurons with positive kernels fire when variable is positive or increases , while neurons with negative kernels fire when the variable is negative or decreases . Moreover , we set the cost terms and at small values , ensuring that our objective function is dominated by the estimation error , compare Eqn . ( 4 ) . As a consequence , the estimate closely tracks the true variable . Albeit small , the cost terms are crucial for generating biologically realistic spike trains . Without them , a single neuron may for example fire at extremely high rates while all others stay completely silent . The regularizing influence of the cost terms is described in more detail in Text S1 . For , the network is a perfect integrator of a noisy sensory signal . The neural activities resemble the firing rates of LIP neurons that integrate sensory information during a slow motion-discrimination task [1] . In the absence of sensory stimulation , the network sustains a constant firing rate ( Fig . 1 C after sec ) , similar to line attractor networks [29]–[31] . In fact , as long as the dynamics of the system are slower than the decoder ( ) , the instantaneous firing rates approximate a ( leaky ) integration of the sensory signals . On the other hand , if the system varies faster than the decoder ( i . e . ) , then neural firing rates track the sensory signal , and model neurons have transient responses at the start or end of sensory stimulation , followed by a decay to a lower sustained rate ( Fig . 1 D ) . These responses are similar to the adaptive and transient responses observed in most sensory areas . The overall effect of the lateral connections depends on the relative time scales of the variable and the decoder ( Fig . 1 B ) . For neurons with similar selectivity ( or equal read-out kernels , ) , the postsynaptic potentials are given by ( assuming ) , ( 18 ) For neurons with opposite read-out kernels , we obtain just a sign reversal . When ( ) , the interplay of fast inhibition with slower excitation results in a bi-phasic interaction between neurons of similar selectivity ( Fig . 1 B , left ) . Moreover , the network activity persists after the disappearance of the stimulus . In the extreme case of the perfect integrator ( ) , the temporal integral of this PSP is exactly zero , which explains how the mean network activity can remain perfectly stable ( neither increase nor decrease ) in the absence of any sensory stimulation . In contrast , lateral interactions are entirely inhibitory when the network tracks the stimulus on a faster time scale than the decoder ( i . e . , Fig . 1 B , right ) . The dominance of lateral inhibition explains the transient nature of the network responses and the absence of persistent activity . Other response properties of the model units are illustrated in Fig . 2 . We define the tuning curves of the neurons as the mean spike count in response to a 1 s presentation of a constant stimulus . Firing rates monotonically increase ( for positive kernels ) or decrease ( for negative kernels ) as a function of and are rectified at zero , resulting in rectified linear tuning curves ( Fig . 2 A ) . Since all neurons have identical kernels ( i . e . all or ) , neurons with the same kernel signs have identical tuning curves . However , such a homogeneous network is rather implausible since it assumes all-to-all lateral connectivity with identical weights , so that all units in the network receive exactly the same synaptic input and have the same membrane potential . To move to more realistic and heterogeneous networks , we can choose randomized decoding kernels . Even then , however , the connectivity matrix is strongly constrained . For negligible costs , , the weight matrix has rank one ( since ) . A lot more flexibility can be introduced in the network connections by simultaneously tracking variables with identical dynamics and identical control , rather than a single scalar variable . Thus the variable and the kernels have dimensions and . We then define the actual network output , , as the mean of those variables ( simply obtained by summing all network outputs ) . The network estimation error , , is an upper bound on , ensuring similar performance as before ( see Fig . 3 ) . Importantly , we can choose the output kernels to fit connectivity constraints imposed by biology . For instance , the output kernels can be made random and sparse ( i . e . with many zero elements ) , resulting in random and sparse ( but symmetrical ) connection matrices . In such a network , the tuning curves are still rectified-linear , but have different gains for different neurons ( Fig . 2 B ) . Output spike trains of both homogeneous and inhomogeneous networks are asynchronous and highly variable from trial to trial ( see raster plots in Fig . 1 C , D and Fig . 2 ) . Fano factors ( measured during periods of constant firing rates ) , CV1 , and CV2 , were all found to be narrowly distributed around one . The interspike interval ( ISI ) distribution was close to exponential ( Fig . 2 C ) . Moreover , noise correlations between neurons are extremely small and do not exceed 0 . 001 ( noise correlations are defined as the cross-correlation coefficient of spike count in a time window of 1 s in response to a constant variable ) . Finally , analysis of auto and cross-correlograms reveals the presence of high-frequency oscillations at the level of the population ( Fig . 2 D ) . These high frequency oscillations are not visible on Fig . 2 C since the size of the bin ( 5 ms ) is larger than the period of the oscillations ( 1 ms ) . Note that if we add a realistic amount of jitter noise ( ms ) to spike times , these high frequency oscillations disappear without affecting the response properties of the network . In contrast to the output spike trains , the membrane potentials of different neurons are highly correlated , since neurons with similar kernels ( ) receive highly correlated feed-forward and lateral inputs ( Fig . 4 A , B ) . In the homogeneous networks without quadratic cost ( ) , these inputs are even identical , resulting in membrane potentials that only differ by the background noise ( Fig . 4 A ) . Despite these strong correlations of the membrane potentials , the neurons fire rarely and asynchronously . Fig . 4 C illustrates why this is the case: let us consider a population of neurons with identical output kernels , maintaining an estimate of a constant positive ( top panel , blue line ) . Due to the decoder leak , the network needs to fire periodically in order to maintain its estimate at the level of ( top panel , red line ) . However , the exact order at which the different neurons fire does not matter , since they all contribute equally . The period between two spikes can be called an “integration cycle” . Within one integration cycle , the prediction errors and thus the membrane potentials , , rise for all neurons ( bottom panel , red line ) . Since all kernels are identical , however , all neurons compute the same prediction error and will reach their firing thresholds at approximately the same time . Only chance ( in this case , the background noise ) will decide which neuron reaches threshold first . This first neuron is the only one firing in this integration cycle ( middle panel , colored bars ) since it immediately inhibits itself and all other neurons . This starts a new integration cycle . As a result of this mechanism , while the population of neurons fire at regular intervals ( hence the high frequency oscillations in Fig . 2 D ) only one neuron fires in each cycle , and its identity is essentially random . The resulting variability has no impact on the network estimate , since all spike orders give the same output . In the presence of a quadratic cost ( ) , neurons that did not fire recently have a higher probability of reaching threshold first ( their membrane potential is not penalized by ) . When the cost term is large compared to the background noise ( i . e . when , which is not the case in the example provided here ) , this tends to regularize the output spike trains and leads to s smaller than 1 . However , this regularization is not observed in inhomogeneous networks . The inhomogeneous network behaves similarly , except that all neurons do not receive the same inputs and do not reach threshold at the same time ( Fig . 4 B ) . In this case , we can even dispense of the background noise ( i . e . ) since fluctuations due to past network activity will result in a different neuron reaching threshold first in each cycle . The individual ups and downs caused by the synaptic inputs from other neurons will nonetheless appear like random noise when observing a single neuron ( Fig . 4 B , D ) . Furthermore , even in this deterministic regime , the spike trains exhibit Poisson statistics . In fact , changing the timing of a single spike results in a total reordering of later spikes , suggesting that the network is chaotic ( as illustrated in Fig . 3 ) . We now apply this approach to the tracking of a 2D point-mass arm based on an efferent motor command . The dynamical variable has dimensions corresponding to the arm positions and the arm velocities . The arm dynamics are determined by elementary physics , so that ( 19 ) ( 20 ) where is a 2D ( control ) force exerted onto the arm , and captures possible friction forces . To simulate this system , we studied an arm moving from a central position towards different equidistant targets . This reaching out arm movement was obtained by “push-pull” control forces providing strong acceleration at the beginning of the movement , and fast deceleration at the end of the movement ( Fig . 5 A , top panel ) . As previously , the network predicts the trajectory of the arm perfectly based on the forces exerted on it ( Fig . 5 A , bottom panel; we again use relatively small cost terms and ) . The resulting spike trains are asynchronous , decorrelated , and Poisson-like , with unpredictable spike times ( rasters in Fig . 5 A; Fano factor and CVs close to 1 ) . The membrane potential of neurons with similar kernels are correlated while output spike trains are asynchronous and decorrelated . The effective postsynaptic potentials have biphasic shapes reflecting the integrative nature of the network for small friction forces ( ) . To measure tuning curves in this “center out” reaching task , we varied the speed and direction of the movement , as well as the starting position of the arm . Neural activity was defined as the mean spike count measured during movement . As illustrated in Fig . 5 B , C , D , instantaneous firing rates are modulated by arm position , velocity and force . We found that tuning curves to arm position are rectified linear , with varying thresholds and slopes ( as in Fig . 2 B ) . Such linear-rectified gain curves with posture have been reported in premotor and motor cortical areas [32] , [33] . In contrast , tuning curves to circular symmetric variables such as movement direction or arm angle are bell-shaped ( Fig . 5 B , C , D ) . In addition , direction tuning curves are gain modulated by arm speed , such that responses are stronger for larger speed when the arm moves in the preferred direction , and weaker when the arm moves in the anti-preferred direction ( Fig . 5 B ) . Finally , arm positions have both an additive and a gain modulating effect on the tuning curve , and these modulation can be monotonically increasing ( Fig . 5 C ) or decreasing ( Fig . 5 D ) with arm position . These observations have a simple geometric explanation , schematized in Fig . 5 E for the velocity space , . A neuron is maximally active ( ; assuming ) when its kernel ( , thick vector in Fig . 5 E ) points in the direction of the derivative of the prediction error , . Since the decoder leak is faster than the arm dynamics , this error mostly points in the direction opposite to the leak , ( thin vectors ) . Within the velocity space , the kernel thus defines the neuron's preferred movement direction ( dashed line and filled circles ) . The neurons is less often recruited when the arm moves away from the kernel's direction ( empty circles ) , resulting in a bell-shaped tuning curve . Finally , since the vector gets larger at larger speeds , more spikes are required to track the arm state resulting in a linear tuning to movement speed . The same reasoning applies for the position space . These predictions are independent of the choice of kernels and are in direct agreement with experimental data from the pre-motor and motor cortices [32] , [33] . We chose to present a sensory integrator and an arm controller for their biological relevance and simplicity . However , any linear dynamical system can be implemented within our framework , and our networks are not limited to performing integration . To illustrate the generality of the approach , we applied the framework to two additional examples . In Fig . 6 A , we simulated a “leaky differentiator” with a transition matrix . This system of differential equations is designed so that the variable approximates a temporal derivative of a command signal . The command signal , , is shown in the top panel of Fig . 6 A; the input signal is zero . We used neurons with kernels drawn from a normal distribution , and then normalized to a constant norm of . As in the other examples , the firing statistics are close to Poisson , with a . In Fig . 6 B , we simulated a network that implements a damped harmonic oscillator . Here we chose a transition matrix . The oscillator is initially kicked out of its resting state through a force given by the command signal , as plotted on the top panel . The input signal is zero . We used neurons with kernels drawn from a normal distribution , and normalized to a constant norm of . The network tracks the position and speed of the damped oscillator until position and speed are too close to zero to allow a reliable approximation . The firing statistics of single units are again Poisson-like , with . Note that in these two examples , the dynamics implemented by the network are faster than the decoder's time scale . Accordingly , our networks can track changes faster than the time scale of the decoder . This speed-independence relies on a simple scheme: Spikes from neurons with positive kernel weight , , represent instantaneous increases in , whereas spikes from neurons with negative kernel weight represent instantaneous decreases in . Even if the inter-spike interval is much shorter that , the decoder can therefore still provide an efficient staircase approximation for . In conclusion , the temporal accuracy of these networks is not limited by , but by .
Our current work both generalizes and modifies our earlier work in which we applied the principle of predictive coding with spikes to a Bayesian inference problem [20] . This model tracked a log-probability distribution and implemented a non-linear drift-diffusion model , rather than a generic linear differential equation . In addition , we here introduced cost terms which provided us with greater flexibility in regulating and controlling the dynamics of the spiking networks . A quite general framework for designing networks of neurons that implement arbitrary dynamical systems has previously been described in the “neuro-engineering” approach [30] . This approach relies on linearly combining the non-linear rate transfer function of LIF neurons . In its essence , the method is therefore based on firing rates , and makes few predictions about the spiking statistics of cortical neurons . A recently developed model , the “ReFiRe network” [34] provides a recipe for designing networks maintaining stable memories , and shares some of the features of our networks . Just as the neuro-engineering framework , however , its design is essentially based on firing rates . Here we have designed a network based on the principle of predictive coding with spikes . Even though indistinguishable from older models on the single cell level , our work is different in several important respects . A first major difference of our approach is that it predicts a detailed balance between excitation and inhibition , rather than imposing it upfront . This balance follows from the attempt of the network to minimize the loss function , Eqn . ( 4 ) , which in turn implies that the membrane potential of neurons represents a prediction error and that neurons spike only when this prediction error exceeds a certain value—a form of predictive coding . Any increase in excitation causes an increase in prediction error , immediately compensated by an increase in inhibition to bring down the prediction error ( and vice versa ) . This interplay causes a tight temporal correlation between excitation and inhibition at the time scale of the stimulus but also at a much finer time scale , within a single ISI ( Fig . 7 A ) . Note that this balance only holds when considering all inputs . In the leaky integrator , for instance , all lateral connections are inhibitory ( Fig . 1 B , right panel ) . However , the network is still globally balanced when taking into account the contribution from the feedforward connections . Such a tight balance between excitation and inhibition has been observed at several levels of cortical processing [22]–[26] . Accordingly , spike trains in our network usually resemble independent Poisson processes , with rates tuned to the variable . We note that spike trains can also be more regular if the networks are smaller and the noise is not too large . A simple example is a network composed of a single neuron ( ) , for which we provide an analytical solution in Text S1 . Such a LIF neuron responds to a constant positive input with a perfectly regular spike train . In practice , regular firing is observed when only a few neurons are simultaneously co-active ( e . g . for networks composed of less than neurons ) . Firing becomes irregular when many neurons are co-active ( e . g . for networks of several hundreds of neurons or more ) . Increasing synaptic background noise tends to make firing less regular , while increasing the quadratic metabolic costs makes firing more regular . However , for large networks , these effects are small and remain within the range of Fano-factors or CVs observed in cortex . The amount of regularity has no impact on the network performance . Despite the variability observed in large networks , one cannot replace or approximate one of our spiking networks with an equivalent rate model composed of Poisson spike generators , a second major difference to other network models . This point is illustrated in Fig . 7 B , C for the homogeneous integrator model , where we removed the fast connections in the network and replaced the integrate-and-fire dynamics by independent Poisson processes ( see Material and Methods ) . The performance of the resulting Poisson generator network is strongly degraded , even though it has the same instantaneous firing rates and slow connections as the LIF network . We can quantify the benefit of using a deterministic firing rule compared to stochastic rate units by considering how the estimation error scales with the network size . The integrator model tracks the dynamical variable with a precision defined by the size of a kernel . Estimation errors larger than are immediately corrected by a spike . As the network size increases , maintaining the same firing rates in single units requires that the kernels , and thus , the estimation error , scale with ( see Material and Methods ) . In contrast , the error made when averaging over a population of independent Poisson neurons diminishes with . Intuitively , the predictive coding network achieves higher reliability because its neurons communicate shared information with each other via the fast synapses , whereas the independent Poisson neurons do not . The communicated information actively anti-correlates all spike trains , which , for networks composed of more than a dozen neurons , will be indistinguishable from the active decorrelation of pairwise spike trains that has recently been observed in vivo [35] . Therefore , the precision of the neural code cannot be interpolated from single-cell recordings in our network , and combining spike trains recorded in different trials results in a strong degradation of the estimate ( Fig . 7 D ) . A third major difference between our network model and those proposed previously concerns the scaling of the network connectivity . Most previous approaches assumed sparse networks and weak connectivity in which the probability of connections ( and/or connection strengths ) scales as or . This weak connectivity leads to uncorrelated excitation and inhibition and thus neurons driven by random fluctuations in their input [15] , [36] . For comparison , the connectivity in our network is finite ( once the membrane have been rescaled by the kernel norm to occupy a fixed range of voltage ) . Our approach is therefore reminiscent of a recent model with finite connection probability [17] . As in our model , stronger connectivity leads to correlation between excitation and inhibition but uncorrelated spike trains . The strong network connectivity in turn swamps the membrane potential of each neuron with currents . The excitatory and inhibitory currents driving the neural response grow linearly with the number of neurons , , and are thus much larger than the membrane potential ( prediction error ) , which is bounded by the ( fixed ) threshold . In turn , the leak currents become negligible in large networks . For example , the integrator network in Fig . 1 C has neurons and can maintain information for 100 s ( it takes 100 seconds for the network activity to decay by half ) despite the fact that the membrane time constant ( ) is only 0 . 1 s . Thus , according to our model , spiking neurons can fire persistently and thereby maintain information because their leaks are dwarfed by the currents they receive from recurrent connections . There are several non-trivial circumstances under which our network models may fail . First , we notice that the spiking rule that we derive amounts to a greedy optimization of the loss function . Future costs are not taken into account . This may cause problems in real neurons which can only communicate with time delays , but it may also cause problems when neurons have opposing kernels . For instance , two neurons with opposing kernels may become engaged in rapidly firing volleys of spikes , each trying in fast succession to decrease the error introduced by the previous spike from the other neuron ( see Text S1 ) , a problem that we call the “ping-pong” effect . This effect can become a serious problem if the network dynamics is corrupted by strongly correlated noise , which may occur in the presence of synaptic failures . However , it is usually possible to diminish or eliminate this effect by increasing the spike count cost ( see Text S1 ) . Second , the leak term we introduced in the voltage equation provides only an approximation to the actual voltage equation ( see Material and Methods ) . Specifically , the term we approximate is times smaller than the other terms in the membrane potential dynamics . In practice , we can therefore always increase the network size to reach an acceptable level of performance . For a given network size , however , the approximation may break down when becomes too large or when both and are too small ( of order ) . Third , the speed at which the linear dynamical system can evolve will be limited from a practical point of view , even in the limit of large networks . While the time scale of the decoder , does not put any strict limitations on the speed ( since spikes corresponding to positive and negative kernels can always provide an efficient stair-case approximation to any time-varying function ) , faster dynamics can only be obtained if the linear dynamical system compensates for the decoder filtering . This compensation or inversion process is a case of deconvolution , and bound to be severely limited in practice due to the noise inherent in all physical systems . Finally , the network requires finely tuned lateral connections in order to balance excitation and inhibition ( from feed-forward and lateral connections ) . In particular , the strength of the fast connections between two neurons corresponds to minus the correlation coefficient of their feed-forward connections ( and thus , to their level of shared inputs ) . Whether such finely tuned motifs exist in biological networks is still an open question . We showed recently that fast lateral connections can be learnt using unsupervised Hebbian learning [37] , suggesting that networks with the appropriate form of plasticity would be able to develop and maintain this tight balance . We note that the performance of the networks is quite sensitive to global perturbations of the balance between excitation and inhibition , an issue that we discuss in more detail in Text S1 . The most crucial work left to the future will be to test the predictions derived from this theory , three of which are described here . First , one could test how the decoding error scales with the numbers of simultaneously recorded neurons . A single unit in the model network ( considered in isolation ) is in fact exactly as reliable as a Poisson spike generator with the same rate . As the number of simultaneously recorded neurons increases , the decoding error initially decreases as , similar to a Poisson rate model . However , as the number of neurons reaches a certain threshold ( 10% for the network models simulated here ) , the error from the spiking network decreases faster than predicted for a Poisson rate model ( Fig . 7 E ) . So far , single-unit recordings or multi-electrode recordings have only sampled from a very small subpart of the population , making it impossible to see this difference ( and in turn , potentially leading to an under-estimation of the precision of the neural code ) . However , with newer techniques , such as dense multi-electrode arrays or optical imaging , as well as with focusing on smaller networks ( such as the oculomotor integrator or insect systems ) , these model predictions are nowadays within experimental reach . We note that one has to carefully account for the effect of shared sensory noise ( ) to see the predicted scaling effect . Shared noise ( absent in Fig . 7 E ) introduces correlations between neurons and results in a saturation of the error with . In our network , such a saturation would only be seen if there were limits to the sensory information available in the first place; saturation would not be seen as a consequence of neural noise or correlations ( as proposed for example in [38] , [39] ) . Second , one could look at the global interaction between neurons of similar selectivity , for example by applying a GLM model to the data [28] . The model predicts that neurons involved in slow integration tasks or working memory tasks should inhibit and excite each other at different delays . In particular , neurons with similar selectivities should be ( paradoxically ) negatively correlated at short delays . Thus , applying GLM analysis even on a small sub-population can uncover the effective PSPs caused by the lateral connections and , indirectly , the dynamical equation implemented by the network . Fig . 7 F shows the GLM filters learnt from the inhomogeneous integrator network during working memory ( i . e . sustained activity in the absence of sensory input ) . The analysis recovered the shape of the filters between neurons of similar kernels and opposite kernels , despite the fact that only 10 simultaneously recorded neurons ( 2 . 5% of the population ) were used in this analysis . Third , the spiking network is by essence self-correcting and will thus be resilient to lesions or many sudden perturbations ( an exception being perturbations of the global balance of excitation and inhibition , see above ) . Equipping neural networks with such resilience or robustness has been a well-studied theoretical problem . For the specific example of the neural integrator , solutions range from constructing discrete attractor states [40] , [41] , [42] , adding adaptation or learning mechanisms to a network [43] , [44] , or changing the nature of network feedback [45] , [46] . In the case of the neural integrator , the robustness of our network could likely be interpreted as a case of derivative feedback [46] . While we know that biological neural networks are quite robust against partial lesions , their response to sudden , yet partial perturbations is less well known . For example , suddenly inactivating half of the active neurons in our sensory integrator increases the firing rates of the remaining neurons but has essentially no effect on the network performance ( Fig . 7 G ) . This instantaneous increase in firing rates without performance loss generates a strong prediction for our network model , a prediction that distinguishes our network from previously proposed solutions to the robustness problem . Indeed , as long as the pool of available kernels remains sufficient to track , and as long as increased firing rates are not affected by saturation , inactivation will not affect the network's computation . This prediction could be tested using for example optogenetic methods .
We here derive the network equations using compact matrix-vector notation . In Text S1 , we also consider the special case of a homogeneous network and a single neuron , for which the derivations are simpler . We consider the error function , Eqn . ( 4 ) , which is given by ( 21 ) The -th neuron should spike at time if ( 22 ) A spike by the -th neuron adds a single delta-function to its spike train . This additional spike enters the right-hand-side of the read-out equation , Eqn . ( 2 ) . Integration of this extra delta-function amounts to adding a decaying exponential kernel , to the read-out . Hence , if neuron spikes at time , we have ( 23 ) ( 24 ) where the latter equation describes the instantaneous change in firing rate due to the additional spike . Note that the standard Eucledian basis vector is a vector in which the -th element is one , and all others are zero . Each spike influences the read-out several time intervals into the future . To see whether a spike leads to a decrease of the error function , we therefore need to look into the future ( from time onwards ) . For a future time with , the spiking rule in Eqn . ( 22 ) translates into ( 25 ) We can expand the terms on the left-hand-side , and then eliminate identical terms on both sides . For that , we remind the reader that the relation holds for the norm , whereas holds for the norm in our case , since all elements ( firing rates ) are positive by definition . Hence we obtain ( 26 ) We rearrange the inequality by moving all terms that depend on the dynamical variables , the estimates , or the firing rates to the left , and all other terms to the right , and we then multiply both sides by minus one , to obtain ( 27 ) Moving the kernels to the front of the integrals and noticing that for , we obtain ( 28 ) The integral on the left-hand-side weights the influence of the spike , as given by the decaying exponential kernel , , against the future development of the error signal , , and firing rate . These future signals are unknown: while we may be able to extrapolate , given its dynamical equation , we cannot safely extrapolate or , since this would require knowledge of all future spikes . We therefore choose a “greedy” approximation in which we only look a time into the future . For the relevant times , we can then approximate the integrands as constants so that ( using for ) ( 29 ) which is our decision to spike , and corresponds exactly to Eqns . ( 5–7 ) . We notice that the right-hand-side is a constant whereas the left-hand-side is a dynamical quantity which corresponds to the projection of the prediction error , , onto the output kernel of the -th neuron , , subtracted by a term depending on the firing rate of the -th neuron . Given this threshold rule , it seems only natural to identify the left-hand-side with the membrane voltage of the -th neuron and the right-hand-side with its spiking threshold , , which is what we did in the main text . If we write the voltage of all neurons as one long vector , , then we can write ( 30 ) We generally assume that there are more neurons than variables to represent so that . We also assume that the output kernel matrix , , has rank , and that the dynamical variables are not degenerate or linearly dependent on each other . In this case , the left pseudo-inverse of exists and is given by ( 31 ) so that . Note that is an -matrix , while has size . In turn , we can solve the voltage equation for by multiplying with the pseudo-inverse from the left so that ( 32 ) Taking the derivative of the voltages , we obtain ( 33 ) Replacing , , and with their respective equations , Eqns . ( 1–3 ) , we obtain ( 34 ) In turn , we can replace with Eqn . ( 32 ) to obtain ( 35 ) Sorting some of the terms , and remembering that , we obtain ( 36 ) To evaluate the relative importance of the different terms , we consider the limit of large networks , i . e . , the limit . First , we impose that the average firing rates of individual neurons should remain constant in this limit . Second , we require that the read-out does not change . Given the scaling of the firing rates , and since , the output kernels must scale with . Accordingly , the pseudo-inverse scales with . Finally , we need to choose how the cost terms and , scale with respect to the read-out error . The linear and quadratic error terms and scale with . To avoid a contribution of the cost term increasing with network size , and should scale ( at the least ) with . However , even if the cost terms scale with , they will still dominate the network dynamics . For instance , the threshold , Eqn . ( 7 ) becomes independent of the output kernel , while the contribution of fast lateral connections becomes negligible . In practice , this causes the performance to degrade quickly with increasing network size . A better choice is to require and to scale with , keeping the relative contribution of the kernel and cost to each neuron's dynamics fixed . With such scaling , large networks can still track the variable while the performance increase with network size . Given the scaling of the output kernels and , the threshold scales with , compare Eqn . ( 7 ) . In turn , since the voltage is bounded by the threshold from above ( and bounded from below due to the existence of neurons with opposing kernels; see also below ) , the voltage also scales with . Accordingly , in a large network , the first , voltage-dependent term in Eqn . ( 36 ) scales with , as do the terms and . In contrast , the terms and represent a sum over all neurons in the population , and thus scale with , similar to the inputs . For large networks , we can therefore neglect the terms that scale with . We note that none of the terms involving delta functions ( i . e . ) can be neglected . We keep a generic leak term , , although the term is essentially irrelevant in large networks , and may be detrimental in very small ones ( e . g . , less than 10 neurons ) . Hence , we approximate Eqn . ( 36 ) by ( 37 ) with ( 38 ) ( 39 ) Since and since , we can define the effective connectivities ( 40 ) to obtain the voltage equation ( 41 ) which is the vectorized version of Eqn . ( 8 ) without the noise term . In the homogeneous integrator network ( with low noise and small costs ) , the membrane potentials of neurons with identical kernels are approximately equal , which allows us to write down an analytical solution ( see Text S1 ) . Briefly , the population inter-spike interval , i . e . the interval between two successive spikes from any neuron , corresponds to the time it takes for this “common” membrane potential to rise from the reset potential to the threshold . We call this time period an “integration cycle” . Note that this interval is typically much shorter than the ISI of an individual neuron or the time constant of the decoder . During this short time interval , the leak term can be neglected , and the derivative of the membrane potential , , is approximately constant . The population ISI is thus simply given by the time it takes to integrate from the reset , , to the threshold , , so that . All neurons with the same kernel have identical firing rates , and , since only half of the population is spiking at any value of ( in the limit of small noise ) , the firing rates of individual neurons are equal to the population firing divided by . Thus , the firing rate of each neuron can be approximated as . To construct the Poisson generator network , we removed the fast connections ( but not the slow connections ) and replaced the LIF neurons by Poisson spike generators with the same instantaneous firing rate , i . e . , . The resulting recurrent network roughly matches the instantaneous firing rates ( but not the performance ) in the LIF network . The match could be enhanced , for example by adding a small baseline firing rate or a refractory period; However , these changes can only decrease the performance of the Poisson rate model . To obtain the filters in the integrator network ( Fig . 7 F ) , we performed the following procedure: The inhomogeneous integrator was driven by an input sampled from Gaussian white noise ( with mean , standard deviation ) and convolved by an exponential filter of width ms . The spike trains of the ten “recorded” neurons were modeled as independent Poisson processes with instantaneous firing rates ( 54 ) The feed-forward weights and lateral filters were estimated by maximizing the log-likelihood of the spike trains , following the method of [28] . Briefly , the filters were discretized in 500 time bins of , and conjugate gradient ascent of the log likelihood was performed on the value of the filters in each time bin for the equivalent of 5 hours of recording . The of a spike train is defined as ( 55 ) where is the total number of spike in the spike train , and is the duration of the inter-spike interval . The reported in the paper are the value of measured in each neuron and averaged over the population . | Two observations about the cortex have puzzled and fascinated neuroscientists for a long time . First , neural responses are highly variable . Second , the level of excitation and inhibition received by each neuron is tightly balanced at all times . Here , we demonstrate that both properties are necessary consequences of neural networks representing information reliably and with a small number of spikes . To achieve such efficiency , spikes of individual neurons must communicate prediction errors about a common population-level signal , automatically resulting in balanced excitation and inhibition and highly variable neural responses . We illustrate our approach by focusing on the implementation of linear dynamical systems . Among other things , this allows us to construct a network of spiking neurons that can integrate input signals , yet is robust against many perturbations . Most importantly , our approach shows that neural variability cannot be equated to noise . Despite exhibiting the same single unit properties as other widely used network models , our balanced networks are orders of magnitudes more reliable . Our results suggest that the precision of cortical representations has been strongly underestimated . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [] | 2013 | Predictive Coding of Dynamical Variables in Balanced Spiking Networks |
The relationship between chromatin structure and gene expression is a subject of intense study . The universal transcriptional activator Gal4 removes promoter nucleosomes as it triggers transcription , but how it does so has remained obscure . The reverse process , repression of transcription , has often been correlated with the presence of nucleosomes . But it is not known whether nucleosomes are required for that effect . A new quantitative assay describes , for any given location , the fraction of DNA molecules in the population that bears a nucleosome at any given instant . This allows us to follow the time courses of nucleosome removal and reformation , in wild-type and mutant cells , upon activation ( by galactose ) and repression ( by glucose ) of the GAL genes of yeast . We show that upon being freed of its inhibitor Gal80 by the action of galactose , Gal4 quickly recruits SWI/SNF to the genes , and that nucleosome “remodeler” rapidly removes promoter nucleosomes . In the absence of SWI/SNF , Gal4′s action also results in nucleosome removal and the activation of transcription , but both processes are significantly delayed . Addition of glucose to cells growing in galactose represses transcription . But if galactose remains present , Gal4 continues to work , recruiting SWI/SNF and maintaining the promoter nucleosome-free despite it being repressed . This requirement for galactose is obviated in a mutant in which Gal4 works constitutively . These results show how an activator's recruiting function can control chromatin structure both during gene activation and repression . Thus , both under activating and repressing conditions , the activator can recruit an enzymatic machine that removes promoter nucleosomes . Our results show that whereas promoter nucleosome removal invariably accompanies activation , reformation of nucleosomes is not required for repression . The finding that there are two routes to nucleosome removal and activation of transcription—one that requires the action of SWI/SNF recruited by the activator , and a slower one that does not—clarifies our understanding of the early events of gene activation , and in particular corrects earlier reports that SWI/SNF plays no role in GAL gene induction . Our finding that chromatin structure is irrelevant for repression as studied here—that is , repression sets in as efficiently whether or not promoter nucleosomes are allowed to reform—contradicts the widely held , but little tested , idea that nucleosomes are required for repression . These findings were made possible by our nucleosome occupancy assay . The assay , we believe , will prove useful in studying other outstanding issues in the field .
Gal4 is an intensively studied transcriptional activator found in the yeast Saccharomyces cerevisiae . Galactose , added to the growth medium , frees Gal4 of its inhibitor Gal80 , and the DNA-bound activator quickly and strongly induces genes required to metabolize the sugar . Two such genes are the divergently transcribed GAL1 and GAL10 , between which lie four Gal4 binding sites comprising the so-called upstream activating sequence , galactose ( UASg ) . A wide array of studies shows that Gal4 recruits to nearby yeast genes protein complexes required for transcription [1 , 2] . Gal4 also activates any of a wide array of genes in higher eukaryotes when ectopically expressed , provided the target gene bears Gal4 binding sites nearby . This ability to activate so many genes in so many different organisms probably reflects its ability to bind , and thereby recruit , a wide array of targets . For example , Gal4 contacts at least three yeast protein complexes ( called SAGA , TFIID , and Mediator ) [3 , 4] , and thereby activates transcription of genes that require different subsets of these complexes [5 , 6] . Addition of glucose , a preferred carbon source , to cells growing in galactose inhibits expression of the GAL genes in several ways . The strongest direct effect is repression of GAL4 and of GAL2 , which encodes the galactose permease . A smaller effect is that the GAL1 , 10 genes are also directly repressed ( see Discussion ) [7–10] . Gal4 , like other eukaryotic transcriptional activators , must work despite the fact that DNA is wrapped in nucleosomes . For example , nucleosomes in the GAL1 and GAL10 promoters ( referred to as promoter nucleosomes ) would cover DNA that must be available for the transcriptional complex to form . And indeed , several experiments show that these nucleosomes , present on the inactive promoters , are missing when the genes are transcribed [11–17] . One mechanism for this nucleosome loss would be that the recruited machinery simply competes them away . Consistent with this idea , fusion proteins bearing a DNA binding domain attached to one or another subunit of the transcriptional machinery ( e . g . , LexA-Gal11 ) can activate transcription to a high level . Such fusion proteins presumably directly recruit the transcriptional machinery to the promoter without removing nucleosomes in a separate step [18] . A second possibility for nucleosome removal by an activator would be that it recruits a function that removes nucleosomes in a step separate from recruitment of the transcriptional machinery itself . This scenario has been shown to hold for the PHO8 gene of yeast: in this case , the nucleosome remodeling complex SWI/SNF , recruited to a gene by the activator , removes promoter nucleosomes in an early step in the process of gene induction [19] . It has been reported , however , that SWI/SNF plays no role in ordinary induction of the GAL genes [20 , 21] . Just as removal of promoter nucleosomes is correlated with activation of transcription , so is their reformation typically correlated with the turning off of transcription . For example , when cells are transferred from galactose to glucose , and GAL gene transcription ceases , promoter nucleosomes rapidly reform at these genes [15 , 17] . Whether this reformation of promoter nucleosomes is required for gene silencing , or rather is a consequence of that inactivity , is not known . In the typical analysis of glucose repression of the GAL genes , Gal4 is either inactive ( due to the absence of galactose ) or is depleted by one of the long-term effects of glucose as mentioned above . The possibility that Gal4 might continue to function in the simultaneous presence of galactose and glucose , and if so to what end , has not , to our knowledge , heretofore been considered . Here , we reexamine these matters using a quantitative micrococcal nuclease protection assay that measures , at any given moment , and for any specified DNA fragment , the fraction of the population that is occupied by a nucleosome in vivo . We show that in an early step of gene activation , Gal4 recruits SWI/SNF and quickly removes promoter nucleosomes . In the absence of SWI/SNF , a high level of transcription is also reached and promoters cleared of nucleosomes , but the process proceeds significantly more slowly . We confirm that upon transferring cells from galactose to glucose , transcription quickly diminishes , and nucleosomes rapidly reform at the promoter [17] . In contrast , however , if glucose is added to cells growing in galactose , although mRNA production also quickly diminishes , nucleosomes do not rapidly reform . We show that , under these repressive conditions , galactose continues to counter the inhibitory effect of Gal80 . Gal4 continues to recruit SWI/SNF , which , in turn , prevents nucleosome reformation despite the onset of repression .
It has long been recognized that nucleosomes protect from micrococcal nuclease digestion fragments of DNA of about 140–160 bp . In a typical modern version of such an experiment , cells are fixed with formaldehyde , and isolated chromatin is lightly digested with a single dose of nuclease for a fixed time . Cross-linking is then reversed , and mononucleosomal-sized DNA fragments of about 150 bp are isolated . These recovered fragments are identified by PCR , microarray analysis , and/or DNA sequencing [22–24] . The novel aspect of our assay is that cross-linked chromatin is digested with nuclease over a wide range of nuclease concentrations . We then use a series of overlapping primer pairs ( amplicons , ∼60 bp each ) to quantitate the reaction products by real-time PCR without any prior fractionation for the size of protected fragments . Our method , like others , allows us to delineate arrayed nucleosomes ( as confirmed by chromatin immunoprecipitation [ChIP] analysis ) and to distinguish them from randomly positioned nucleosomes . Beyond that , however , we shall see that for any DNA segment potentially bearing a positioned nucleosome ( such as in the GAL1 promoter ) , we can determine the fraction of that DNA segment , in the population , that is occupied by a nucleosome at the moment of cross-linking . This allows us to determine nucleosome positioning and occupancy prior to induction and to follow the time courses of nucleosome removal and reformation upon induction and repression . Consistent with many previous studies using micrococcal nuclease ( e . g . , [11–13] ) , the transcriptional machinery itself , recruited to the promoter , does not protect against micrococcal nuclease digestion . ( see Figure S1 ) . In initial experiments , we found , as expected , that DNA , purified from cells not exposed to the cross-linking reagent ( hereafter referred to simply as purified DNA ) , yielded , in every case , a first-order decay function when digested and analyzed as outlined above . The digestion rates of the segments ( k , blue lines in Figure 1 ) varied in value as much as 10-fold , indicating differences in the intrinsic sensitivities of different DNA sequences to micrococcal nuclease . When this experiment was performed on cross-linked chromatin , a few locations yielded a monophasic digestion pattern like that of purified DNA , indicating the presence of a single species in the population . The absolute values of the relevant digestion rates varied from experiment to experiment depending upon the specific activity of the nuclease , DNA concentrations , and impurities in the chromatin preparations . We normalize these digestion rates by setting that of one of these locations ( which we call naked , or hypersensitive [HS] ) equal to that of its counterpart in purified DNA . We then find that the rates of digestion of the other HS sites , compared to the first , are predicted by their relative intrinsic sensitivities as determined with purified DNA ( see Figure 1A for an example ) . Exceptions to this rule—that the digestion rate describing a monophasic curve is predicted by the digestion rate of the corresponding purified DNA—are found in the UASg , a matter we return to below . Most chromatin locations yield curves that , unlike those just discussed , are biphasic , consisting of rapidly digested and slowly digested portions , indicating the presence of two subpopulations . Considering only the rapidly digested portion , again the first order rates of the reaction were related to one another as were those of their counterparts in purified DNA . We call this subpopulation naked . In striking contrast , for the remaining portion of each biphasic curve , the digestion rate ( again first order ) was some 200-fold slower than that in the faster digesting portion ( compare k1 and k2 in Figure 1B and 1C ) . The fact that this degree of protection is so constant over so many locations ( varying no more than plus or minus some 2-fold ) suggests that it is caused by a common factor bound to DNA . As this and other evidence confirms , the typical protecting factor is a nucleosome . Thus , analysis of each biphasic curve reveals , for the corresponding DNA fragment , the fraction in the population that bears a nucleosome , and the fraction naked , at the moment of cross-linking . For example , the chromatin fragment of Figure 1B comprises a population about 46% naked and 54% occupied . For the case of Figure 1C , the corresponding fractions are 61% and 39% . Of the approximately 500 DNA segments we have examined from around the genome , most yield biphasic digestion curves , and these curves differ from each other primarily in the percent protected as in the examples shown . Figure 2 shows that in cells in which the GAL1 , 10 genes were not expressed , rather precisely positioned nucleosomes were found flanking the UASg . This conclusion was reached , in part , by experiments in which we measured the nuclease sensitivities of many overlapping short ( ∼60 bp ) DNA segments . The nuclease sensitivities of 40 such segments are represented by the colored bars above the gene schematic ( Figure 2B ) . Each bar represents a single DNA fragment , indicated by its position , assayed using a specific amplicon . The fraction of each bar that is green corresponds to the fraction of the population of the corresponding DNA fragment that was protected ( occupied ) in a typical experiment , and vice versa for the fraction that is red . Thus , each bar represents a digestion curve , four examples of which are shown in Figure 2A . Whenever a bar crosses one of the paired vertical dashed lines , it is largely red , indicating that that site is naked ( unprotected , HS ) in essentially every member ( >90% ) of the population in vivo . About 130 bp ( plus or minus a few base pairs ) separate the two HS sites flanking the UASg , but 160 bp separate the HS sites to the right and left of the UASg . Two less well-defined HS sites lie further downstream and upstream , separated again from their neighboring HS sites by about 160 bp . The repeat length of 160 bp is that expected if nucleosomes are positioned as shown in Figure 2C ( green ovals ) , each protecting about 160 bp from digestion and separated from each other by 10–15 bp . The ChIP experiment of Figure 2E , in which FLAG-tagged H2B was probed , supports the conclusion that the 160-bp regions between HS sites in Figure 2 are occupied by nucleosomes . This experiment ( a “high-resolution” ChIP ) included a step in which cross-linked chromatin was lightly digested with micrococcal nuclease prior to immunoprecipitation [25] . Absent this step , the nucleosome positioning is less well defined ( see Figure S2 ) . These nucleosomes are positioned on DNA sequences crucial for formation of the transcription complex . Thus the two nucleosomes on the GAL1 promoter around the TATA box , and the single GAL10 promoter nucleosome span the distance between the UASg and the transcription start site . In the remainder of this paper , we refer to the nucleosome positioned just to the right of the UASg in Figure 2C as the GAL1 upstream nucleosome , and to the three nucleosomes of the figure as promoter nucleosomes . The results of these micrococcal nuclease digestion experiments and the ChIP experiments were essentially the same whether cells were grown in raffinose ( a noninducing sugar ) or glucose . The depicted positions of these nucleosomes are consistent with earlier analyses [11 , 14 , 16] . Figure 2B also shows that as indicated by the positions of the largely or completely green bars , prior to induction , DNA segments in the UASg were protected in essentially every member of the population . Several experiments indicate that this protection is caused by some molecule bound to the UASg and that this molecule is neither Gal4 nor a nucleosome . First , the UASg isolated from cross-linked chromatin ( see Figure 2A ) is digested more slowly than predicted from the rate of digestion of its purified counterpart , indicating that the nuclease resistance of this fragment is not an intrinsic property of its sequence . Second , the pattern of protection shown in Figure 2B , obtained with wild-type cells grown in raffinose , was unchanged by deletion of GAL4 ( unpublished data ) . Third , as noted above , the ChIP experiment of Figure 2E shows the region to be free of histone H2B . Fourth , the rate of digestion of the protected fragment was significantly faster than that predicted were the protecting factor a nucleosome ( see Figure S3 ) . Fifth , as shown in Figure 2C , the size of the protected UASg fragment , defined by the distance separating the flanking HS sites , is considerably smaller than that of the repeat length of a nucleosome ( 130 bp versus 160 bp ) . Finally , as we shall see ( Figure 2D ) , the promoter nucleosomes are removed upon induction , rendering the DNA naked in our assay , whereas the UASg remains protected throughout our experiments . We do not know the identity of the putative molecule bound to the UASg , nor do we know its function , if any . Others have noted that some molecule other than a nucleosome or Gal4 can occupy the UASg [14 , 26 , 27] . We draw attention to this molecule here only because , as we shall see , the protection it confers , which remains constant throughout our experiments , serves as a useful reference point . We return now to the nucleosomes flanking the UASg . Figure 2D shows that at the moment of cross-linking , a significant fraction of the population is missing one or another of the depicted nucleosomes prior to induction . Thus , columns 1 , 5 , and 6 , which represent DNA segments found at the centers of the three positioned nucleosomes , show that prior to induction ( the top three rows ) , only about half of each segment in the population is protected . In contrast , as shown in column 3 , the UASg is 100% protected . Although more complicated scenarios might be imagined ( see Discussion ) , a simple explanation for these results is that at any given instant , one or more of the depicted nucleosomes is absent from about 50% of the population . This level of occupancy was essentially the same in wild-type cells grown in either glucose or raffinose , and was unchanged by deletion of GAL4 ( unpublished data ) . Our experiments , as well as those of others , indicate a nucleosome disposition in the GAL1 ORF different from that found in the promoter . Thus , ChIP experiments indicate the presence of histones more or less uniformly across the ORF ( see [28] and Figure S4 ) Because the nucleosomes in the ORF are not regularly positioned as they are in the promoter , it is difficult to measure precisely the typical level of occupancy of an individual ORF nucleosome . We do , however , estimate that level to be significantly higher in the ORF than in the promoter ( see Figure S4 ) . Figure 2E shows that as analyzed by a ChIP experiment , GAL1 promoter nucleosomes , present before induction , were absent from cells grown for many generations in galactose . The time course of removal of these nucleosomes following the addition of galactose to cells growing in raffinose is revealed by our protection assay ( Figure 2D ) . Consider , for example , the fragments represented in columns 1 , 5 , and 6 in the figure . For each of these fragments , the fraction of the population occupied by a nucleosome steadily decreased as induction proceeded . Nucleosome removal began about 5 min after addition of galactose and was complete by about 12–16 min . In contrast , the UASg remained highly protected . The naked regions flanking the UASg , as well as that separating the two nucleosomes to the right of the UASg , remained naked . An induction experiment in which we simultaneously measured nucleosome removal ( using the nuclease protection assay ) and recruitment of the transcriptional machinery to the GAL1 promoter ( using the ChIP assay as in [1] ) , showed that nucleosome removal was approximately coincident with the appearance at the promoter of the transcriptional machinery . We also found that as previously reported [29] , SWI/SNF is quickly recruited to the UASg by Gal4 ( Figure 3D ) . Figure 3A shows the progressive removal of the GAL1 upstream nucleosome in a series of mutant strains . The figure shows that deletion of the SAGA component SPT20 , which drastically reduces formation of the transcription complex [1 , 30 , 31] , had no effect on the time course of nucleosome removal . Deletion of any of the following genes also had no effect on the rate of this reaction: GCN5 , which encodes the histone acetyltransferase in SAGA; SPT7 , which encodes a core SAGA component; and SRB2 and PGD1 , which encode Mediator components . The figure also shows , in striking contrast , that deletion of SNF2 , the catalytic subunit of the SWI/SNF complex , dramatically delayed nucleosome removal and the onset of transcription ( Figure 3C ) . Figure 3B shows that certain mutations that had no effect on nucleosome removal nevertheless had strong deleterious effects on transcription . Thus , at 15 min postinduction , GAL1 mRNA levels were strongly diminished in the spt20Δ and snf2Δ strains , moderately diminished in the spt7Δ , pgd1Δ , and srb2Δ strains , and essentially at wild-type levels in the gcn5Δ strain . The other two promoter nucleosomes behaved identically to the GAL1 upstream nucleosome in these experiments ( G . O . Bryant and M . Ptashne , unpublished data ) . Figure 3C , a time course of mRNA production following induction , shows that in snf2Δ mutants , mRNA production reached levels obtained with the wild-type strain , but only over a much longer time course that paralleled promoter nucleosome loss . These experiments also support the finding alluded to above that the recruited transcriptional machinery , readily detected by ChIP assays , does not protect against micrococcal nuclease digestion . The conclusion that SWI/SNF is required for rapid nucleosome removal , and that delayed nucleosome removal also delays the onset of transcription , stands in contrast to a report that mutation of SNF2 has no effect on induction of transcription of the GAL genes [20] . We therefore obtained the snf2Δ strain of Kundu et al . , and found that this strain , as well as the snf2Δ strain to which a wild-type SNF2 allele had been added , behaved identically to our corresponding strains in an assay for the rate of synthesis of mRNA upon induction ( Figure S5 ) . Figure 4A shows that in cells grown in galactose , washed , and then resuspended in glucose , the GAL1 upstream nucleosome quickly ( within 10 min ) reformed on a fraction of templates equal to that occupied prior to induction . The other two nucleosomes in Figure 2 reformed with an indistinguishable time course ( G . O . Bryant and M . Ptashne , unpublished data ) . This time course of promoter nucleosome reassembly mirrored the time course of loss of mRNA production ( Figure 4B , green line ) . A strikingly different result was obtained if , instead of transferring cells from galactose to glucose , the cells were resuspended in medium containing glucose ( 2% ) plus galactose ( 2% ) . Figure 4A ( red line ) shows that in this case , over the first few hours following this transfer , nucleosomes reformed only slowly . As indicated in Figure 4B , however , transcription decreased as dramatically as in the previous experiment . Thus , for example , the early phase of glucose repression ( as assayed by mRNA production in Figure 4B ) was complete by 30 min , but nucleosome formation only reached about half of its original value by 3 h . Figure 4C , a ChIP experiment , shows that several components of the transcriptional machinery ( RNA polymerase II , Gal11 , and TFIIE ) , each of which had bound to the promoter in cells grown in galactose , were quickly depleted from the promoter upon resuspension in glucose plus galactose , and the time course of this depletion mirrored the time course of repression of transcription . ChIP analysis also showed that Gal4 was bound to the UASg over the course of the experiments of Figure 4 ( Figure S6 ) . The glucose repression of the GAL genes we observed upon transfer to glucose plus galactose was not due to a nonspecific effect on transcription . Thus , three other genes ( HHF1 , ACT1 , and RPB11 ) , which are constitutively active , remained so in the presence of glucose in our experiments ( unpublished data ) . Figure 5A shows that although repression of transcription by glucose was not affected by the simultaneous presence of galactose , nucleosome reassembly was . Thus , at some 20 min after addition of glucose , the extent of nucleosome reassembly was approximately inversely proportional to the concentration over a 20-fold range , of galactose . The results indicate that even in the presence of glucose , galactose activates Gal4 ( by removing the inhibitory effect of Gal80 ) , and Gal4 recruits some function that keeps the region nucleosome-free despite the repression of transcription . The following experiments support this surmise . First , ChIP experiments show that Gal4 remained present at the UASg over these time courses ( Figure S6 ) . Second , as shown in Figure 5B , in the absence of Gal80 , galactose had no effect on the extent of nucleosome redeposition . In this case , Gal4 worked constitutively to keep the promoters nucleosome-free even though transcription was repressed . And finally , two additional experiments involving glucose repression identify a crucial function recruited by Gal4 as—once again—SWI/SNF . First , the ChIP experiment of Figure 5D shows that following resuspension in medium containing galactose and glucose , Snf2 continued to be recruited to the promoter , whereas if glucose was substituted for galactose , Snf2 recruitment ceased . Second , we induced snf2Δ mutant cells by growth in galactose ( a process that takes many hours to be fully realized—see Figure 3C ) , and then resuspended the cells in medium containing glucose and galactose . In this case , nucleosomes reformed more quickly than in wild-type cells ( see Figure 5C ) . We conclude that recruited SWI/SNF plays a significant role in keeping the promoter nucleosome-free following transfer of cells from galactose to medium containing galactose and glucose . These experiments were also performed simply by adding glucose to cells growing in galactose , with results essentially the same as those obtained by resuspending cells in glucose plus galactose ( unpublished data ) .
Our finding that none of several components of the transcriptional machinery is required for rapid removal of promoter nucleosomes indicates that the action of SWI/SNF , recruited by Gal4 and perhaps assisted by chaperones [32] , suffices to remove nucleosomes in an early step of transcriptional activation . In several strains mutated for components of SAGA and the Mediator , mRNA production was severely delayed or diminished with no effect on nucleosome removal . We do not know whether histone acetylation aids in this reaction , but we saw no difference in the rate of nucleosome removal between wild-type cells and cells deleted for the histone acetyltransferase GCN5 . Our results also show that in the absence of SWI/SNF , nucleosome removal and transcription were elicited by the action of Gal4 , but the reactions were considerably slower than in the presence of SWI/SNF ( requiring hours versus minutes ) . In this case , assuming the recruited machinery competes away the promoter nucleosomes , the reaction may be facilitated by the fact that these nucleosomes spontaneously vacate the promoter relatively frequently ( see Promoter Nucleosomes below ) . Such a scenario might also explain the ability of fusion proteins , bearing a DNA binding domain attached to a component of the transcriptional machinery ( e . g . , LexA-Gal11 ) , to activate transcription to a high level . Consistent with this idea , we have found that activation by such a fusion protein follows a slow time course approximating that triggered by Gal4 in a strain mutant for SWI/SNF ( X . Wu , M . Floer , and M . Ptashne , unpublished data ) . A previous claim , emanating from this laboratory , that SWI/SNF is not involved in induction of the GAL genes [21] , is explained by a failure to examine the stages in the course of the reaction , relying instead upon the end result observed after many hours of induction . In a different report , analyzed in the text , it was claimed that SWI/SNF had no effect on the onset of transcription [20] . Our results are in contrast to that claim . We had no reason to anticipate our finding that galactose can continue to signal to Gal80 , removing its inhibitory effect on Gal4 , even as transcription is repressed by glucose . This was first suggested by the observation that the extent of nucleosome reformation some 30 min following addition of glucose was approximately inversely proportional to the concentration in the medium of galactose . This suggestion was confirmed by our finding that in a strain deleted for GAL80 , the promoter remained nucleosome-free at this time point even in the absence of galactose . We surmised that Gal4 maintains the promoter nucleosome-free under repressive conditions just as it does when activating transcription , namely , by recruiting SWI/SNF . As predicted by this scenario , in a strain mutated for SWI/SNF , promoter nucleosomes rapidly reformed upon addition of glucose whether or not galactose remained present . The ability of Gal4 to work at early times following the addition of glucose , maintaining the promoters nucleosome-free , does not contradict the known mechanisms for glucose repression alluded to in the Introduction . Thus , direct repression of GAL4 and GAL2 , even if immediate , would have an effect on GAL1 , 10 expression only as the previously synthesized Gal4 and Gal2 proteins were diluted away . And direct repression of GAL1 , 10 , as measured hours after addition of glucose , is reported to be weak , only some 2–3-fold [7 , 9] . This direct repression is believed to be caused by recruitment of the Tup1 repressing complex to the GAL1 , 10 region by the specific DNA binding protein Mig1 , and in preliminary experiments , we have found little if any alleviation of early glucose repression of GAL1 , 10 by deleting MIG1 ( G . O . Bryant and M . Ptashne , unpublished data ) . As expected from these various considerations , in our experiments , nucleosomes do slowly reform at promoters in the presence of galactose and glucose , presumably a consequence of depletion of Gal4 and Gal2 . An implication of our findings is that early negative effects of glucose on transcription cannot be ascribed to an elimination of all Gal4 recruiting activities . We do not know whether other recruiting activities of Gal4 remain functional , but if so , it is possible that glucose somehow causes destruction of any transcription complex that might be recruited to the promoter . This notion would be consistent with previous suggestions that certain mutations in that complex can diminish the negative effects of glucose [33–36] . Our analysis has equated protection from nuclease digestion with nucleosome occupancy ( excluding the exceptional case of the UASg ) . Thus , for example , where we find DNA locations that yield biphasic digestion curves , indicating two subpopulations , we have identified the slow digesting fraction as occupied by a nucleosome , the fast digesting portions as naked , i . e . , simply lacking nucleosomes . And , we have argued , that the progressive increase in the fraction naked , as induction proceeds , reflects nucleosome loss . A more complicated description for the naked fraction could be imagined . Thus , for example , perhaps prior to induction , the naked regions bear nucleosomes in some aberrant configuration that would expose a segment of DNA so as to render it “naked” in our experiments , and according to this notion , as induction proceeds , instead of falling off , the nucleosomes increasingly adopt that aberrant configuration . It is difficult to completely exclude such a scenario . However , our ChIP analysis , which probed for histone H2B ( Figure 2E ) , as well as ChIP analyses probing additionally for histone H3 and H4 [15 , 17] , all show a clear drop in each of these histone signals upon induction . Also , our finding that as many as 50% of the promoter nucleosome sites register as naked prior to induction is consistent with other studies indicating a low nucleosome density at various yeast promoters , and it has been reported that , for many yeast genes , promoter nucleosomes turn over more rapidly than do ORF nucleosomes [37 , 38] . Taken together , these results are simply explained by the idea that promoter nucleosomes are often vacant prior to induction and increasingly so as induction proceeds . It is possible , but not directly demonstrated , that the relative absence of promoter nucleosomes prior to induction is determined by the intrinsic sequence of those nucleosome-forming sites . The fact that promoter nucleosomes must be removed for rapid activation of transcription indicates that even the relatively infrequent formation of these nucleosomes ( compared , for example , to that observed for the ORFs ) suffices to significantly compete with formation of the large , multicomponent transcription complex . Our assay measures two aspects of nuclease protection conferred by a DNA bound molecule: the location of the bound molecule , the fraction of the population that bears it , and the degree of protection it imparts . Most DNA segments , as we have seen , yield biphasic digestion curves , and for segments bearing positioned nucleosomes , the ratio of the fast-digesting and slow-digesting fractions is very large , invariably close to a value of 200 . Two unusual cases illustrate further how we can separate the degree of protection from the fractional occupancy . First , as we have seen , the molecule occupying 100% of the UASg's in the population imparts a degree of protection some 10-fold less than that imparted by a nucleosome , and indeed this property is one of the indications that it is not a nucleosome . Second , we have introduced into the GAL locus a sequence predicted to have a high propensity to form a nucleosome ( [39] and E . Segal and J . Widom , personal correspondence ) . This segment indeed forms a nucleosome ( as measured in a ChIP experiment ) . In this case , 100% of the population is occupied , and the degree of protection is just that found for the typical nucleosome ( i . e . , some 200-fold ) ( X . Wang , G . O . Bryant , and M . Ptashne , unpublished data ) . Our results do not exclude the possibility that nucleosomes , even positioned nucleosomes , can have some small degree of mobility along the DNA . In fact , the HS sites that lie between positioned nucleosomes are , as we have noted , naked to about the 90% level prior to induction , a value that decreases still further upon induction . Thus , perhaps , even the positioned nucleosomes can vary a few base pairs in their exact location in different members of the population .
Strains , both wild type and deletions ( except gal4Δ and FLAG-tagged H2B [40] ) , were derived from BY4741 ( MATa his3Δ1 leu2Δ0 met15Δ0 ura3Δ0 ) obtained from EUROSCARF ( European Saccharomyces Cerevisiae Archive for Functional Analysis ) . Additional strains used were CY1069 ( snf2Δ ) and its corresponding wild type [20] . We also added a wild-type SNF2 expression plasmid , pM4724 [41] , to both the BY4741 snf2Δ derivative and to CY1069 . For all experiments reported , cells were grown exponentially for at least 16 h at 30 °C prior to harvesting in synthetic complete medium ( SC ) or , in the case of snf2Δ and its wild-type control , in yeast extract peptone medium ( YP ) . All sugars were added at a final concentration of 2% unless otherwise indicated . For the galactose induction experiments , prewarmed and aerated galactose was added directly to the media . In cases in which harvested time points where less then 5 min apart , medium containing 4% galactose was added at a one-to-one ratio to the growing cells . For the glucose repression experiments , exponentially growing cells were precipitated , washed with the original medium , and then added to prewarmed and aerated glucose containing medium under conditions in which the original medium was diluted greater then 100-fold . For all experiments , cells were harvested at an optical density at 600 nm ( OD600 ) between 0 . 5–0 . 9 by fixing the cells with freshly diluted formaldehyde at a final concentration at 0 . 5% for 1 to 5 min . The fixing reaction was stopped by addition of glycine at a final concentration of 0 . 125 M . DNA and cDNA were quantitated by real-time quantitative polymerase chain reaction ( real-time PCR ) . A 2× reaction buffer: 20 mM Tris-HCl ( pH 8 . 3 ) , 13 mM MgCl2 , 100 mM KCl , 400 μM dNTPs , 4% DMSO , 2× SYBR Green I ( Molecular Probes ) , 0 . 01% Tween 20 , 0 . 01% NP40 , 1–4 ng/μl of each oligo primer , and 0 . 025–0 . 1 U/μl of Taq polymerase ( Roche ) was mixed with an equal volume of DNA being quantitated , resulting in a total reaction volume of 5 μl . The time that the oligo primers are mixed with the 2× reaction buffer is limited to less then 10 min before it is mixed with the DNA sample to limit primer dimer formation . A typical real-time PCR reaction measured 80 unknown samples plus 16 known samples ( two standard curves ) in quadruplicate , i . e . , 96 samples transferred into four different positions of a 384-well plate . Seven of the eight samples within a standard curve consisted of 3 . 33-fold dilution series of yeast chromosomal DNA . The final sample of the standard curve contained no DNA . All real-time PCR reactions were performed on the Light Cycle 480/384 from Roche . Reactions were run for 40 to 50 cycles ( depending on the primer pair ) at 95 °C for 4 s , 59 °C for 26 s , and 72 °C for 4 s with the florescence of the SYBR Green being read at the 72 °C step . Since the specific activity of Taq polymerase varied considerable from lot to lot , care was taken to test by titration each batch of Taq to find its optimal concentration . Outliers of quadruplicate measurements were eliminated if dropping one of the four measurement reduced the standard deviation by greater then 2-fold and the original standard deviation was above the 50th percentile for the plate . An average quadruplicate measurement was eliminated if it was not greater then 2-fold above the measured value of the no DNA control . The DNA concentration was then determined by comparing real-time PCR measured values to a linear fit of the known chromosomal concentrations . mRNA was isolated from 10 ml ( growing volume ) of cells by a modified version of the hot acidic phenol technique [42] in which the 65 °C incubation step was extended to 3 h to ensure that the formaldehyde crosslink was completely reversed . One twentieth of the isolated mRNA was reverse transcribed by AMV reverse transcriptase ( Roche ) as per the manufacturer's instructions . cDNA was quantitated by real-time PCR ( see above ) at GAL1 , GAL10 , GAL7 , GAL3 , and GAL2 and for controls HHF1 , ACT1 , and RBP11 ( see Primer Pair List , Table S1 ) . The three control genes were used to normalize the varying yields of cDNA from each sample . ChIP assays were performed as described [1] . Reaction conditions: 100–200 ml of a yeast cell culture were spun down and resuspended in 500 μl of FA lysis buffer without EDTA: 50 mM Hepes-KOH ( pH 7 . 5 ) , 140 mM NaCl , 1% Triton X-100 , 0 . 1% sodium deoxycholate . The resuspended cells were sonicated twice for 10 s using a Branson Sonifier 250 equipped with a micro tip with the output set at 4 . Cell debris was then spun down and the chromatin supernatant transferred to a new tube . A total of 26 μl was then distributed to 16 separate tubes , and 120 μl of FA lysis buffer without EDTA was added . To each tube , 10 μl of a micrococcal nuclease solution in H2O was added at a range of concentrations from 4 U to 0 . 000488 U in a 2-fold dilution series; two tubes had no nuclease . The reaction was started by adding 5 . 6 μl of 2 mM CaCl2 to each tube . The reactions were incubated for 1 . 5 h at 37 °C and stopped by the addition of 8 . 8 μl of 0 . 5 M EDTA each . Ten microliters of a solution containing 200 mM Tris ( pH 7 . 4 ) , 4 M NaCl , and 0 . 2 μl of Protease K ( recombinant , Roche ) were added to each tube , followed by incubation at 42 °C for 1 h and at 65 °C for at least 4 h . The DNA was purified using the QIAquick 96 PCR Purification Kit ( Qiagen ) . DNA was typically quantitated at 16 or more positions ( see Primer Pair List in Table S1 ) near the GAL genes ( GAL1 , GAL10 , GAL7 , GAL2 , and GAL3 , along with three positions within the UASg ) and eight loci near the control genes TUB2 and PHO5 . For purposes of discussion , the DNA from each level of micrococcal digestion will be referred to as a sample . All 16 samples from the same harvest digested at the range of micrococcal nuclease concentrations will be referred to as a group . The real-time PCR-measured value for each sample at each position ( locus ) will be referred to as the sample-locus value , and the group of values at each locus as group-locus values . For each group-locus , the undigested DNA concentration was rescaled to one by dividing each sample-locus value of a group-locus by the average of the undigested sample-locus values . To compensate for the varying yields of DNA for each sample within a group ( e . g . , differing efficiencies of DNA recovery in the DNA purification step ) , the rescaled values for each group-locus measured at the UASg was fit to the one-state decay function e− ( kMN ) , where MN is the concentration of micrococcal nuclease and k is the adjustable parameter representing the rate of digestion . Each rescaled sample-locus is normalized by dividing it by the average of the sample-locus value ( at the UASg ) divided by its calculated one-step curve fit value . The normalized rescaled values for each group-locus is then fit to the two-state decay function ( 1 − fr2 ) e− ( k1MN ) + fr2 e− ( k2MN ) , where MN is the concentration of micrococcal nuclease , k1 is the adjustable parameter representing the rate of digestion of the unprotected DNA , k2 is the adjustable parameter representing the rate of digestion of the protected DNA , and fr2 is the adjustable parameter representing the fraction of the DNA that is protected . The curve is fit by adjusting all three parameters to minimize the sum of the squares of the differences between the two-state decay function and the normalized rescaled group-locus values , where k1 is at least 50-fold greater than k2 , and k1 is no less than a cutoff value set for each group ( this cutoff value is typically 10–30 times greater than the protection seen within the UASg for the group ) . Sample-locus outliers from this fit were eliminated if the absolute difference of the normalized rescaled sample-locus value compared to its respective curve was greater then five times the average absolute difference for the entire group . The curves were then fit again to the remaining data points as described above . The error for each adjustable parameter was calculated by incrementally adjusting the parameter away from its best fit while allowing the other two parameters to adjust to their minimum until the sum of the squares of the differences increased by greater than 10% . A slight systematic variation in fr2 values was seen at the control loci ( the variation was less then 15% ) . To correct for this , the average value for each control was assumed to be its true value . For each group , the measured fr2 values at the control loci were plotted against their average values , and a curve fit was then performed on this plot . The curve used was the line segments defined by the points ( ( 0 , 0 ) , ( x , y ) ) and ( ( x , y ) , ( 1 , 1 ) ) , where the x-axis is the measured fr2 control values and the y-axis is the average control values . A least-squares fit was performed by adjusting x and y under conditions in which the slope of either line segment was between 0 . 5 and 2 . The fit curve was then used to rescale all fr2 measurements from the group . | In this paper , we examine activation and repression of transcription of a gene in yeast . This gene , like the typical human gene , is wrapped in DNA-protein packets called nucleosomes . It is widely believed that these condensed packets are unwrapped , in a process called nucleosome removal , as transcription begins . Here , we describe a new quantitative nucleosome assay that allows us to measure the time course of nucleosome removal and replacement as the gene is activated and repressed . The yeast activator Gal4 , bound to DNA , effects activation of gene transcription in two separate steps . First , it recruits to the gene an enzyme that strips off nucleosomes; and second ( as we had shown previously ) , it recruits the transcriptional machinery . We also show that transcription of the gene can be turned off well before nucleosomes have been returned to the gene . In this case , the activator continues to recruit the nucleosome-remover , but either the transcriptional machinery is not recruited , or if it is , it is soon destroyed . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"molecular",
"biology"
] | 2008 | Activator Control of Nucleosome Occupancy in Activation and Repression of Transcription |
The aim of this study is to model the association between weekly time series of dengue case counts and meteorological variables , in a high-incidence city of Colombia , applying Bayesian hierarchical dynamic generalized linear models over the period January 2008 to August 2015 . Additionally , we evaluate the model’s short-term performance for predicting dengue cases . The methodology shows dynamic Poisson log link models including constant or time-varying coefficients for the meteorological variables . Calendar effects were modeled using constant or first- or second-order random walk time-varying coefficients . The meteorological variables were modeled using constant coefficients and first-order random walk time-varying coefficients . We applied Markov Chain Monte Carlo simulations for parameter estimation , and deviance information criterion statistic ( DIC ) for model selection . We assessed the short-term predictive performance of the selected final model , at several time points within the study period using the mean absolute percentage error . The results showed the best model including first-order random walk time-varying coefficients for calendar trend and first-order random walk time-varying coefficients for the meteorological variables . Besides the computational challenges , interpreting the results implies a complete analysis of the time series of dengue with respect to the parameter estimates of the meteorological effects . We found small values of the mean absolute percentage errors at one or two weeks out-of-sample predictions for most prediction points , associated with low volatility periods in the dengue counts . We discuss the advantages and limitations of the dynamic Poisson models for studying the association between time series of dengue disease and meteorological variables . The key conclusion of the study is that dynamic Poisson models account for the dynamic nature of the variables involved in the modeling of time series of dengue disease , producing useful models for decision-making in public health .
Dengue is an arboviral disease caused by a Flavivirus , leading to high morbidity in children and adults in tropical countries of Asia and Latin America [1] . There are four genetically distinct but antigenically related ( different serotypes ) dengue viruses named DEN-1 , DEN-2 , DEN-3 , and DEN-4 . All serotypes can cause a spectrum of illness ranging from unapparent or mild fever to the potentially fatal syndrome characterized by hemorrhage , fever and shock syndrome [2] . The infective female Aedes aegypti mosquito is the main vector involved in transmiting the viruses causing dengue . The mosquito acquires the virus when it feeds on the blood of an infected human . Several studies show that climate is associated with the mosquito ecology , the infectious agents they carry , and the arboviral transmission of dengue disease [3] [4] [5] . Naish et al . ( 2014 ) [3] reviewed the studies associating climatic factors and dengue transmission , concluding that higher temperatures affect the rate of larval development , shorten the emergence of adult mosquitoes , increase the biting behavior of mosquitos , and accelerates virus replication within the mosquitos . Meanwhile , the combined effect of temperature and relative humidity impact mosquito feeding behavior , vector survival and the probability to be infected and the ability to transmit dengue . Epidemiological research on dengue incidence is based on passive surveillance data from case reports [5] [6] . Racloz et al . ( 2012 ) [5] reviewed early warning modelling in dengue disease , concluding that epidemiological modeling is constrained by limited data sources . Authors encouraged the collection of information at the spatial and temporal level of climatic and socio environmental variables to develop models with stronger predictive capabilities , while Runge-Ranzinger et al . ( 2014 ) [6] concluded that passive surveillance provides the baseline for outbreak alert , which should be strengthened through the definition of appropriate alert thresholds . For the time series analysis of dengue case counts associated with meteorological variables , diverse methodologies have been employed , including auto-regressive integrated moving average ( ARIMA ) models [7] [8] [9] [10] [11] [12] [13] [14] [15] , Poisson multivariate regression forecasting models [16] [17] [18] , distributed lag non-linear models ( DLNM ) [19] [20] , decision trees with cross-validation [21] , multiresolution analysis and fuzzy systems [22] , stepwise negative binomial multivariate linear regression analysis [23] , wavelet time series analysis [24] , probabilistic random walks [25] [26] , and dynamic generalized linear models ( DGLM ) [27] [28] [29] . DGLMs are extensions of the dynamic linear models [30] [31] , based on two sets of equations , a measurement or observation equation and the transition or state equations . The observation equation establishes a link between observations and unobserved variables , and the transition equations describe the evolution of state variables . DGLMs allow the inclusion of components modeling seasonality , trend , cyclicity and covariates [31] . The classic models for calendar trend are the first-order random walk model , the local linear trend model ( first-order random walk plus trend ) and the second-order random walk [32] . Modeling seasonality and cyclicity is accomplished through dummy variables or trigonometric series defined in the transition equations , and covariates are included with constant or time-varying coefficients [32] . DGLM parameter estimations have followed different approaches . Linear Bayes estimation with conjugate updating [30] [31] or iteratively weighted Kalman filter and smoother , accompanied by the expectation-maximization ( EM ) algorithm for the estimation of unknown hyperparameters [32] , was applied by Chiogna and Gaetan [33] to explore the association between pollution covariates and respiratory diseases . Shepard et al . [34] applied likelihood base inference for non-Gaussian state space parameters , based on importance sampling . DGLMs estimated by Markov Chain Monte Carlo ( MCMC ) simulations have been explored by Gamerman [35] , Ferreira and Gamerman [27] ( modeling Dengue disease and meningitis with covariates and seasonal terms ) , Schmidt and Pereira [28] and Alves et al . [36] including covariates with constant coefficients for time accompanied by covariates modeled by transfer functions . Malhão et al . [29] implemented DGLM for time series of dengue cases , capturing temporal dependencies not explained by covariates , and modeling dengue over-mortality . Colombia is one of the countries with the highest incidence of dengue disease in the tropics , and it is testing dengue control by vaccination [37] , a topic of interest among the research community [38] . The country possesses climatic , environmental and socio-geographic conditions favoring the growth and development of the dengue vector . The Aedes aegypti mosquito is found across more than 80% of the territory , which has an altitude of 1000 m and 2200 m above sea level , and the Aedes albopictus ( forest and urban dengue vector ) has also been reported [39] . Bucaramanga is among the Colombian cities with the highest annual dengue incidence for the 2008–2015 period . In 2010 and 2012 the city experienced incidence rates of 1515 and 279 . 93 cases per 100 , 000 people , respectively , while for the same years the incidence rates for the country were 657 and 221 . 9 cases per 100 , 000 , respectively [39] [40] . The Aedes aegypti mosquito has been reported as the dengue vector in the city of Bucaramanga . While vectorial surveillance studies did not exist in 2008–2015 to quantify the presence of vectors , their abundance , occurrence , distribution and other epidemiological parameters at monthly or weekly temporal scales for Bucaramanga , information of climatic variables such as environmental temperature , rainfall , solar radiation , and relative humidity are available from several sources at these temporal scales . These data offer opportunities to analyze the relation between time series of dengue cases and climatic variables , as Rúa-Uribe et al . ( 2013 ) [8] show for another Colombian city . The aim of this study is to model the association between time series of dengue case counts and meteorological variables , in a high-incidence city of Colombia , applying Bayesian hierarchical dynamic generalized linear models , during the period January 2008 to August 2015 . Additionally , we evaluate the model’s performance in short-term prediction of dengue cases .
Bucaramanga is a medium-sized city in Colombia , at 959 meters above sea level , with a population of 527 , 913 people ( projected population , 2015 ) , at the coordinates 7°07′07″N , 73°06′58″W . We collected dengue case counts for 2008–2015 in metropolitan Bucaramanga from the Surveillance National System of Public Health ( SIVIGILA ) . The total dengue case counts ( probable and confirmed cases of dengue and severe dengue plus dengue mortality ) by epidemiological week ( EW ) were computed in the interval between the first EW of January 2008 to the last EW of August 2015 , for a total of 396 EW . For the meteorological variables ( MV ) , daily maximum temperature ( °C ) , daily total rain fall ( mm ) , daily maximum solar radiation ( Watts/m2 ) and daily maximum relative humidity ( % ) were obtained from three stations of the Defense Corporation of the Bucaramanga Plateau ( CDMB ) . Daily maximum temperature ( °C ) and daily total rain fall ( mm/m2 ) were obtained from the Institute of Hydrology , Meteorology and Environmental Studies of Colombia ( IDEAM ) for two meteorological stations . Daily values for every variable were averaged by EW and by station , and then the weekly averages of all stations were averaged , obtaining one value per MV and EW . We fitted Bayesian hierarchical dynamic Poisson models to dengue case counts . Let yt be the case count for dengue in EW t ( t = 1 , ⋯ , T and T = 396 ) , and y t ∼ Poisson ( λ t ) ( 1 ) The logarithm of the mean λt is modeled with two options . The first option is the inclusion of a constant coefficient α for the calendar trend , log ( λ t ) = {α α + ∑ j = 1 J β j x t - 1 , j α + ∑ j = 1 J b t , j x t - 1 , j ( 2 ) where α is Normal with mean 0 and variance 10 , which allows flexibility for the exploration of the parameter space . The second option is the inclusion of time-varying coefficients αt for the calendar trend , log ( λ t ) = {α t α t + ∑ j = 1 J β j x t - 1 , j α t + ∑ j = 1 J b t , j x t - 1 , j ( 3 ) where the time-varying coefficients αt are defined with Normal random walk 1 ( RW1 ) or Normal random walk 2 ( RW2 ) priors . The Normal RW1 priors for αt are defined as α 1 ∼ Normal ( 3 , 0 . 2 ) α t ∼ Normal ( α t - 1 , τ α ) ; ( 2 ≤ t ≤ T ) and the Normal RW2 priors for αt follow α 1 , α 2 ∼ Normal ( 3 , 0 . 2 ) α t ∼ Normal ( 2 α t - 1 - α t - 2 , τ α ) ; ( 3 ≤ t ≤ T ) where for the Normal ( 3 , 0 . 2 ) prior , the mean of 3 for α1 and α2 in the exponential scale is close to the observed dengue case counts at time points 1 and 2 , and 0 . 2 is a precision ( variance of 20 ) that allows flexibility for these parameters . τα is the precision parameter with Gamma ( 1 , 0 . 1 ) hyperprior , which represents a Gamma prior noninformative distribution centered at 10 with variance of 100 . In Eqs 2 and 3 , the xt−1 , j ( j = 1 , ⋯ , J and J = 4 ) are the mean centered MVs temperature ( j = 1 ) , rainfall ( j = 2 ) , solar radiation ( j = 3 ) and relative humidity ( j = 4 ) . The βj are constant coefficients for lag-one MV , and bt , j are time-varying coefficients for lag-one MV . Normal priors with mean 0 and variance 10 were assigned to the constant coefficients β for the covariates . The time-varying coefficients for the lag-one covariates received first-order Normal RW1 priors , b 1 , j ∼ Normal ( 0 , 0 . 1 ) b t , j ∼ Normal ( b t - 1 , j , τ b j ) ( 2 ≤ t ≤ T ) where for the Normal ( 0 , 0 . 1 ) , we let b1 , j start centered at zero , with a 0 . 1 precision ( variance of 10 ) , allowing a large space for exploring the parameter . Gamma ( 1 , 0 . 001 ) prior distributions ( Gamma centered at 1000 with variance of 100 , 000 ) are assigned to the precision parameters τbj . The reason for this prior is that we constrain the variance of the bt , j to be very small , smoothing the trend of the time-varying coefficients and allowing us to visualize the smoothed trend of the covariate effects . We modeled missing data in the covariates by imputing the empty values , assuming a Normal ( μt−1 , τj ) prior for t = 1 , ⋯ , T and T = 396 , where μt−1 is the value of the lag-one week meteorological centered variable , where τj is a precision parameter with Gamma ( 0 . 1 , 0 . 1 ) priors for temperature , for rainfall , solar radiation , and relative humidity , where the Gamma prior is an informative prior centered at 0 . 1 with dispersion 10 , slightly constraining the imputed values of the covariates to have a small variance , without restricting to high variance values . Models were fitted applying MCMC using WinBUGS 1 . 4 software [41] , with 3 chains , 50 , 000 iterations total , 46 , 000 iterations burn-in and thinning of 4 , obtaining a final sample of 1000 iterations per chain . Convergence was assessed by Gelman-Rubin diagnostic [42] and visual inspection of the simulations chains . Model selection was accomplished using deviance information criteria ( DIC ) [43] . When DIC measures are used for model selection , models with small deviance D ¯ , a small number of parameters pD and a small DIC are selected for inference . After fitting all models , and selecting the final model for inferences , we were interested in evaluating the short-term prediction performance of the selected final model . We obtained predictions at several time points , during the study period T = 396 . We selected estimation periods 1 to t , where t was in increments of 20 EWs , starting in the 20th EW of the study period and ending in the 380th EW . We obtained 19 upper bounds for the estimation period 1 to t . Then we fitted models for periods 1 to p , where p = t + k ( k = 1 , ⋯ , 4 ) , and the k are prediction periods ( one , two , three or four weeks ahead ) . We used the same conditions defined above for the MCMC simulations . Samples from the posterior predicted distribution for the prediction periods k were obtained , and the mean and 95% credible intervals ( CIs ) for the cases of dengue were calculated . To evaluate the prediction performance from the final model , we calculated the mean absolute percentage error ( MAPE ) per MCMC iteration between the predicted cases of dengue ypredk and the observed case count yk , at prediction periods k ( ∑k | ( ypredk − yk ) /yk|/k ) . We present the median MAPE of the posterior predictive distribution for all the estimation periods t for one , two , three and four weeks ahead as a measure of short-term model performance for predicting dengue case counts .
The total number of cases of dengue disease for the study period was 26 , 755 . The weekly case count averaged 67 . 6 , with a median of 52 ( range 7 to 247 ) . There were three dengue disease outbreaks in 2010 , 2013 and 2014 , with small case counts in 2011 and 2012 ( Fig 1 ) . The partial autocorrelation function for the time series of dengue case counts ( Fig 1 ) suggest a first- or second-order autoregressive process . Maximum weekly temperature averaged 27°C , with a minimum of 23 . 6°C , a maximum of 30 . 4°C , and 18 missing values . Mean and median values of weekly rainfall were 2 . 7 mm/m2 and 3 . 6 mm/m2 , respectively , with a minimum of 0 , a maximum of 24 . 8 mm/m2 , and 11 missing values . Weekly maximum solar radiation averaged 946 . 5 Watts/m2 , with median of 940 . 9 Watts/m2 , a minimum of 733 . 5 Watts/m2 , a maximum of 1279 Watts/m2 , and 66 missing values . Maximum weekly relative humidity averaged 94 . 2% , with a minimum of 79 . 2% , a maximum of 99 . 5% , and 63 missing values . Fig 2 shows plots of time series for MVs , and plots of the average dengue case counts by intervals of the MVs . While time series for temperature and relative humidity display an upward trend over the 396 EWs , solar radiation decreases , and precipitation shows highly volatile behavior . Dengue disease case counts are positively correlated with temperature , and negatively correlated with solar radiation . There is no apparent association between dengue case counts and precipitation or relative humidity . In Fig 3 , linear correlations between the meteorological variables and dengue case counts show positive and moderate correlation with temperature and negative and moderate linear correlation with relative humidity , solar radiation and rainfall . Relative humidity and solar radiation display high positive correlations with their own lag-1 and lag-2 values , followed by temperature and rainfall . Rainfall , relative humidity and solar radiation are positively and moderately correlated , while rainfall and temperature show negative and moderate correlation . Finally , we highlight the negative and low correlation between solar radiation and temperature . In this section , we begin by presenting the results from the models without covariates ( only constant coefficient ( CC ) ( α ) or RW1 or RW2 time-varying coefficients ( TVCs ) ( αt ) for calendar trend ) . We define calendar trend as the pattern observed in the model’s parameters over the EWs in the entire study period ( 2008–2015 ) , not the trends observed over any given epidemiological year . We then present the results from models including CC ( βj ) for covariates , and CC ( α ) or RW1 or RW2 TVCs ( αt ) for calendar trend . Finally , we exhibit the results from models including RW1 TVCs ( bt , j ) for the covariates with CC ( α ) or RW1 or RW2 TVCs ( αt ) for calendar trend .
In this report , DGLMs are employed to model time series of dengue disease case counts and meteorological variables . DGLMs for the data at hand included two components: the first substracts the temporal pattern , and the second models the covariate effect . We observed weak time-varying associations between cases of dengue disease and solar radiation and temperature . Time-varying associations mean that the dengue case counts are associated with solar radiation and temperature changes over time , where some intervals show a positive association , while in other intervals the association is negative . DGLMs are a straightforward way to deal with count data , without the need to transform or alter the response variable , accounting for covariates with natural time-varying behavior . For parameter estimation , we applied MCMC using WinBUGS 1 . 4 , providing the flexibility to include constant and time-varying coefficients for calendar trend and covariates . There are few examples of studies including time-varying coefficients . Lee and Shaddick ( 2008 ) [44] fit DGLMs to pollution data and respiratory diseases , based on the block sampling algorithm from Knorr-Held ( 1999 ) [45] . Ruiz-Cardenas et al . ( 2012 ) [46] employed Integrated Laplace Approximation ( INLA ) to illustrate the fit of simulated and real time series of counts , using augmented data with the inclusion of time varying-coefficients for calendar trend and covariates . Our findings can be summarized as follows: in the models without covariates , the best model was the RW1 TVCs ( α ) for trend . Within the models with CC ( βj ) for covariates , we found the worst fit in models with CC ( α ) for trend , which display strong association ( 95% CIs not including zero ) between weekly cases of dengue and temperature , solar radiation and rainfall , but not with relative humidity . However , models with RW1 or RW2 TVCs ( αt ) for calendar trend had a good fit , revealing a weak association between dengue and the covariates . These findings are important because simple and multiple Poisson regression models with constant coefficients for the covariates are statistical methods commonly employed to model counts of infectious diseases like dengue [4] . For example , Hii et al . [16] modeled dengue and weather variables , applying a Poisson multiple regression model with piecewise linear spline functions for the covariates and constant coefficient terms to model autoregression , seasonality and trend . They validated the model by forecasting cases of dengue for week 1 of 2011 up to week 16 of 2012 using weather data alone . In the class of models with RW1 TVCs ( bt , j ) for the covariates , the best model corresponds to the simple dynamic Poisson model with RW1 TVCs ( αt ) for calendar trend . After fitting the simple dynamic regression models , we fitted multiple dynamic regression models , with several combinations of TVCs ( bt , j ) for the covariates , and we selected the model including all the meteorological variables . Our final model delineates the time-varying association between the covariates and cases of dengue , although the inspection of the mean estimates and 95% CIs of the RW1 TVCs ( bt , j ) for the covariates shows a weak association . In the literature associating dengue and weather variables , many of the modeling strategies show strong association ( evidenced by low p-values ) between dengue and meteorological variables , with different lag periods . As an example , Xu et al . [19] established an association between absolute humidity ( relative humidity adjusted by temperature ) and dengue cases using a Poisson distributed lag non-linear model , with cubic splines for the covariates and accounting autoregression with constant coefficients for the lag-one and lag-two response . We also evaluate the short-term predictive performance of the selected model , concluding that it enables relatively accurate ( < 25% error ) prediction of weekly dengue case counts at one or two weeks ahead although the predictions are strongly influenced by volatility in the weeks preceding the prediction periods , with high volatility associated with high MAPE in the predictions , as occurred in the peak of the 2010 , 2013 and 2014 outbreaks in Bucaramanga . Before finishing our discussion , we acknowledge some study limitations . The dengue case counts used in the data corresponded to the probable and confirmed cases reported to the official public health surveillance system in Colombia . The weekly dengue data was the sum of the the dengue and severe dengue cases per EW . Romero-Vega et al . ( 2014 ) [47] concluded that the expansion factor ( the factor by which the reported cases should be multiplied to adjust for underreporting ) of dengue was 7 . 6 for 2013 , which is high . This implies that efforts to decrease underreporting must be undertaken to improve data quality for the entire surveillance system . It would be difficult to quantify the impact of underreporting in our conclusions , but still , the methods we used are valid for adjusted time series of dengue . The covariates data ( time series of temperature , rainfall , solar radiation and temperature ) were a composition of several time series at daily and hourly temporal scales from several meteorological stations at different locations in the city . We summarized the data , averaging them for the different temporal scales and stations and consequently losing some data . However at some point the analyst must decide how to summarize the information to input variables for a modeling exercise . If the temporal scale is reduced ( from weekly to daily data ) the dengue case counts will be lower , and the Poisson models presented in the study could fit the data much better than Normal models . One of this study’s referees remarked on the absence of vector data in the study . We explored several sources of vector data in the city , but we did not find any data at the temporal scale of the study . We recognize that the inclusion of data for the distribution , presence and ecology of the vector would improve the conclusions of the study , but this is an opportunity to show that dengue in Colombia , and particularly in Bucaramanga , is a neglected disease , despite its huge impact on the population and the allocation of resources for dengue research ( Villabona-Arenas et al . , 2016 ) [38] . One interesting experience in ongoing vectorial surveillance is in the city of Medellín , Colombia . Rúa-Uribe ( 2016 ) [48] reported that the Health Office of this city designed an entomological surveillance system using mosquito larval traps . We hope that the results of this interaction between the public sector and the research community will be disseminated to the country , and similar surveillance systems will be applied in all Colombian cities affected by arboviral diseases . In the mean-time , for the city of Bucaramanga , we applied a dynamic Poisson model with time-varying coefficients for the covariates and calendar trend , which helps to establish the association between climatic factors and dengue case counts at a small temporal scale , providing a prediction model within the bounds of the limitations presented in the study . Forecasting models are commonly deployed in dengue research literature . Earnest et al . [10] compare the forecasting ability of the ARIMA model and the two-component Knorr-Held model ( seasonal and epidemic Bayesian hierarchical time series model ) to predict out-of sample cases of dengue . They found similar predictive ability ( lower MAPE values ) for the Bayesian K-H model and the ARIMA model . Forecasting models of dengue disease usually account cyclical or seasonal behavior of the time series at hand . Earnest et al . [10] and Hii et al . [16] included seasonal trend by means of sinusoidal terms with trigonometric series structure . In a previous stage , we included seasonal terms , but we removed them from the models , allowing the time-varying coefficients for calendar trend alone account for dengue incidence trends . We establish the short-term predictive performance of a model with time-varying coefficients ( αt ) for calendar trend and time-varying coefficients ( bt , j ) for meteorological covariates . We found a moderate predictive ability from the model to forecast cases of dengue disease at one or two weeks , which could be used by public health authorities interested in employing predictive models to help in the labors of dengue surveillance and control in Colombia . For the future , we will explore the study models in different datasets from other cities of Colombia because , the enviromental and physical conditions are generally similar between many cities and municipalities . The models presented in the study are not only available for use with climatic variables . They can also include data from vectorial studies , socioeconomic variables and many more , if these are available at weekly or monthly temporal scales . In conclusion , we found that dynamic generalized linear models can forecast dengue cases at one or two weeks in Bucaramanga , based on temperature , rainfall , solar radiation and relative humidity , and the models allow us to explore the association between weekly cases of dengue and these covariates through the time . | Time series analysis of dengue disease case counts are currently employed to establish associations between dengue disease and environmental , socioeconomic and climatic variables and to predict the evolution of dengue epidemics . Nowadays there is acceptance that climatic factors like environmental temperature , rainfall and relative humidity modify the behavior of the dengue vectors , affecting the transmission of the disease . Thus , in the absence of vector data , climatic factors are commonly used to input transmission models of dengue disease on several temporal and spatial scales . We applied hierarchical Bayesian dynamic generalized models to dengue diseases case counts in a medium-sized city in Colombia , with constant and time-varying coefficients for calendar trend , and constant and time-varying coefficients for meteorological variables ( temperature , rainfall , solar radiation and relative humidity ) . We selected a final model useful for exploring of the time-varying association between climatic variables and dengue , and the short-term out-of-sample predictions of dengue counts within the study period . We illustrate the modeling process so a data analyst on a multidisciplinary research team could integrate a time series model accounting for the time-varying nature of the data . | [
"Abstract",
"Introduction",
"Materials",
"and",
"methods",
"Results",
"Discussion"
] | [
"medicine",
"and",
"health",
"sciences",
"atmospheric",
"science",
"geographical",
"locations",
"mathematical",
"models",
"electromagnetic",
"radiation",
"mathematics",
"forecasting",
"statistics",
"(mathematics)",
"atmospheric",
"dynamics",
"humidity",
"research",
"and",
"analysis",
"methods",
"random",
"walk",
"south",
"america",
"epidemiology",
"mathematical",
"and",
"statistical",
"techniques",
"physics",
"people",
"and",
"places",
"solar",
"radiation",
"rain",
"colombia",
"meteorology",
"earth",
"sciences",
"disease",
"surveillance",
"geophysics",
"physical",
"sciences",
"atmospheric",
"physics",
"statistical",
"methods"
] | 2017 | Bayesian dynamic modeling of time series of dengue disease case counts |
As the more recent next-generation sequencing ( NGS ) technologies provide longer read sequences , the use of sequencing datasets for complete haplotype phasing is fast becoming a reality , allowing haplotype reconstruction of a single sequenced genome . Nearly all previous haplotype reconstruction studies have focused on diploid genomes and are rarely scalable to genomes with higher ploidy . Yet computational investigations into polyploid genomes carry great importance , impacting plant , yeast and fish genomics , as well as the studies of the evolution of modern-day eukaryotes and ( epi ) genetic interactions between copies of genes . In this paper , we describe a novel maximum-likelihood estimation framework , HapTree , for polyploid haplotype assembly of an individual genome using NGS read datasets . We evaluate the performance of HapTree on simulated polyploid sequencing read data modeled after Illumina sequencing technologies . For triploid and higher ploidy genomes , we demonstrate that HapTree substantially improves haplotype assembly accuracy and efficiency over the state-of-the-art; moreover , HapTree is the first scalable polyplotyping method for higher ploidy . As a proof of concept , we also test our method on real sequencing data from NA12878 ( 1000 Genomes Project ) and evaluate the quality of assembled haplotypes with respect to trio-based diplotype annotation as the ground truth . The results indicate that HapTree significantly improves the switch accuracy within phased haplotype blocks as compared to existing haplotype assembly methods , while producing comparable minimum error correction ( MEC ) values . A summary of this paper appears in the proceedings of the RECOMB 2014 conference , April 2–5 .
While human and other eukaryotic genomes typically contain two copies of every chromosome , plants , yeast and fish such as salmon can have strictly more than two copies of each chromosome . By running standard genotype calling tools , it is possible to accurately identify the number of “wild type” and “mutant” alleles ( A , C , G , or T ) for each single-nucleotide polymorphism ( SNP ) site . However , in the case of two heterozygous SNP sites , genotype calling tools cannot determine whether “mutant” alleles from different SNP loci are on the same or different chromosomes ( i . e . compound heterozygote ) . While the former would be healthy , in many cases the latter can cause loss of function; it is therefore necessary to identify the phase ( phasing ) —the copies of a chromosome on which the mutant alleles occur—in addition to the genotype ( Figure 1 ) . This necessitates efficient algorithms to obtain accurate and comprehensive phase information directly from the next-generation-sequencing read data in higher ploidy species . Various sources of information can be utilized for the computational identification of an individual's diplotype/polyplotype: pedigree ( e . g . trio-based phasing ) [1]–[3] , population structure of variants ( e . g . phasing by linkage disequilibrium ) [3]–[6] and more recently by identity-by-descent in unrelated individuals [7] , [8] , as well as sequencing read datasets [9]–[13] . Among these approaches , methods for sequence-based haplotype phasing are the only viable approach for haplotype phasing on a single individual member of a species ( assuming homologous chromosomes are sequenced together ) , as other approaches either require family members or a population . For an individual diploid genome , the problem of reconstructing the diplotype using sequence information , the diploid phasing problem , is equivalent to the identification of the sequence of alleles on either parental haplotype . If this sequence is correctly inferred , then the other haplotype will automatically carry the corresponding opposite alleles ( reference or alternative ) . Solving an error-free version of the diploid haplotype reconstruction problem is straightforward: the haplotype of each connected ( by reads ) component of heterozygous SNPs can be obtained by propagating allele information within reads . In reality , however , sequencing errors as well as false read mappings cause conflicts within sequence information , requiring a mathematical formulation of the haplotype reconstruction problem . Among various formulations suggested for this problem , the most commonly used is an NP-hard minimum error correction ( MEC ) definition [14] , [15] , which aims to identify the smallest set of nucleotide changes required within mapped fragments that would allow a conflict-free separation of reads into two separate homologous chromosomes ( or a bipartite separation of the fragment conflict graph ) . Some of the solutions proposed for this problem include: HapCUT[9] , an algorithm for optimizing MEC score based on computing max-cuts of the fragment graph; Fast Hare [16] , a heuristic that clusters reads into two sets in a greedy fashion , and HapCompass [10] , a spanning tree based approach for minimizing fragment conflicts . Unlike diploid genomes , computational identification of common chromosomal variants in polyploid genomes using sequencing data has received little attention , except in the pioneering work of Aguiar & Istrail [8] . Polyploidy studies are of importance as they allow a comprehensive investigation of variants within plant , fish , and yeast genomes and help understand mechanisms of eukaryotic evolution . However , haplotype reconstruction in polyploid genomes is fundamentally more complex , even in the error-free version of the problem ( without sequencing errors or false read mappings ) . Due to the newness of the NGS-based biological research in polyploid genomes , the mathematical foundations of the polyploid phasing problem have not yet been established . The solution proposed by Aguiar & Istrail for single individual polyplotyping problem is based on phasing all possible SNP loci pairs independently while further consolidating this information in a separate stage in order to infer a set of haplotypes . Diploid phasing methods focus on a given list of heterozygous variants that are guaranteed to contain a single reference allele , as well as an alternative allele ( assuming all heterozygous loci are bi-allelic ) . In contrast , in the polyploid phasing problem , there is no such guarantee of a single type of heterozygous SNP . Each heterozygous locus for a -ploid chromosome can potentially contain from up to alternative alleles within the heterozygous loci , significantly increasing the complexity of the phasing problem in comparison to the diploid case . Furthermore , in a diploid phasing setting , there are always two possible options for phasing a pair of SNP loci , regardless of what other SNPs they are phased with . These two options can be thought as parallel ( alternative allele pairs and reference allele pairs are matched within themselves ) or switched ( each alternative allele is matched with the other reference allele ) . These two options are no longer relevant when the genome contains more than two copies of each chromosome , due to the fact that there are up to options when merging a phased haplotype block with another . In this paper , we introduce a maximum-likelihood formulation of the polyploid full haplotype reconstruction problem and present a haplotype assembly algorithm , HapTree , which concurrently performs SNP-pair phasing and full haplotype assembly based on a probabilistic framework . We observe that , on simulated polyploid data , HapTree substantially improves the phasing capabilities and performance of any existing program . Because real polyploid data is hard to come by , we also evaluate HapTree on real human diploid data and find that , when compared to the more accurate trio-based data as the ground truth [17] , HapTree significantly reduces the number of switch errors , while remaining on par in terms of MEC score over existing single-individual haplotype assembly methods for diploid genomes . We also introduce a relative likelihood ( RL ) score definition for annotation-free evaluation of phasing quality for polyploid haplotype assembly as an alternative to MEC score . Using simulated polyploid sequencing datasets , we demonstrate that RL-score performs significantly better at capturing haplotype assembly quality than MEC-score as ploidy increases .
The HapTree pipeline is designed to perform phasing and full haplotype assembly of a single genome . The key component of HapTree is a relative likelihood function which measures the concordance between the aligned read data and a given haplotype phase under a probabilistic model that also accounts for possible sequencing errors . To identify a phasing solution of maximal likelihood , HapTree finds a collection of high-likelihood solutions for phases of the first SNP loci and extends those to high likelihood phases of the first SNP loci , for each incremental . In each step , HapTree maintains only the set of likely partial phases to be extended in next steps . Finally , a phase of maximal likelihood for all loci is obtained after the extension of the last SNP locus . Broadly speaking , HapTree aims to discover the best , or maximum likelihood , haplotype based on the read data available . Theoretically , one could enumerate all possible haplotypes , compute the likelihood of each being the true haplotype ( using formulas described below ) , and choose the most likely one; in most cases this approach is intractable as there are exponentially many possible haplotypes . HapTree therefore has a variety of ways of trimming down the solution set from all possible haplotypes to a much smaller set of more likely solutions , making the problem tractable . It does so by taking an inductive approach , generating a collection of likely phasing solutions for the first two SNPs in the genome , and then extending those to phasing solutions of the first three SNPs , and those to the first four SNPs , and so on . When extending any particular solution , HapTree chooses ( based on computing likelihoods ) how the alleles of the newly added SNP may be assigned to chromosomes; it includes only those assignments that are sufficiently likely . Additionally , if HapTree finds after extending all solutions to include the next SNP that there are too many likely solutions , it throws the worst ( least likely ) solutions away . Upon including all SNPs to be phased , HapTree randomly chooses a solution of maximum likelihood from amongst the solutions it has found . An implementation of our method , HapTree , is available for download at: http://groups . csail . mit . edu/cb/haptree/ We describe below the problem of sequence-based polyploid haplotype assembly and provide basic technical notation that will be useful for describing our method . We assume for now that each SNP locus to be phased is bi-allelic ( i . e . contains only two possible alleles , one being the reference allele ) . We further assume that for each SNP locus , the genotype of is known and is defined to be the number of chromosomes carrying the alternative allele ( denoted by ) . If denotes the ploidy , can range from to for heterozygous loci . At this point , we would like to note that these two assumptions are made for the sake of simplicity of method description and implementation , though the genotype information does tend to be available . After describing our method we also describe the changes needed to our original approach to accommodate multi-allelic and genotype-oblivious polyploid haplotype assembly . At this time our implementation accommodates the aforementioned simpler case of bi-allelic SNPs and known genotypes; it is simple to extend this implementation to the more general case , and we describe such an extension in Discussion . We denote the sequence of observed nucleotides of a fragment simply as a “read” ( independent from single/paired-end reads and sub-reads of a strobe read structure ) . The set of all reads is denoted as . We define a read as a vector with entries where a denotes the reference allele , a the alternative allele , and a indicates one of two possibilities: First , that the read does not overlap with the corresponding SNP locus , or second , that neither the reference nor alternative allele is present and hence there must be a read error . A read contains a SNP if . A read can also be represented as a dictionary or mapping with keys the positions ( from amongst the SNPs to be phased ) of SNP loci it contains and values of either reference allele or alternative allele , represented by 0 and 1 respectively ( e . g . ) . As current sequencing technologies generate read data with a certain rate of sequencing errors , some of the positions within a read likely contain false nucleotide information . Among these erroneous bases , unless they are located at SNP loci and contain opposite allele information , we ignore them by representing them with , and thus keep only confounding sequencing errors that can affect phased haplotype results . For each read and for each SNP locus , we assume an error rate of and a probability of opposite false allele information is equal to . We modify this error rate by a factor of two-thirds because conditional on there being an error , we model the error as equally likely to be any of the three other alleles . Two of the three of these alleles are neither the reference nor the alternative allele and thus we know that an error has been made in this case . Therefore , two-thirds of the time the erroneous alleles produced are known as such and may be thrown out , leaving a true error only one-third of the time . We represent these error rates as matrices . At this time our method assumes uniform error rates with respect to the SNP position; the error rate is supplied by the user and ought to depend on the read sequencing technologies used . Upon the set of SNP loci and read set ; we define a Read Graph , , such that there is a vertex for each SNP locus and an edge between any two vertices if there is some read containing both and ; equivalently if . Without loss of generality , we assume that is connected; otherwise each connected component can be processed independently . Our main approach to solving the single individual polyploid haplotype assembly problem is by finding highly probable solutions on SNPs and extending those to highly probable solutions on SNPs . Our algorithm has two fundamental parts: branching and pruning . For each connected component of the , , we inductively generate a collection of high likelihood phases on the first SNPs . For each of these phases , we branch them to phases on SNPs by considering all possible orderings of alleles for position and including branches for those which occur with probability above a certain threshold . After doing so , we prune the tree of phases by removing all leaves that occur with probability sufficiently less than the most probable leaf . We discuss both parts in more detail below . We note that although a dynamic programming algorithm can be directly applied to infer the best solutions under HapTree's likelihood model , we instead developed HapTree , which is substantially faster than exact dynamic programming but with nearly identical empirical performance .
Determining the quality of a phasing solution depends on whether the true phase is known . When no such information is avaliable , the Minimum Error Correction ( MEC ) score [15] is a widely used scoring function to measure the quality of phasing solutions . The MEC score is defined as the minimum ( amongst chromosomes ) number of mismatches between a phase and the read set . A number of existing programs , including HapCut [9] , find phasing solutions by optimizing the MEC score in diploid cases . For higher ploidy the MEC score can no longer be reliably used because unlike in the diploid case , the phase of any one chromosome does not determine the phases of the others . Moreover , the MEC score does not distinguish between two separate phases of a pair of SNP loci with different non-zero counts of in their vector sets . Finally , unlike in the diploid case , a phase of a pair of SNP loci containing a set of parallel alleles does not prevent it from containing a set of switched alleles as well . To demonstrate these issues , consider two possible vector sets corresponding to phases of a pair of triploid SNPs both with genotype 2: and . If the read data is , it is clear from a probabilistic standpoint that phase is a better fit , but both and have equal MEC scores . This effect is exaggerated as increases . When a true phase is available , there are a variety ways to evaluate how accurate any predicted phase is . A widely used measure in diploid phasing is switch error , which is calculated as the number of positions where the two chromosomes of a proposed phase must be switched in order to agree with the true phase . For polyploid phasing , we generalize switch error to vector error . In higher ploidy cases , at any SNP locus , it is possible for no chromosomes in a proposed phase to require a switch or anywhere from to chromosomes to require switches , in order for a proposed phase to agree with the true phase . We do not wish to penalize a solution where only two vectors must be switched at a given position with the same penalty to be used for a solution in which all vectors must be switched . The vector error of a proposed phase ( with respect to the true phase ) is defined by the minimum number of segments on all chromosomes for which a switch must occur; for the diploid case this score is exactly twice the switch error . One may also think of the vector error as the minimum number of segments a proposed phase and the true phase have in common , less the ploidy . Even for triploid genomes , the vector error is more discriminative than switch error . Consider the following example in Figure 2: In Figure 2 phase ( i ) is a more accurate phase than ( ii ) , and phase ( ii ) more accurate than phase ( iii ) . The segments are broken up by row and color: phase ( i ) having five segments , phase ( ii ) having six , and phase ( iii ) having seven . Note that there may be several ways to break a vector set into a minimal number of segments; phase ( ii ) is such an example . Finally , we remark that vector error can be computed in time , where is the ploidy and the block size .
Not only does HapTree outperform HapCompass on phasing quality , it is also significantly faster , especially for longer block length . The median runtimes for block length 10 and 10× coverage were seconds for HapTree and HapCompass , respectively; for block length of 40 and 40× coverage , they were seconds , respectively . As seen in the results of Geraci et al . [18] , there is no perfect solution for diploid phasing . HapCUT is one of the methods reported that consistently performs best or close-to-best for a variety of experiments . For a proof of concept of how HapTree would perform on real data , we ran HapTree and HapCUT using 454 and Illumina sequencing data of the well-studied NA12878 genome ( 1000 Genomes Project Phase 1 ) [17] , and compared MEC scores as well as switch errors to a trio phasing annotation accepted as ground truth; we present these results in Table 1 . The trio phasing annotation represents a high quality diplotype of NA12878 for all SNP sites where either parent ( NA12891 or NA12892 ) is homozygous [17] . Note that we computed the number of switch errors within connected SNP components only , against SNPs whose phase has been determined by the trio-based phasing; we then sum over components . In this case , HapTree was run with a uniform error rate of , an EXTEND threshold , and primarily with a PRUNE threshold of . We begin to prune more aggressively when we have at least or possible haplotypes with thresholds of respectively . For the vector set prior , from examining the read data , we ran HapTree with parallel bias . We found that HapTree and HapCUT perform almost identically in MEC scores , with HapCUT having marginally smaller scores for both 454 and Illumina data sets . It is worth noting that HapCUT optimizes MEC score , and MEC score measures only the consistency between a phasing solution and read data , not with the true phase . Notably , when comparing to the ground-truth phase as determined by trio-based phasing , we found HapTree significantly outperforms HapCut in terms of switch error rate for the phasing experiments on the NA12878 genome for 454 and Illumina datasets . Although our method is not primarily designed for phasing diploid genomes , it is still able to achieve better phasing results , when compared to the state-of-the-art diploid method . Again , the results on real-world read datasets showed the superiority of our likelihood function over MEC score for NGS-based phasing .
We have presented a scalable algorithm , HapTree , for polyplotyping using NGS sequencing data and a new metric for measuring accuracy in this context . We have described an efficient algorithm to identify phases that maximize our RL metric , a relative likelihood function which measures the quality of a given phase according to the read dataset . We have demonstrated the advantages of such a likelihood formulation over the existing MEC score in phasing both polyploid and diploid genomes . HapTree not only substantially improves the efficiency and phasing accuracy of the state-of-the-art in polyploid phasing , but also produces more accurate phased haplotype blocks for diploid genomes , as compared to HapCUT , which is designed for diploid phasing by MEC score optimization . Our results indicate that HapTree can be used in phasing individual triploid and tetraploid genomes , as well as improving phasing of real diploid genomes . HapTree also easily scales to genomes of higher ploidy . Our algorithm can be easily extended to phase data with multi-allelic SNPs and with unknown genotype information as well . With unknown genotype information and multi-allelic SNPs , instead of allele permutations , there are possibilities , since all 4 alleles ( A , C , G , T ) are possible for all chromosomes . For bi-allelic SNPs with unknown genotypes , there are , as all possible reference and alliterative allele permutations are allowable . Finally , when the genotype is known but a SNP is multi-allelic , we may use multinomial coefficients to compute the number of allele permutations allowable: , where denotes the number of alleles according to the genotype , where . The only change to HapTree in these cases is that at each EXTEND step , we allow all allele permutation possibilities as dictated by whatever genotypic is available: we compute the probabilities for all , , or possibilities ( depending on the situation ) as opposed to and EXTEND accordingly . Moreover , the type of information available does not need to be the same for all SNPs , since it only determines which allele permutations we introduce at the EXTEND step . A future application of HapTree is genotype imputation , which can predict missing genotype from phasing results . As polyploid sequencing data becomes available , HapTree will be useful for the investigation of the role of heterozygosity in plant , fish , and other species . Moreover , accurate individual phases of diploid haplotypes can be assembled without the use of pedigree or population information . A summary of this paper appears in the proceedings of the RECOMB 2014 conference , April 2–5 [19] . | While human and other eukaryotic genomes typically contain two copies of every chromosome , plants , yeast and fish such as salmon can have strictly more than two copies of each chromosome . By running standard genotype calling tools , it is possible to accurately identify the number of “wild type” and “mutant” alleles ( A , C , G , or T ) for each single-nucleotide polymorphism ( SNP ) site . However , in the case of two heterozygous SNP sites , genotype calling tools cannot determine whether “mutant” alleles from different SNP loci are on the same or different chromosomes . While the former would be healthy , in many cases the latter can cause loss of function; it is therefore necessary to identify the phase—the copies of a chromosome on which the mutant alleles occur—in addition to the genotype . This necessitates efficient algorithms to obtain accurate and comprehensive phase information directly from the next-generation-sequencing read data in higher ploidy species . We introduce an efficient statistical method for this task and show that our method significantly outperforms previous ones , in both accuracy and speed , for phasing triploid and higher ploidy genomes . Our method performs well on human diploid genomes as well , as demonstrated by our improved phasing of the well known NA12878 ( 1000 Genomes Project ) . | [
"Abstract",
"Introduction",
"Method",
"Results",
"Discussion",
"Discussion"
] | [
"haplotypes",
"biology",
"and",
"life",
"sciences",
"population",
"genetics",
"computational",
"biology",
"evolutionary",
"biology"
] | 2014 | HapTree: A Novel Bayesian Framework for Single Individual Polyplotyping Using NGS Data |
The Drosophila Alp/Enigma family protein Zasp52 localizes to myotendinous junctions and Z-discs . It is required for terminal muscle differentiation and muscle attachment . Its vertebrate ortholog ZASP/Cypher also localizes to Z-discs , interacts with α-actinin through its PDZ domain , and is involved in Z-disc maintenance . Human mutations in ZASP cause myopathies and cardiomyopathies . Here we show that Drosophila Zasp52 is one of the earliest markers of Z-disc assembly , and we use a Zasp52-GFP fusion to document myofibril assembly by live imaging . We demonstrate that Zasp52 is required for adult Z-disc stability and pupal myofibril assembly . In addition , we show that two closely related proteins , Zasp66 and the newly identified Zasp67 , are also required for adult Z-disc stability and are participating with Zasp52 in Z-disc assembly resulting in more severe , synergistic myofibril defects in double mutants . Zasp52 and Zasp66 directly bind to α-actinin , and they can also form a ternary complex . Our results indicate that Alp/Enigma family members cooperate in Z-disc assembly and myofibril formation; and we propose , based on sequence analysis , a novel class of PDZ domain likely involved in α-actinin binding .
Vertebrate muscles consist of three major types , skeletal , cardiac , and smooth muscles , which correspond in Drosophila to body wall , heart , and visceral muscle . Common to all is an actomyosin contractile system with thin filaments anchored at Z-discs . A crucial component of Z-discs is α-actinin , which anchors actin filaments at the Z-disc . In addition , proteins of the Alp/Enigma family function in maintenance of Z-discs [1] and have also been proposed to play an important role in myofibril assembly [2] , [3] . In vertebrates , the Alp/Enigma family comprises ZASP/Cypher/LDB3/PDLIM6 , ENH/PDLIM5 , ENIGMA/PDLIM7 , PDLIM1/CLP36 , PDLIM2/Mystique , ALP/PDLIM3 , and PDLIM4/RIL [4] . All vertebrate family members have one N-terminal PDZ domain , and one or three C-terminal LIM domains . In Drosophila , Zasp52 is a member of the Alp/Enigma family with a PDZ domain , Zasp-like motif and four LIM domains; another potential member , Zasp66 , lacks the LIM domains , but features a similar PDZ domain and Zasp-like motif , and also localizes to Z-discs [2] , [5] . Zasp52 was identified in an RNAi screen for spreading defects of S2R+ tissue culture cells [2] , [6] . We could show that Zasp52 is a focal adhesion component and is required for cell spreading downstream of integrins . It also co-localizes with integrins at myotendinous junctions and is required for muscle attachment . Finally , it co-localizes with α-actinin to Z-discs and plays a role in embryonic Z-disc assembly [2] . Other groups proposed a role mainly in Z-disc maintenance [7] , [8] . More recently we documented that Zasp52 occurs as at least 13 different splice isoforms and localizes to Z-discs in all muscle types in Drosophila [9] . Mutations of Zasp52 orthologs in vertebrates cause similar defects , ranging from improper formation of somites and heart in zebrafish to fragmented Z-discs in skeletal and cardiac muscles in mice [10] , [11] . The single C . elegans ortholog ALP-1 displays defects in actin filament organization , but motility defects are much milder than in vertebrates or Drosophila [12] , [13] , [14] . Mutations in the human ortholog ZASP result in phenotypes of variable severity from congenital myopathy with fetal lethality to late-onset cardiomyopathy [1] . In this study , we show by live imaging of embryos that GFP-Zasp52 first assembles into repetitively spaced clusters , putative Z bodies , at the cortex of myotubes , which then coalesce to form Z-discs . We also show by antibody stainings that Zasp52 is among the first repetitively spaced Z-disc markers in indirect flight muscle ( IFM ) development , indicating that Zasp52 plays a general role in Z-disc assembly . We demonstrate a role for Zasp52 , Zasp66 , and the newly identified Zasp67 in IFM assembly , and show that Zasp52 acts together with Zasp66 , both of which bind directly to α-actinin . Finally from sequence analyses we propose the name Zasp PDZ domain for PDZ domains with a putative α-actinin binding motif , which can be found in all vertebrate Alp/Enigma family members , as well as in Zasp66 and Zasp67 in Drosophila , and myopodin and CHAP in vertebrates .
Zasp52 depletion causes partial embryonic lethality and defects in embryonic myofibril assembly , in particular Z-discs are not properly aligned and do not properly recruit α-actinin [2] . This suggests that Zasp52 could be an early marker suitable to follow myofibril assembly in real time . We also previously confirmed that line G00189 is a GFP-Zasp52 fusion that faithfully represents endogenous protein localization and is fully viable and functional [2] . We therefore used GFP-Zasp52 for live imaging to document Z-disc , and by extension , myofibril assembly ( Figure 1A and Video S1 ) . We focused on the period between late stage 16 embryos , when Zasp52 localizes only to myotendinous junctions and is evenly distributed within myotubes , and late stage 17 embryos , when Zasp52 localizes distinctly to both myotendinous junctions and Z-discs [2] . Our observations show: 1 ) GFP-Zasp52 gradually accumulates in clusters that steadily increase in size . 2 ) GFP-Zasp52 is first cleared from the area next to the myotendinous junctions . 3 ) GFP-Zasp52 clusters sort out gradually into future Z-discs . 4 ) GFP-Zasp52 clusters can first be observed close to the sarcolemma ( Figure 1B ) . 5 ) Eventually , GFP-Zasp52 clusters coalesce to form the final Z-disc . 6 ) Sorting of GFP-Zasp52 clusters into future Z-discs correlates with a gradual increase in contractility , with initial contractility observed concomitant with GFP-Zasp52 clusters , after about 55 min in Video S1 . These observations are consistent with our proposed model of myofibril assembly [3] . They demonstrate that GFP-Zasp52 is an early marker for myofibril assembly and a suitable tool for live imaging studies of myofibril assembly . As Zasp52 localizes to Z-discs in all muscle types [9] , we asked whether Zasp52 has a general role in Z-disc assembly . We therefore determined Zasp52 localization during IFM development . IFM development is distinct from that of embryonic body wall muscle in several respects . First , development takes much longer , approximately 96 h at 25°C from puparium formation to the emerging fly , and second , sarcomeres grow over time , from about 1 . 7 µm to a final length of 3 . 3 µm [15] . Due to the extended period of IFM development , we were able to perform antibody stainings at different stages of development . As during embryonic myogenesis , Zasp52 can be detected in a repetitive pattern at very early stages of pupal muscle development ( Figure 2 ) . At 30 h after puparium formation ( APF ) , Zasp52 has a punctate distribution along the forming myofibrils and co-localizes with α-actinin . At this stage actin staining shows undifferentiated strands with no H-zones visible . At 48 h APF , Zasp52 and α-actinin appear as broad dots at the Z-discs . Myofibrils are narrow and actin labelling now shows a regular periodicity with evenly spaced H-zones . At 72 h APF , sarcomeres have grown in length , with Zasp52 and α-actinin appearing as elongated bands in the growing Z-disc . In the adult , Zasp52 and α-actinin are labeled in clear striations . A strong hypomorphic mutation in Zasp52 deleting most splice variants causes late embryonic to larval lethality [2] . We therefore decided to only deplete Zasp52 long isoforms with an RNAi transgene targeting the last exon of Zasp52 ( KK101276 , called UAS-iZasp52ex20 in this study ) , which allows us to study the function of Zasp52 in IFM . We used the pan-muscle driver Dmef2-Gal4 to knock down Zasp52 in all muscles . We verified the knockdown efficiency by immunoblotting ( Figure 3A ) . Long isoforms are almost absent in immunoblots from isolated IFM , and are completely gone with the addition of UAS-Dicer ( Dcr ) , which enhances the generation of siRNAs . iZasp52ex20 mutant flies lacking long isoforms encoding LIM domains 2–4 are completely flightless ( Figure 3B ) . To rule out off-target effects , we generated a second RNAi transgene against exon 16 with an shRNA transgene we call UAS-iZasp52ex16 ( exon numbering according to [9] ) . Only by using Dcr , we were able to obtain a phenotype with this transgene . Dcr iZasp52ex16 mutants knock down long isoforms to a similar , but smaller degree as judged by immunoblotting ( Figure 3A ) , and consistent with this , their flight ability is less impaired ( Figure 3B ) . We next analyzed adult IFM of iZasp52ex20 and Dcr iZasp52ex16 mutants by antibody staining and confocal microscopy ( Figure 3C ) . In wild type myofibrils , obscurin and kettin , a titin isoform , label in straight bands at M-lines and Z-discs , respectively . H-zones are always straight and evenly spaced . In iZasp52ex20 and Dcr iZasp52ex16 knockdown flies , kettin labeling at the Z-disc appears normal , whereas obscurin in the M-line is occasionally wavy . Associated bent H-zones are frequently observed , indicating irregular thin filament lengths . In some areas of the IFM in iZasp52ex20 knockdown flies , unstable and frayed myofibrils are seen . This becomes more apparent when the phenotype is enhanced with Dcr showing myofibrils with distorted Z-discs and M-lines throughout the sample ( Figure 3C ) . Overall , the phenotypes of both Zasp52 RNAi transgenes are similar , consistent with them having no off-target effects . More importantly , the phenotypes indicate that Zasp52 is required for proper adult myofibril IFM structure , a muscle very different in structure and function from embryonic body wall muscles . We wondered if Zasp52 defects arise already during development , or are maintenance defects due to muscle contractility of adult IFM . To address this question , we stained IFM muscles of wild type and Dcr iZasp52ex20 knockdown flies at different stages of pupal development ( Figure 4 ) . Disruptions of myofibrils of Dcr iZasp52ex20 knockdown flies become apparent at 48 h APF . At this stage myofibrils are thinner than in the wild type , without clearly defined H-zones . Kettin and obscurin label in fuzzy dots at Z-discs and M-lines , respectively . Kettin and obscurin appear less ordered than in the wild type , indicating that sarcomeres are not properly assembled . At 72 h APF , some of the myofibrils are frayed and kettin and obscurin labeling is in wavy stripes as seen in the adult fly . Overall these observations indicate that a lack of Zasp52 affects myofibril assembly during pupal development . We wondered if some functions of Zasp52 are masked by redundancy , and therefore performed a detailed database search with the Zasp PDZ domain among Drosophila and human proteins . We uncovered two Drosophila proteins , Zasp66 and CG14168 , which we name Zasp67 owing to its cytogenetic location and similarity to Zasp66 . Zasp66 and Zasp67 have a similar PDZ domain followed by the Zasp-like motif , and can therefore be classified as novel Alp/Enigma family members ( Figure S4B ) . We also found two human proteins , CHAP and myopodin , which have a PDZ domain highly similar to Alp/Enigma family proteins with the putative amino acids required for α-actinin binding that are absent in the next-closest PDZ domain protein LMO7 ( Figure 5 and Figure S1 ) [11] , [16] . As CHAP and myopodin lack both LIM domains and the Zasp-like motif , we do not classify them as new members of the Alp/Enigma family . The PDZ domain-ligand interaction network was recently determined in humans [17] . Their algorithm predicts α-actinin as a likely ligand for all PDZ domains of this group , but not for LMO7 ( Figure S1 ) . Multiple sequence alignment and phylogenetic tree analysis shows that Zasp66 is the most distantly related member of this group of PDZ domains ( Figure S2 and S3 ) . We therefore decided to also investigate Zasp66 , to obtain a better idea of functions potentially applying to all PDZ domains in this group . We also initiated characterization of Zasp67 to gather additional evidence on conserved functions of this protein family . Zasp66 was first identified with a GFP trap generating an endogenous GFP fusion protein demonstrating that Zasp66 localizes to Z-disks [5] . Zasp66 is an alternatively spliced gene on chromosome arm 3L with at least 13 annotated transcripts ( FlyBase; see Figure S4A for three representative transcripts ) . Zasp66 and Zasp52 both have very similar expression profiles with peak expression during embryonic and pupal myofibril assembly ( FlyBase ) . They share an N-terminal PDZ domain and a Zasp-like motif that are highly similar to each other ( Figure S4B ) . Four Zasp66 transcripts encode both the PDZ domain and the Zasp-like motif , three transcripts encode a truncated PDZ domain and the Zasp-like motif , and the remaining six transcripts encode only the Zasp-like motif ( Figure S4A ) . It is known that Z-disc or M-line proteins can distribute differentially throughout the diameter of their respective bands . For example , the vertebrate M-line proteins obscurin and Obsl1 localize in a non-overlapping pattern at the M-line , where obscurin is at the periphery of the myofibril , and Obsl1 in the core [18] , [19] . We therefore co-immunostained GFP-Zasp66 with anti-Zasp52 antibody to determine if there are subtle differences in localization . Both co-localize indistinguishably throughout the entire diameter of the Z-disc ( Figure 6A ) . The genetic tools to analyze Zasp66 are limited , because it is localized in a haploinsufficient region , and therefore no deficiencies are available . However , there is one hairpin RNA transgene ( KK112973 , called UAS-iZasp66 in this study ) that targets a 200 bp Zasp66 exon common to all 13 transcripts annotated by FlyBase ( see Figure S4A ) . This construct has zero predicted off targets and should therefore knock down all known Zasp66 transcripts with high specificity . We verified the knockdown of Zasp66 with RT-PCR and qPCR demonstrating efficient knockdown ( Figure 6B , 6C ) . Depletion of Zasp66 with Dmef2-Gal4 at 29°C results in high pupal lethality . With the inclusion of Dcr to enhance the phenotype , we observed almost complete pupal lethality and flightless adults ( Figure 6D ) . We first investigated the IFM of rare adult escaper flies by confocal microscopy . They look largely normal with an occasional thickening and bending of Z-discs ( Figure 7A , arrows ) . Rarely , we observed more severe disruptions in Z-disc structure ( left panels in Figure 7A ) . Because 98% of Dcr iZasp66 knockdown flies die as pupae , we also investigated if developing IFM exhibit defects . We only analyzed the 48 h APF time point , because Dcr iZasp66 pupae at 72 h are largely dead or dying . Pupal IFM stained with phalloidin , anti-obscurin , and anti-kettin antibody reveal thinner and frayed myofibrils with irregular Z-discs ( Figure 7B ) . Altogether , we examined 16 IFMs from individual 48 h knockdown pupae , all of which had a similar phenotype . These data indicate that Zasp66 , like Zasp52 , contributes to Z-disc assembly during development . We next asked if Zasp66 also contributes to Z-disc stability . Given that hatched flies without Dcr enhancement could fly , we did not expect strong defects , and therefore we compared sarcomeric organization of IFM by transmission electron microscopy ( TEM ) . Zasp66 RNAi mutants exhibit mild defects in Z-disc structure ( Figure 8A ) . Wild type Z-discs are always completely regular , whereas Zasp66 mutant Z-discs often show small pockets , where Z-disc material is missing . These defects are weaker than defects caused by depletion of iZasp52ex20 . When long isoforms of Zasp52 are depleted , there is very little Z-disc material left ( Figure 8A ) . Apart from occasionally shifted H-zones , the rest of the sarcomere is unaffected , with correctly arranged thick and thin filaments . As seen by confocal microscopy , this phenotype is enhanced by the use of Dcr , with frequently distorted Z-discs and H-zones . In the case of Zasp66 , the use of Dcr results in a phenotype similar in strength to the phenotype observed when iZasp52ex20 is depleted without Dcr . As in the majority of cases the overall integrity of sarcomeres was not lost , we asked if Z-discs are unstable and fall apart due to the conditions used for sample preparation for TEM . During sample preparation , IFMs were treated with glycerol and Triton X-100 in order to be able to place sarcomeres into rigor state . This allows detailed structural analysis of sarcomere components and normally does not interfere with sarcomere architecture [20] . We analyzed iZasp66 and iZasp52ex20 mutant IFM without the use of glycerol and Triton X-100 . In this case the Z-discs of iZasp52ex20 knockdown flies appear more complete , but are much thinner with many irregularities . iZasp66 Z-discs look like wild type ( Figure 8B ) . This demonstrates that both Zasp52 and Zasp66 are crucial for Z-disc stability . We finally wanted to know if and how Zasp52 and Zasp66 act together in Z-disc assembly . We showed previously that Zasp52 can co-immunoprecipitate α-actinin [2] , and in vertebrate ZASP/Cypher , the PDZ domain is crucial for interaction with α-actinin [11] , [21] , [22] . Zasp52 and Zasp66 both carry an N-terminal PDZ domain . We therefore tested whether both Zasp52 and Zasp66 can bind α-actinin by pulling down endogenous GFP fusions to Zasp52 and Zasp66 with anti-GFP antibody-coupled beads . Both GFP-Zasp52 and GFP-Zasp66 robustly co-immunoprecipitate α-actinin , whereas extracts from y w control flies do not ( Figure 9A ) . We next asked if this interaction is direct . We overexpressed and purified His-tagged Zasp66 and GST-tagged Zasp52 from bacteria , and tested direct interaction with rabbit skeletal muscle α-actinin . Both Zasp52 and Zasp66 directly interact with α-actinin ( Figure 9B ) . We can also show that GFP-Zasp66 co-immunoprecipitates Zasp52 , whereas control flies expressing GFP alone do not ( Figure 9C ) . These data indicate that Zasp52 , Zasp66 and α-actinin form a ternary complex . Finally , we determined if Zasp66 genetically interacts with Zasp52 . To this end , we removed one copy of Zasp52 in the background of Zasp66 depletion by RNAi and measured pupal lethality . While knocking down Zasp66 on its own or only removing Zasp52 heterozygously shows mild pupal lethality at 25°C , removing Zasp66 together with one copy of Zasp52 substantially increases pupal lethality , demonstrating a genetic interaction between Zasp52 and Zasp66 ( Figure 9D ) . These data indicate that Zasp52 and Zasp66 cooperate in Z-disc assembly and that both are direct binding partners for α-actinin . To address the issue of potential redundancy between Zasp52 and Zasp66 in Z-disc assembly , we investigated the double mutant phenotype . Using the pan-muscle driver Dmef2-Gal4 , iZasp52ex20 and iZasp66 double mutants die at the earliest stage of pupal development precluding analysis of developing IFM . We therefore used the IFM-specific driver Act88F-Gal4 [23] , in order to obtain adult double knockdown flies . As with Dmef2-Gal4 , iZasp66 knockdown flies are able to fly and show no severe phenotype when analyzed by electron microscopy using glycerol and Triton X-100 extraction ( Figure 10 ) . Sarcomeres are properly arranged and Z-discs have small pockets with missing Z-disc material . In iZasp52ex20 single knockdown flies the phenotype was as observed when driven with Dmef2-Gal4 . There is almost no Z-disc material left and H-zones and Z-discs are distorted occasionally . No frayed myofibrils can be seen ( Figure 10 ) . In iZasp52ex20 iZasp66 double knockdown flies , we observe a more severe phenotype than would be expected by additive effects of single knockdowns . Myofibrils are frequently frayed and unstable , with severely distorted Z-discs and H-zones ( Figure 10 ) . This synergistic defect indicates that Zasp52 and Zasp66 function partially redundantly during myofibril assembly in the IFM and cooperate in stabilizing Z-discs . Zasp67 is exclusively expressed during pupal stages at a time when pupal myofibrils assemble ( FlyBase ) , but we do not know if Zasp67 protein localizes to Z-discs similar to Zasp52 and Zasp66 . We tested two available Zasp67 RNAi transgenes , GD8245 and KK111478 , which both result in flight-impaired flies , when expressed in muscles with Dmef2-Gal4 . We continued to work with KK111478 , which we call iZasp67 . We can show by RT-PCR and qPCR that Zasp67 is efficiently knocked down ( Figure S5B , S5C ) . We then characterized IFMs of iZasp67 mutants by electron microscopy . They exhibit a phenotype very similar to knocking down the long isoforms of Zasp52 ( Figure S5A ) . We also analyzed Zasp67 Zasp52 double mutants , which look similar to Zasp66 Zasp52 double mutants , but even more severe ( Figure S5A ) . We lastly checked α-actinin localization in various mutant combinations , which all still express one or several Zasp isoforms . We observe normal α-actinin localization in all mutant combinations ( Figure S6 , see Discussion ) . These results indicate that Alp/Enigma family members in Drosophila act partially redundantly in the same pathway , the assembly of Z-discs .
Our live imaging with GFP-Zasp52 is the first time-lapse recording of myofibril assembly in a whole animal . The only other live imaging in whole animals was done in zebrafish skeletal muscle for fluorescence recovery after photobleaching [26] . Other time-lapse recording studies used reassembly of myofibrils in tissue culture cells [27] . A further difference to previous studies is that GFP-Zasp52 is endogenous , fully functional and viable . Still , both our and previous studies agree on two points . First , Z-disc proteins initially form separate clusters close to the sarcolemma , presumably corresponding to the Z bodies described in electron microscopy studies . Second , these clusters coalesce and align into Z-discs . Our study additionally documents that GFP-Zasp52 clusters are initially evenly distributed and gradually sort out to the future Z-disc , while at the same time growing in size . We also show a clear correlation between Z-disc assembly and an increase in contractility . We notice one important difference: in avian heart and in zebrafish skeletal muscle as well as in IFM , myofibril assembly is approximately an order of magnitude slower than in the Drosophila embryo . Also , non-embryonic myofibril assembly involves an increase in sarcomere length , or a premyofibril stage , which employs non-muscle myosin [15] , [28] . In embryonic myofibril assembly , there is no increase in sarcomere length , and initial spacing corresponds to the final sarcomere length . This is likely due to time constraints of the very fast development of Drosophila embryos . This timelapse study fits very well with a model we have proposed recently [3] , and also with a computational modeling study indicating that actin clusters cross-linked at the barbed end ( Z bodies ) together with actin filament treadmilling is sufficient for establishment of sarcomere arrays [29] . In the IFM , a very different muscle type , Zasp52 also localizes to Z-disc precursors at the earliest stages of pupal IFM development , further strengthening the notion that myofibril assembly is highly conserved across muscle types . Live imaging of myofibril assembly in Drosophila embryos provides a suitable model system , because it occurs very quickly , and without the complications of sarcomere growth . The adult Zasp52 IFM phenotypes confirm and extend our previous observations on body wall muscles of embryos [2] . We used UAS-iZasp52ex20 to study Zasp52 phenotypes in IFM . It will form a 573 nt hairpin targeting the last exon of Zasp52 [30] . Even though no off-targets are predicted for this construct , we wanted to independently verify our phenotype with a different construct , and therefore generated an shRNA construct targeting only 19 nt within exon 16 ( UAS-iZasp52ex16 ) , also without predicted off-targets . As judged by immunoblotting ( Figure 3A ) , both constructs target only the long isoforms of Zasp52 , though Dcr iZasp52ex16 is slightly less efficient . They produce similar IFM defects , with the slightly weaker phenotype of Dcr iZasp52ex16 being consistent with its apparent reduced knockdown efficiency ( Figure 3B , 3C ) . Surprisingly , Dcr iZasp52ex16 causes stronger pupal lethality ( Figure 3B ) . This could be due to an off-target effect , a stronger reduction of a critical embryonic or larval isoform with iZasp52ex16 , or the difference between knocking down all long isoforms versus exon 16-containing long isoforms . Muscle defects are very similar with both the pan-muscle driver Dmef2-Gal4 and the IFM-specific Act88F-Gal4 driver ( Figure 8 , Figure 10 ) . This is consistent with inefficient knockdown of Zasp52 long isoforms during larval stages using Dmef2-Gal4 ( A . K . , unpublished observations ) . We demonstrate that the stability of Z-discs is severely compromised upon depletion of Zasp52 or Zasp66 , because significant amounts of Z-disc proteins can be lost simply by detergent extraction ( Figure 8 ) . The impaired stability likely gives rise to the misalignment of Z-discs and H-zones that we observe during IFM myofibrillogenesis ( Figure 4 ) . If IFM muscles contract during assembly or are under tension as we show for embryonic body wall muscles , then unstable Z-discs should lead to the misalignment of thin and thick filaments , resulting in wavy H-zones and M-lines . These developmental defects appear very early , consistent with our imaging data on embryonic myofibril assembly and our proposed role for Zasp52 as an organizer for Z body assembly [3] . We observe a similar developmental defect in Zasp66-depleted pupal IFM myofibrils ( Figure 7B ) . The Z-disc defects in single knockdowns of Zasp52 long isoforms are also similar to α-actinin mutants [31] , supporting the interdependence of Zasp PDZ domain proteins and α-actinin at the Z-disc . We still observe α-actinin at the Z-disc in various mutant combinations ( Figure S6 ) . We do not believe that this result is contradictory to our previous observation of reduced α-actinin recruitment to Z-discs in embryonic body wall muscles [2] , because even in double mutants , there are still several Zasp proteins expressed . For example , the most severe double mutant ( iZasp52ex20 iZasp67 ) still expresses the short Zasp52 isoforms and all Zasp66 isoforms . Moreover , the double mutant phenotypes are much more severe than the α-actinin null mutant phenotype [31] , consistent with Zasp proteins being upstream of α-actinin in Z-disc assembly . Finally , the phenotypic features we see in flies are similar to human myopathies [32] , supporting the use of fly muscles as a model system . Importantly , the Zasp52 Zasp66 double mutant phenotype is considerably more severe than would be expected from additive defects of single knockdowns ( Figure 10 ) , and the same is true for the Zasp52 Zasp67 double mutant ( Figure S5 ) , indicating a synergistic mechanism , where Zasp52 , Zasp66 , and Zasp67 cooperate in Z-disc assembly . A possible mechanism is the formation of a multiprotein complex consisting of α-actinin , Zasp52 , Zasp66 , and Zasp67 at the forming Z-disc , which helps in assembly and stabilization of the Z-disc . The genetic interaction of Zasp52 and Zasp66 and the direct binding of Zasp52 and Zasp66 to α-actinin support this model ( Figure 9 ) . Our results suggest that several Zasp-like proteins are required together with α-actinin to form a critical complex for Z body and Z-disc assembly . Such multiprotein complexes have already been reported for ENH , Cypher , calsarcin and myotilin , and have been inferred from RNAi studies for Zasp52 , non-muscle myosin and α-actinin [8] , [33] . The Alp/Enigma family comprises ZASP , ENH , ENIGMA , PDLIM1 , PDLIM2 , ALP , and PDLIM4 , and characteristically contains an amino-terminal PDZ domain , a Zasp-like motif , and carboxy-terminal LIM domains [4] . The Drosophila ortholog with the same domain organization is Zasp52 , also called Zasp [2] , [5] . We propose to include Drosophila Zasp66 and Zasp67 as novel family members , because they share a similar amino-terminal PDZ domain followed by the characteristic Zasp-like motif ( Figure S4B ) . Zasp66 and Zasp67 do not encode LIM domains , however , both Zasp52 and other Alp/Enigma family members encode protein isoforms without LIM domains [4] , [9] , [34] , indicating the existence of functional Alp/Enigma proteins without LIM domains . We uncovered two additional proteins , myopodin and CHAP , with a highly related PDZ domain ( Figure 5 ) . These proteins lack both LIM domains and the Zasp-like motif , therefore they are likely not Alp/Enigma family members . We propose the name Zasp PDZ domain for PDZ domains with an amino-terminal PWGFRLxGG motif , which is likely required for α-actinin binding [11] , [16] . We chose the name Zasp PDZ domain for two reasons: 1 ) the first PDZ domain that was crystallized and functionally analyzed , is from ZASP/Cypher [16] , [21] , [22] . 2 ) ZASP was also the first gene for which mutations in humans causing myopathies were identified [1] , [32] , [35] , [36] . Six of the proteins with a Zasp PDZ domain , ZASP , ENH , PDLIM1 , PDLIM2 , Alp , and PDLIM4 bind α-actinin via their PDZ domain [21] , [22] , [37]–[41] . In addition , ZASP , Enigma , ENH , PDLIM1 , and Alp localize to Z-discs [21] , [22] , [37] , [39] , [42]–[44] , while PDLIM2 and PDLIM4 localize to actin stress fibers in non-muscle cells [38] , [40] . CHAP plays an important role in myofibril assembly and co-localizes with α-actinin , but whether the PDZ domain is involved in α-actinin binding has not been clarified [45] . For myopodin , only an isoform that lacks the PDZ domain and functions in skeletal muscles has been analyzed [46] , but heart muscle expresses a 95 kD isoform localizing to Z-discs that could correspond to a PDZ-containing isoform [47] . Given that the most diverging Alp/Enigma protein in this group , Zasp66 ( see Figure 5 and Figure S3 ) , as well as Zasp52 , also interact directly with α-actinin ( Figure 9B ) , it appears likely that all proteins containing a Zasp PDZ domain can do so . In contrast , the closest relative , LMO7 , cannot bind α-actinin through its PDZ domain [48] . Finally , the Alp/Enigma family proteins ZASP , PDLIM1 , and ALP contain an additional area partially overlapping with the Zasp-like motif , which interacts with the α-actinin rod domain , giving rise to the possibility that one Zasp molecule may bind two α-actinin dimers or one α-actinin dimer in antiparallel configuration [43] , [49]–[51] . Recently , a comparative evolutionary study showed that only four ortholog groups localize to Z-discs in all bilaterian species: Zasp , α-actinin , titin , and MLP proteins , suggesting that these four protein groups could be sufficient for assembly and function of Z-discs [52] . Our results indicate that multiple members of Zasp PDZ domain proteins may be required to provide the critical mass for Z body assembly . Together with the well-documented role of ZASP mutations in human disease , our data indicate that these proteins occupy a central place in muscle assembly and function .
The following fly stocks were used: G00189 ( GFP-Zasp52 ) , zcl0663 ( GFP-Zasp66 ) , Dmef2-Gal4 , UAS-Dcr Dmef2-Gal4 , Gla/CTG [CyO , P{GAL4-twi . G}2 . 2 , P{UAS-2×EGFP}AH2 . 2] from the Bloomington Drosophila Stock Center , Zasp52 [2] , Act88F-Gal4 [23] ( kindly provided by RM Cripps ) , UAS-iZasp66 ( KK112973 , transformant 102980 ) , UAS-iZasp52ex20 ( KK101276 , transformant 106177 ) , UAS-iZasp67 ( GD8245 , transformant 17414 and KK111478 , transformant 103225 ) from the Vienna Drosophila RNAi Center , and UAS-iZasp52ex16 ( this study ) . Act88F-Gal4; UAS-iZasp66 was generated by standard genetic crosses . For the genetic interaction assay , UAS-iZasp66 was crossed to Dmef2-Gal4 , and Zasp52/CTG; Dmef2-Gal4 was crossed to UAS-iZasp66 or y w and incubated at 25°C . After 14 days of incubation , pupal lethality was scored ( ratio of non-green pupae to straight-winged adults ) . Live imaging was performed as described [53] . Briefly , two-hour egg-lays were aged for 24 h at 18°C to obtain late stage 16 embryos . Embryos were dechorionated in 50% bleach for 2 min , rinsed , dried and mounted in halocarbon 27 oil on a gas-permeable membrane ( Coy Lab Products , MI , USA ) . Micrographs were taken on a Zeiss LSM 510 Meta laser scanning confocal microscope at room temperature with a 40×1 . 3 Plan-NEOFLUAR oil immersion objective at 2× zoom . Every minute , 7 z-sections were captured at 512×512 resolution , 2× scan average , with each slice being separated by 1 µm ( total scan time: 14 sec ) . After collection , sections were separated and exported as TIFF files using Volocity software ( PerkinElmer , Ontario , Canada ) . For RT-PCR and qPCR , the average of two independent experiments of triplicate-PCR reactions is presented . Total RNA was isolated from 20 adult flies using Trizol , and reverse transcribed using SuperScript II Reverse Transriptase according to the manufacturer's instructions ( Life Technologies , Ontario , Canada ) and run on a T3000 Thermocycler for RT-PCR ( Biometra , Montreal Biotech Inc . , Quebec , Canada ) . Quantitative PCR reactions were performed with the iQ SYBR Green Supermix Kit on a C1000 Thermocycler ( Bio-Rad , Ontario , Canada ) . Quantification was performed with the comparative threshold cycle method on Bio-Rad CFX Manager software . Both rp49 and β-Tubulin were used as normalization controls in a single experiment . Primer pairs used: Zasp66-F TACCGTACAACTCCGCTGGT , Zasp66-R TCATGGTAGTCGTGTCCTGG , Zasp67-F CTTAATGGTGGGCAGCAAGTC , Zasp67-R GACAGTGAGGTGCCGAATTT , tubulin-F ACATCCCGCCCCGTGGTC , tubulin-R AGAAAGCCTTGCGCCTGAACATAG , rp49-F TACAGGCCCAAGATCGTGAAG , rp49-R GACGCACTCTGTTGTCGATACC . Zcl0663 ( GFP-Zasp66 ) was verified by PCR with primer pair GFP-fwd CGACCACTACCAGCAGAACA and Zasp66-rev GATGCACCTACGCCACTTTT . For UAS-iZasp52ex16 we designed oligos generating a 21 nt siRNA targeting exon 16 with the DSIR algorithm ( http://biodev . extra . cea . fr/DSIR/DSIR . html ) . It should deplete all long isoforms except one containing exon 17 ( exon numbering according to [9] . Oligos ctagcagtCTGCACATTGCAGCTGTTGCAtagttatattcaagcataTGCAACAGCTGCAATGTGCAGgcg and aattcgcCTGCACATTGCAGCTGTTGCAtatgcttgaatataactaTGCAACAGCTGCAATGTGCAGactg were annealed and cloned into Valium20 [54] . A sequence-verified clone was injected into vermillion attP2 ( 3L ) flies ( Genetic Services Inc . , MA , USA ) . 50 adult fly thoraces were cut in half and were homogenized in lysis buffer ( 25 mM Tris-HCl pH 8 , 150 mM NaCl , 1 mM EDTA , 0 . 5% TritonX-100 , 5% glycerol , and complete EDTA-free Protease inhibitor cocktail; Roche , Quebec , Canada ) . Protein extracts were then incubated with prewashed GFP-Trap-M anti-GFP beads ( ChromoTek , Germany ) for 2 h at 4°C . After incubation the beads were washed three times with wash buffer ( 10 mM Tris-HCl pH 8 , 150 mM NaCl , 0 . 1% TritonX-100 and complete EDTA-free Protease inhibitor cocktail ) , and bound proteins were eluted by boiling in 2× SDS sample buffer . Eluates were analyzed by SDS-PAGE and by immunoblotting . Antibodies were used at the following ratios: rat anti-α-actinin antibody at 1∶2000 ( Babraham Institute , UK ) ; rabbit anti-GFP antibody at 1∶400 ( Clontech , CA , USA ) . The immunoreaction was visualized by ECL ( GE Healthcare , Ontario , Canada ) . Zasp66-RB was synthesized by GenScript ( New Jersey , USA ) , and cloned into pRSETA ( Life Technologies , Ontario , Canada ) ; GST-Zasp52Alp ( amino acids 1–357 containing the PDZ domain , the Zasp-like motif and the LIM1 domain ) was cloned from EST LP01550 into pGEX-5X-1 ( GE Healthcare , Ontario , Canada ) , then overexpressed and purified by standard procedures . For pull- down assays GST-ZaspAlp was added to glutathione paramagnetic beads ( Promega , WI , USA ) in 20 mM Tris-HCl pH 8 , 100 mM NaCl , 1 mM MgCl2 , 1 mM DTT , 0 . 2% TritonX-100 , 10% glycerol , and incubated for 2 h at 4°C . This was followed by a 1 h blocking step of GST-ZaspAlp-coupled beads in the above buffer with 5% BSA . Subsequently , rabbit skeletal muscle α-actinin ( Cytoskeleton , CO , USA ) was added and incubated for another 2 h at 4°C . Final washes were in the above buffer with 500 mM NaCl and 0 . 5% TritonX-100 . Beads were resuspended in SDS sample buffer and analyzed by SDS-PAGE and immunoblotting . 6×His-Zasp66-RB was coupled to Ni-NTA agarose beads ( Qiagen , Ontario , Canada ) in 20 mM Tris-HCl pH 8 , 100 mM NaCl , 1 mM MgCl2 , 1 mM DTT , 10 mM Imidazole , 0 . 2% Triton X-100 . α-actinin pull-down and washes were carried out using this buffer . We used the following primary antibodies for immunofluorescent stainings of IFMs: rat anti-Zasp52 [9] , mouse anti-α-actinin [55] , rabbit anti-obscurin Ig14-16 [56] , rat anti-kettin MAC155 [57] . Half thoraces were glycerinated ( 20 mM Na-Phosphate pH 7 . 2 , 2 mM MgCl2 , 2 mM EGTA , 5 mM DTT , 0 . 5% Triton X-100 , 50% glycerol ) overnight at −20°C . IFMs were dissected , washed in relaxing solution ( 20 mM Na-phosphate pH 7 . 2 , 5 mM MgCl2 , 5 mM ATP , 5 mM EGTA ) with protease inhibitors , and separated into single myofibrils or left as a whole [56] . Primary antibody incubation was carried out overnight , followed by washes in relaxing solution , and 1–3 h incubations of secondary antibodies and Alexa 594-phalloidin ( Life Technologies , Ontario , Canada ) . Pupal IFMs were dissected in relaxing solution , fixed in 4% paraformaldehyde in relaxing solution , and labeled with primary and secondary antibodies . Antibodies against Zasp52 , obscurin and kettin were diluted 1∶100 in relaxing solution . Anti-α-actinin antibody was diluted 1∶10 . Fluorescently labeled secondary antibodies of the Alexa series ( Life Technologies , Ontario , Canada ) were used at 1∶400 . Samples were mounted in ProLong Gold antifade solution ( Life Technologies , Ontario , Canada ) . Images were obtained on a LSM 510 Meta confocal microscope using a 63×1 . 4 NA Plan Apo oil immersion objective ( Carl Zeiss , Germany ) . Thoraces were treated with 5 mM MOPS pH 6 . 8 , 150 mM KCl , 5 mM EGTA , 5 mM ATP , 1% Triton X-100 for 2 h at 4°C , followed by overnight incubation in the same buffer without Triton X-100 but 50% glycerol . This was repeated for a second time . Samples were then washed in rigor solution ( 5 mM MOPS pH 6 . 8 , 40 mM KCl , 5 mM EGTA , 5 mM MgCl2 , 5 mM NaN3 ) and fixed in 3% glutaraldehyde , 0 . 2% tannic acid in 20 mM MOPS pH 6 . 8 , 5 mM EGTA , 5 mM MgCl2 , 5 mM NaN3 for 2 h at 4°C . Secondary fixation and embedding were as described before [15] , [20] , [56] . For preparation of non-glycerinated samples , hemithoraces were dissected in rigor solution and directly transferred into primary fixative . Images were acquired on a Tecnai 12 transmission electron microscope ( FEI , Japan ) . | Muscles are comprised of huge , multinucleated cells that feature a highly organized cytoskeletal architecture consisting of variable numbers of myofibrils , whose formation is not well understood . Each myofibril is an array of sarcomeres , the smallest contractile unit of muscles . The contractile system consists of actin filaments anchored at the Z-discs , which border the sarcomere , and myosin filaments anchored at the M-line in the middle of the sarcomere . In this study , we reveal the role of the Alp/Enigma family proteins Zasp52 , Zasp66 , and Zasp67 that are required for both the initial assembly and the stability of myofibrils . We also gain new insights into myofibril assembly by following it via live imaging . We can show that Zasp52 and Zasp66 cooperate in Z-disc assembly by binding directly to α-actinin , by interacting genetically , and by forming a ternary complex with α-actinin . As a result , the combined defects of removing both Zasp52 and Zasp66 or Zasp52 and another family member , Zasp67 , are much more severe than would be expected from the additive defects of the single mutants . Thus , our results suggest that multiple Alp/Enigma family proteins are required to form the critical complex to initiate Z-disc and myofibril assembly . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"immunocytochemistry",
"animal",
"genetics",
"genetic",
"mutation",
"protein",
"interactions",
"gene",
"function",
"developmental",
"biology",
"organism",
"development",
"molecular",
"development",
"molecular",
"genetics",
"morphogenesis",
"sequence",
"analysis",
"cytochemistry",
"proteins",
"adhesion",
"molecules",
"organogenesis",
"biology",
"biochemistry",
"cytoskeletal",
"proteins",
"gene",
"identification",
"and",
"analysis",
"genetics",
"computational",
"biology",
"genetics",
"and",
"genomics"
] | 2013 | Alp/Enigma Family Proteins Cooperate in Z-Disc Formation and Myofibril Assembly |
Sleep contributes to cognitive functioning and is sufficient to alter brain morphology and function . However , mechanisms underlying sleep regulation remain poorly understood . In mammals , tumor necrosis factor-alpha ( TNFα ) is known to regulate sleep , and cytokine expression may represent an evolutionarily ancient mechanism in sleep regulation . Here we show that the Drosophila TNFα homologue , Eiger , mediates sleep in flies . We show that knockdown of Eiger in astrocytes , but not in neurons , significantly reduces sleep duration , and total loss-of-function reduces the homeostatic response to sleep loss . In addition , we show that neuronal , but not astrocyte , expression of the TNFα receptor superfamily member , Wengen , is necessary for sleep deprivation-induced homeostatic response and for mediating increases in sleep in response to human TNFα . These data identify a novel astrocyte-to-neuron signaling mechanism in the regulation of sleep homeostasis and show that the Drosophila cytokine , Eiger , represents an evolutionarily conserved mechanism of sleep regulation across phylogeny .
The function of sleep and the neurobiology underlying the detrimental effects of sleep deprivation on physiological function are poorly understood . Understanding phylogenetic conservation of mechanisms that regulate sleep and the neurobiological consequences associated with sleep loss may provide important clues to sleep function . Although the function of sleep is unknown , sleep is regulated by a combination of genetic and experience-dependent environmental influences . For example , environmental conditions such as temperature , light levels , and social interactions are sufficient to modify sleep duration or sleep architecture [1] . Additionally , many genes have been found to play a role in the regulation of sleep [2–6] . Specifically , in the fruit fly Drosophila melanogaster , molecules involved in inflammation and immunological functioning have been implicated in sleep regulation including nuclear factor kappa-light-chain-enhancer of activated B cells ( NF-kB ) [7] and Jun amino-terminal kinases ( JNK ) [8] . In mammals , another such molecule is the pro-inflammatory cytokine , tumor necrosis factor alpha ( TNFα ) . TNFα functions as a sleep regulatory substance [9 , 10] . TNFα promotes non-REM ( NREM ) sleep when administered centrally or peripherally [9 , 11] , and neural activity [12] and sleep deprivation [13] both increase production of TNFα . TNFα is also an important regulator of homeostatic synaptic scaling [14] and experience-dependent plasticity [15] . TNFα signaling has been implicated in cognitive impairment due to sleep deprivation in humans [16] , and in many human diseases including cancers [17] , depression [18] , and neurodegenerative disorders like Alzheimer’s disease [19–21] . Many of these diseases are associated with sleep abnormalities and impaired cognition . Given the importance TNFα signaling for sleep in mammals and the role of inflammation and immunological genes in sleep regulation in the fly , we examined the Drosophila melanogaster TNFα homologue , Eiger , to determine evolutionarily conserved mechanisms of cytokines in sleep . Eiger was first discovered in a p-element screen to examine cell death mechanisms [22] and was the first cytokine to be discovered in the fruit fly [23] . The majority of the work on Eiger has focused on cell death , apoptosis , infection , and JNK signaling pathways . Here , we assessed behavioral and molecular properties of sleep and Eiger expression in subpopulations of fruit fly central nervous system ( CNS ) cells in order to determine the role of Eiger and the Eiger receptor , Wengen , in the regulation of sleep duration and sleep architecture . The data presented here suggest that Eiger underlies phylogenetically conserved mechanism of sleep regulation and identifies a unique astrocyte-to-neuron mechanism regulating sleep behavior . The fly is an ideal organism to examine phylogenetically conserved mechanisms of sleep regulation . The genetics of the fly are easily manipulated and sleep in the fruit fly shares many of the features of mammalian sleep [24 , 25] . Specifically , sleep in the fruit fly is dependent upon waking experience and homeostatically regulated . Similar to mammals , flies show an increase in sleep following prolonged periods of sleep deprivation [4 , 5 , 26 , 27] . Additionally , courtship conditioning and social enrichment both result in increased sleep [28] . Learning tasks increase the expression of plasticity-related molecules and increase sleep duration [1 , 29] and similar observations have been made in mammalian systems [30 , 31] . In mammals , when a brain area is activated during wake , cytokine expression increases [32 , 33] and the subsequent EEG delta power in that brain area increases [31 , 34 , 35] . Taken together , these data suggest that the fly may be an important model to test for evolutionarily conserved mechanisms of cytokine dependent sleep regulation .
Eiger shares significant homology to human TNFα and TNFα superfamily members [22] . To examine if human TNFα is sufficient to increase sleep in Drosophila similarly to its effect in mammals [34] , we performed intra-thoracic injections of recombinant human TNFα into wild-type flies and measured sleep for 3 days . Sleep duration increased with increasing concentrations of TNFα ( Fig 1A ) . Whereas PBS injected controls did not show significantly different sleep profiles from non-injected controls ( Fig 1B ) . Specifically , total sleep duration was significantly increased in a dose-response manner in the 2 hours immediately following injection ( Fig 1C ) and in the subsequent 8 daytime hours ( Fig 1D ) . TNFα injections ( 100nM ) also resulted in a significant increase in the length of daytime sleep bouts ( Fig 1E ) . Nighttime sleep bout duration ( Fig 1F ) and bout length ( Fig 1G ) were unaffected by injection of TNFα . 48 hours after injection , total sleep duration in all TNFα treated groups was not significantly different from PBS injected controls ( Fig 1H ) . These data indicate that the injection resulted in an acute change in sleep duration and there were no long-term consequences of the varied concentrations of TNFα on sleep duration . TNFα is a potent sleep regulatory substance in mammals . TNFα levels correlate positively with slow wave sleep and sleep duration [36] and inhibition of TNFα reduces sleep duration in rats and rabbits [37 , 38] . To determine the role of Eiger in sleep regulation in the fly , we tested Eiger 1 ( EGR1 ) and Eiger 3 ( EGR3 ) loss-of-function mutants and examined the necessity of Eiger in regulating baseline sleep . Both Eiger mutants displayed a reduction in total sleep duration ( Fig 2A and 2B ) compared to wild-type Canton S ( CS ) control flies . During the day ( Fig 2C–2E ) and the night ( Fig 2F–2H ) , the EGR1 and EGR3 mutants showed decreased sleep duration ( Fig 2C and 2F ) and reduced sleep bout length ( Fig 2D and 2G ) . The number of sleep bouts was reduced during the day ( Fig 2E ) and increased during the night ( Fig 2H ) in both mutant lines compared to control flies . During the night , a reduction of sleep bout length along with an increase in the number of sleep bouts suggests that these flies experience significant nighttime sleep fragmentation . These data are in agreement with the mammalian literature , which suggests that inhibition of TNFα is sufficient to reduce sleep . To assess the effect of the EGR1 and EGR3 mutations on circadian function and behavioral activity levels , we examined groups of CS , EGR1 , and EGR3 flies during 4 days on a 12:12 , Light:Dark cycle and then transitioned all flies to Dark:Dark ( Fig 3A–3C , red star indicates transition to Dark:Dark ) . Then , using the ActogramJ Plugin for ImageJ , we generated double plotted actograms and circadian periods were calculated . None of the three lines tested , including the Eiger mutant lines , showed significantly impaired circadian rhythms during Dark:Dark . Additionally , we plotted the average daily locomotor activity during the Light:Dark period and Dark:Dark periods ( Fig 3D and 3E ) and found that the Eiger mutants were hyperactive at different times of the day compared to CS control flies during the Light:Dark ( Fig 3D ) and Dark:Dark ( Fig 3E ) periods . These activity data show that the active periods in the Eiger mutants correspond to the timing of reduced sleep shown in Fig 2 . Human TNFα is sufficient to increase sleep in wild-type flies ( see above ) , however , it is unknown whether TNFα injections would have the same effect in short-sleeping Eiger mutant flies . To examine if human TNFα is sufficient to increase sleep in the Eiger mutants similarly to wild-type flies , we performed TNFα injections on the EGR1 and EGR3 mutants . Injection of 1000nM human TNFα was sufficient to increase sleep in both the EGR1 ( Fig 4A ) and EGR3 ( Fig 4B ) mutants . The increase in sleep duration was not associated with an increase in average bout length during the day or the night for either mutant line ( Fig 4C–4F ) . In mammalian brains , TNFα is produced in neurons and glia [14 , 39]; in the fly , Eiger is enriched in astrocytes [40] . To determine whether the Eiger loss-of-function sleep duration phenotype is confined to a particular cell type , we used a previously validated UAS-Eiger RNA-interference ( RNAi ) line [22] to knockdown Eiger expression in neurons ( using Elav-Gal4 ) versus astrocytes ( using Alrm-Gal4 ) . We assessed sleep duration in these knockdown lines compared to their corresponding control parental lines . RNAi knockdown in neurons had no effect on total sleep duration ( Fig 5A ) , while astrocyte knockdown significantly reduced total sleep duration ( Fig 5B ) . Wengen was the first TNFα receptor superfamily member found in Drosophila , and it plays a critical role in transducing Eiger mediated cell death signals [41] . To further elucidate the role of Eiger in sleep regulation , we explored the cell-type specificity of the Eiger receptor , Wengen , in regulating baseline sleep and sleep homeostasis . We expressed the previously validated UAS-Wengen RNAi [41] in neurons ( Fig 6A and 6B ) and astrocytes ( Fig 6C and 6D ) using the same Gal4 drivers as presented in Fig 5 ( Elav-Gal4 , neurons; Alrm-Gal4 , astrocytes ) . Baseline sleep was unaffected by neuronal ( Fig 6A ) and astrocyte driven ( Fig 6C ) Wengen-RNAi expression . Following 12 hours of sleep deprivation , neuronal RNAi-meditated knockdown of Wengen showed a significant reduction in sleep rebound compared to control flies ( Fig 6B ) , while RNAi-mediated knockdown of Wengen in astrocytes showed a normal homeostatic response ( Fig 6D ) . This neuronal knockdown of Wengen resulted in a similar sleep homeostasis phenotype as Eiger loss-of-function mutants ( S1 Fig ) , pointing to an astrocyte-to-neuron signaling pathway via Eiger and Wengen , which may be a critical regulator of the homeostatic response to sleep deprivation . To determine whether neuronal Wengen is necessary for TNFα effects on regulating sleep duration or architecture in Drosophila , as in Figs 1 and 4 , we injected 1000nM human recombinant TNFα ( or vehicle ) in UAS-WGN RNAi/ Elav-Gal4 flies and measured changes in baseline sleep . TNFα failed to alter baseline sleep in neuronal expressing Wengen RNAi flies ( Fig 6E–6G ) . These data suggest that neuronal expression of Wengen is required for TNFα induced changes in sleep . Social experience ( i . e . enrichment ) increases sleep duration in Drosophila . This may reflect interactions between synaptic plasticity and sleep need . Given that TNFα mediates experience-dependent plasticity in mammals [15] , and social experience increases sleep duration in flies [1 , 28 , 42] , we next investigated whether Eiger also mediated experience-dependent changes in sleep duration . We exposed CS , EGR1 , and EGR3 flies to either social enrichment or social isolation for 5 days and then measured subsequent changes in sleep . Following 5 days of social enrichment , wild-type CS flies showed the typical increase in sleep ( M = 178 minutes , P<0 . 05 ) , whereas , neither EGR1 ( M = -41 minutes , P>0 . 05 ) nor EGR3 ( M = 8 minutes , P>0 . 05 ) mutants increased sleep ( Fig 7A ) compared to their isolated counterparts . To assess whether astrocytes contribute to this enrichment phenotype , we expressed Eiger-RNAi in astrocytes and performed the same social enrichment paradigm as above . The parental lines for the Eiger-RNAi/+ and Alrm Gal4/+ flies both showed an increase in sleep in response to social enrichment ( black and grey bars in Fig 7B ) . However , expression of the Eiger-RNAi construct in astrocytes was sufficient to block experience-dependent increases in sleep ( Fig 7B ) .
We show that human TNFα is sufficient to increase sleep in the fruit fly , similar to what is seen in mammals [9 , 11] . Conversely , Eiger loss-of-function using the EGR1 and EGR3 mutants resulted in a significant reduction in total sleep; a similar phenotype as seen in mammals when TNFα is inhibited [37 , 38] . These data suggest that TNFα and Eiger share functional homology across species . Additionally , we showed cell-type specificity of Eiger signaling in these sleep phenotypes . Reduction of Eiger in astrocytes , a population of cells with enriched Eiger expression [43] , resulted in a significant reduction in sleep duration compared to genetic controls , while no effect was observed in flies with neuronal Eiger knock-down . In addition to identifying a role of Eiger in baseline sleep regulation , our loss-of function studies found that Eiger was required for experience-dependent increases in sleep following social enrichment . Further , we found that Eiger expression in astrocytes was required for the increase in sleep characteristic of social enrichment . Therefore Eiger signaling in astrocytes regulates sleep in the fly , revealing a novel mechanism for understanding how astrocytes contribute to sleep . Neuronal expression of the Eiger receptor , Wengen , is necessary for a normal homeostatic response to sleep deprivation . RNAi knockdown of Wengen in neurons or astrocytes did not alter baseline sleep duration . However , the neuronal RNAi knockdown , but not the astrocyte-based RNAi knockdown , significantly reduced homeostatic sleep rebound following sleep deprivation . This suggests that the Wengen receptor may be required for cytokine dependent homeostatic sleep regulation and that astrocyte-neuron communication via Eiger/Wengen signaling plays an important role in sleep homeostasis , but not in regulating baseline sleep duration . Our data indicate that increases in sleep following sleep loss requires Wengen receptor signaling in neurons . This finding is supported by the finding that the human TNFα injections in neuronal expressing Wengen RNAi flies failed at increasing sleep duration . We hypothesize that different TNFα receptors may differentially mediate the regulation of baseline sleep duration and sleep homeostasis . Grindelwald , for example , is another Drosophila TNFα receptor superfamily member that has been shown to bind Eiger and whose function differs significantly from Wengen [44] . Therefore , downstream signaling from Eiger via Grindelwald in neurons and/or glia may explain this dissociation between Eiger signaling and baseline sleep duration versus its homeostatic regulation . Future studies are required to elucidate the signaling mechanisms of Eiger and Grindelwald from those between Eiger and Wengen . Such studies would additionally benefit from over-expression analysis and from injection of Eiger instead of TNFα . Genetic background can often be sufficient to affect behaviors such as sleep . However , through the use of multiple experimental approaches to elucidate Eiger function , we gained confidence that the phenotypes described here are not merely artifacts of genetic background . This conclusion is supported by the use of two loss-of-function mutants , one RNAi line , social enrichment experiments , and injections of human TNFα . It is possible that the human TNFα injection studies may have resulted in peripheral activation of a TNFα superfamily receptor that regulates sleep duration . However , our experiments showing that neuronal expression of Wengen is required for human TNFα induced sleep increases rules that out as the sole explanation for our findings . Few studies have investigated how glial-specific factors and mechanisms regulate sleep [45–47] . We previously identified the astrocyte-derived fatty acid-binding protein 7 ( Fabp7 ) as an important regulator of sleep across diverse species [45] . Fabp7 deficiency in astrocytes decreased the activity NF-kB and TNFα following lipopolysaccharide stimulation [48] , which promotes inflammatory cytokine production [49] . Since Fabp7 can influence both NF-kB and TNFα expression in astrocytes , this suggests that Fabp7 may be directing its effects on sleep through cytokine and/or inflammatory pathways . It is currently unknown whether NF-kB and TNFα can in turn affect Fabp7 expression . A reciprocal feedback loop involving astrocytes would provide a unique framework for differentiating baseline sleep regulation from sleep homeostatic processes .
Canton S , w1118 , UAS-Eiger RNAi , UAS-Wengen RNAi , Elav-Gal4 , and Alrm-Gal4 stocks were all obtained from the Bloomington Drosophila Stock Center ( Indiana University ) . EGR1 and EGR3 mutants on CS background were obtained from Dr . Masayuki Miura ( University of Tokyo ) . Flies were cultured at 25°C , 60% humidity , maintained on a 12:12 Light:Dark cycle , on Nutri-fly Bloomington Formulation fly food ( Genesee Scientific , San Diego , CA . ) . Newly eclosed virgin female flies were collected from culture vials daily under CO2 anesthesia and housed in groups of ~30 prior to experimentation . Adult 4–6 day old wild-type w1118 , virgin , female flies were injected at ZT3 with 9 . 2nL of 1nM , 10nM , 100nM or 1000nM recombinant human TNFα ( R&D Systems , Minneapolis , MN ) in phosphate-buffered saline ( PBS ) using a glass pulled pipette and an automatic nanoliter injector ( Drummond Scientific , Broomall , PA ) . Flies were injected under CO2 anesthesia in the ventrolateral surface of the fly thorax and immediately placed into the behavioral monitoring system for measuring sleep . Female flies 4–7 days after eclosion were used for all sleep studies . Flies were mouth aspirated into 5mm x 65mm ( outside diameter x length ) polycarbonate recording tubes ( Trikinetics , Waltham , MA ) with food ( Bloomington Nutrifly formula ) on one end and yarn plugs on the other . Sleep parameters were continuously evaluated using the Trikinetics Drosophila activity monitoring system ( DAMS; Trikinetics , Waltham , MA ) as described previously [24] . One acclimatization day was followed by two days of baseline sleep recording , one 12-hour sleep deprivation period during the dark period , or one 24-hour sleep deprivation period ( see main text ) , and two full days of recovery sleep . Sleep deprivation was performed as previously reported [24] using a custom-built Sleep Nullifying APpartus ( SNAP; Machine Shop , Washington University in Saint Louis ) . Sleep homeostasis was calculated for each individual as the ratio of minutes of sleep gained above baseline during recovery divided by minutes of sleep lost during sleep deprivation ( min gained/min lost ) . To standardize the environmental conditions during critical periods of brain development , virgin female flies were collected upon eclosion and maintained in same-sex vials containing ~30 flies for 2 days . This protocol keeps environmental conditions constant between subsequently isolated and enriched flies for the first 2 days of adult life . Three-day old flies were then divided into a socially isolated group , in which flies were individually housed in 65 mm glass tubes , and a socially enriched group , consisting of 50 female flies housed in a single vial as previously described [1] . After five days of social enrichment/isolation , flies were placed into clean 65 mm glass tubes and sleep was recorded for three days using the Trikinetics DAMS . Statistics were calculated using Graphpad Prism software . Student’s t-test , one-way ANOVA , two-way ANOVA , and Tukey post-hoc analysis were used for analyses . Sleep data were analyzed by averaging across multiple experiments . Flies that did not survive the entire experimental paradigm were removed from data analysis . The ActogramJ plugin for ImageJ was used to quantify circadian period and daily activity levels . | Every animal sleeps , from fruit flies to humans . However , the function of sleep is still currently unknown . Identifying conserved mechanisms of sleep regulation in evolutionarily ancient organisms may help us to understand the function of sleep . Therefore , we have examined whether Eiger , the homologue of the cytokine tumor necrosis factor-alpha ( TNFα ) , regulates sleep in the fruit fly as it does in higher mammals . Cytokines are inflammatory molecules and are typically elevated following infection or fever and may contribute to increased sleepiness when sick . We found that , in the fruit fly , Eiger regulates sleep duration just like TNFα does in mammals: increasing cytokine levels increased sleep duration while decreasing Eiger reduced sleep . In addition , we found that Eiger expression in glial astrocytes , is responsible for the alteration in sleep duration . We also examined the necessity of Eiger receptor activation on neurons and found that astrocyte-to-neuron communication was required for regulating the normal increases in sleep following sleep deprivation . These data show that a novel cytokine mechanism regulates sleep in flies and mammals , and provides insight into conserved roles of astrocytes in sleep behavior . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] | [
"invertebrates",
"innate",
"immune",
"system",
"medicine",
"and",
"health",
"sciences",
"sleep",
"deprivation",
"immune",
"physiology",
"cytokines",
"rna",
"interference",
"sleep",
"astrocytes",
"immunology",
"vertebrates",
"neuroscience",
"macroglial",
"cells",
"animals",
"mammals",
"animal",
"models",
"physiological",
"processes",
"developmental",
"biology",
"drosophila",
"melanogaster",
"model",
"organisms",
"experimental",
"organism",
"systems",
"molecular",
"development",
"epigenetics",
"drosophila",
"research",
"and",
"analysis",
"methods",
"animal",
"cells",
"genetic",
"interference",
"animal",
"studies",
"gene",
"expression",
"glial",
"cells",
"insects",
"immune",
"system",
"arthropoda",
"biochemistry",
"rna",
"cellular",
"neuroscience",
"eukaryota",
"cell",
"biology",
"nucleic",
"acids",
"physiology",
"neurology",
"genetics",
"neurons",
"biology",
"and",
"life",
"sciences",
"cellular",
"types",
"amniotes",
"organisms"
] | 2018 | Astrocyte expression of the Drosophila TNF-alpha homologue, Eiger, regulates sleep in flies |
Bacterial secondary metabolites are widely used as antibiotics , anticancer drugs , insecticides and food additives . Attempts to engineer their biosynthetic gene clusters ( BGCs ) to produce unnatural metabolites with improved properties are often frustrated by the unpredictability and complexity of the enzymes that synthesize these molecules , suggesting that genetic changes within BGCs are limited by specific constraints . Here , by performing a systematic computational analysis of BGC evolution , we derive evidence for three findings that shed light on the ways in which , despite these constraints , nature successfully invents new molecules: 1 ) BGCs for complex molecules often evolve through the successive merger of smaller sub-clusters , which function as independent evolutionary entities . 2 ) An important subset of polyketide synthases and nonribosomal peptide synthetases evolve by concerted evolution , which generates sets of sequence-homogenized domains that may hold promise for engineering efforts since they exhibit a high degree of functional interoperability , 3 ) Individual BGC families evolve in distinct ways , suggesting that design strategies should take into account family-specific functional constraints . These findings suggest novel strategies for using synthetic biology to rationally engineer biosynthetic pathways .
Bacterial secondary metabolites are widely used as pharmaceutical , agricultural , and dietary agents . They consist of many classes of compounds including polyketides ( PKs ) , nonribosomal peptides ( NRPs ) , ribosomally synthesized and post-translationally modified peptides ( RiPPs ) , terpenoids , saccharides , and a plethora of hybrids . The genetic basis for this rich molecular diversity can be found in biosynthetic gene clusters ( BGCs ) , physically clustered groups of genes that encode the enzymatic pathways necessary to construct specific chemicals [1] , [2] . The diversity of extant natural products and BGCs raises important questions about their evolutionary origin . These include the basic question of how Nature invents new molecules , and a series of applied questions relevant to biotechnology: for example , the evolutionary modularity of NRP and PK BGCs has long been seen as a feature that might allow large libraries of new compounds to be generated by mixing and matching their constituent domains and modules [3] . However , although there have been notable successes [4]–[6] , the majority of combinatorially generated pathways appear to be nonfunctional [4] . More recently , advanced synthetic biology approaches to pathway engineering have been frustrated by the complexity and unpredictability of metabolic enzymes , particularly NRPSs and PKSs [7] , [8]: unlike LEGO bricks , their constituent domains and modules do not ‘fit’ together universally , but only function effectively in specific pathway contexts . Regardless of these apparent constraints to genetic change , Nature appears to have been quite successful at engineering biosynthetic pathways through the process of gene cluster evolution: even a conservative estimate suggests that the number of broad biosynthetic gene cluster families that have evolved exceeds 6 , 000 [8] , most of which contain multiple BGCs that synthesize derivatives of a common scaffold . Hence , a detailed study of evolutionary patterns within various BGC families has the potential to offer a new inroad into effective BGC engineering , through mimicry of Nature's evolutionary design strategies . So far , insights into the key principles underlying the evolution of BGC architectures and repertoires have been derived from limited case studies [9]–[13] , which lack sufficient detail about the generality of the underlying mechanisms . Here , we systematically quantify the strategies that make evolution so successful at engineering BGC diversity . Through a detailed computational analysis of a recently generated dataset of 732 known and 10 , 724 predicted prokaryotic BGCs [8] , we find that the rates of evolutionary events , such as insertions , deletions and duplications within BGCs , are much higher than those seen in comparable gene clusters involved in primary metabolism . Furthermore , distinct sub-clusters consisting of co-evolving genes appear to constitute relatively independent building blocks that play key roles in the evolution of larger BGCs encoding the biosynthesis of complex metabolites . Finally , BGC families encoding the production of polyketides and nonribosomal peptides evolve in family-specific modes , in many of which we observe an unexpectedly large role for concerted evolution [14] , [15] driven by internal recombinations . Based on these observations , we offer several recommendations for establishing new modes of evolution-guided BGC engineering .
The large diversity of BGCs observed throughout the prokaryotic tree of life [8] suggests that BGCs evolve rapidly . Indeed , when we systematically quantified different evolutionary events by mutually comparing all gene clusters in our data set ( Table S1 ) , we found not only that they may have been transferred horizontally at high frequency ( Fig . 1a and Figure S1 ) , but also display exceptionally high rates of insertions , deletions , duplications and rearrangements ( Fig . 1b ) . While the percentage of gene cluster pairs related by an indel is independent of gene cluster size , the distribution of indel sizes shows a long tail that includes 195 indels of 10 kb or more ( Fig . 1c ) . As expected , these large indels are more commonly found in larger gene clusters , where they indicate either the merger of one gene cluster fragment with another or the loss of a gene cluster fragment from a larger cluster ( see examples in Figure S2 ) . Phylogenetic profiling [16] showed that many such BGC fragments – here termed sub-clusters – appear to evolve in a correlated fashion: 884 different motifs of adjacent Pfam domains ( out of 7 , 641 found ) were shown to co-evolve significantly more often than not ( P<0 . 001 ) , based on the χ2 test . These motifs comprise 591 different Pfam domains and have an average length of 5 . 3 domains ( Table S2 ) . As expected , they include many well-known and widely conserved motifs that appear to be linked to specific sub-functionalities of gene clusters , such as precursor biosynthesis , transport or synthesis of a specific chemical moiety , and motifs belonging to modular BGC architectures of NRPSs and PKSs ( e . g . , C-A-T and KS-AT-T [17] ) . Earlier evidence has suggested complex mosaic patterns of sub-cluster sharing for some BGCs , such as those involved in the production of glycopeptides [18] . To further explore the role of sub-cluster sharing in the evolution of BGCs , we manually compiled a set of 35 BGCs that are rich in sub-clusters that have a known connection with a specific chemical moiety . We then used this data set to construct a network in which the nodes represent BGCs and the edges denote a sub-cluster that a pair of BGCs has in common ( Fig . 2 ) . Three observations were particularly notable ( Fig . 2 ) . First , >60% of the coding capacity of some BGCs ( e . g . , those encoding vancomycin and rubradirin [19] ) is composed of individually conserved sub-clusters ( note that this is not entirely reflected in the depiction of the rubradirin gene cluster in Fig . 2b , where only those sub-clusters are highlighted that are shared with other depicted BGCs ) . This supports a “bricks and mortar” model of gene cluster evolution in which gene clusters are composed of large , modular “bricks” ( sub-clusters ) that encode key building blocks and individual genes ( the “mortar” ) that encode functions such as tailoring , regulation and transport . During evolution , both bricks and mortar ( scaffold and tailoring ) may remain the same , only the tailoring may change or the scaffold itself may change . Second , the same sub-cluster commonly appears in otherwise unrelated BGCs , and multiple unrelated sub-clusters can be found in a single parent gene cluster , indicating that sub-clusters are independent evolutionary entities . Third , sub-clusters are not static; they are loosely organized around a core set of genes , but gene gain/loss leads to chemical changes in the corresponding part structure: for example , gene clusters encoding molecules such as everninomicin [20] , simocyclinone [21] and polyketomycin [22] have different variants of deoxysugar sub-clusters , which lead to subtle variations in the final chemical structures . Although the complex patterns of sub-cluster sharing , in which various sub-clusters are shared between otherwise completely different gene clusters ( Fig . 2 ) , indicate that BGCs may evolve by the successive merger of sub-clusters , this does not mean that every case where sub-clusters are shared points to an independent sub-cluster transfer event . For example , the KS domains of the diverse range of ansamycin type I PKS BGCs that harbor AHBA sub-clusters are almost completely monophyletic ( Figure S3 ) , indicating that the macrolactam- and AHBA-producing sub-clusters have been co-evolving for a long time ( instead of multiple independent AHBA sub-cluster acquisitions having occurred in different macrolactam-producing polyketide BGCs ) . Hence , the multi-hybrid rubradirin gene cluster might have arisen from a rifamycin-like ancestor ( most rubradirin KS domains are monophyletic with rifamycin KS domains , see Figure S3 ) that already harbored the combination of a modular type I PKS sub-cluster and an AHBA biosynthesis sub-cluster , and which then acquired new sub-clusters for the biosynthesis of the aminocoumarin , 3 , 4-dihydroxydipicolinate and nitrosugar moieties ( which are not found in any other closely related ansamycins ) . Contrary to the shared evolutionary histories of AHBA and ansamycin type I PKS sub-clusters , a clear example of sub-cluster transfer between BGCs of different types can be seen for 6-methylsalicylic acid ( MSAS ) /orsellinic acid ( OSAS ) sub-clusters , as inferred from a maximum-likelihood phylogenetic tree of MSAS/OSAS-producing iterative PKSs ( Figure S4 ) . The topology of this tree strongly indicates that MSAS/OSAS sub-clusters have largely evolved independent of the scaffold types of their parent gene clusters ( Figure S4 ) , and that they have been transferred between multiple types of BGCs during their evolutionary past . In conclusion , in the context of the bricks-and-mortar analogy , some bricks move around between different structures more often than others . Finally , we should note that there are also BGC families which evolve over long periods of time without major changes to the gene cluster architecture or the scaffold of the core molecule made: for example , the large family of over >1 , 000 aryl polyene BGCs that we described recently [8] has not undergone any major sub-cluster transfers , aside from the inclusion of the dialkylresorcinol sub-cluster in the BGCs from some CFB group bacteria . The products of many of these BGCs are likely to be entirely identical , while remaining differences between the molecules mostly concern differential tailoring of the same scaffold . Many chemical scaffold types of secondary metabolite classes are quite distinct , which raises the question of how BGC families encoding the synthesis of distinct scaffolds are related . To assess this question , we calculated the proportion and similarity of Pfam domains shared between all pairs of BGCs within our data set of 732 known gene clusters using multiple sequence alignments for each Pfam domain ( Fig . 3 ) and looked specifically for close homologues of BGCs just outside their immediate family . Even though of course sequence similarity alone does not provide conclusive evidence on evolutionary histories , the analysis did suggest that unexpected evolutionary connections might exist between natural products of different scaffold types . For example , the Streptomyces gene cluster encoding the lipopeptide antibiotic daptomycin [23] is surprisingly similar to Mycobacterium glycopeptidolipid ( GPL ) gene clusters [24] ( Figure S5 ) . When we performed a more in-depth analysis through a phylogenetic analysis of condensation domains , we indeed found that GPL domains consistently cluster together with domains from the NRPSs that synthesize daptomycin ( Figure S6 ) . Although both daptomycin and the GPLs are lipopeptides , the Mycobacterium GPLs are shorter ( tetrapeptide vs . tridecapeptide ) , cell-wall-associated rather than diffusible , linear rather than cyclic , and originate from an actinomycete genus that is not closely related to Streptomyces . Likewise , one of the strongest matches for the gene cluster encoding the immunosuppressant rapamycin [25] , apart from the closely related FK520 [26] and meridamycin [27] , [28] BGCs , was the gene cluster for pladienolide [29] , a polyketide of unrelated structure with a distinct biological activity ( inhibition of the splicing factor SF3b instead of TOR ) . Strikingly , based on phylogenetic trees of their constituent ketosynthase ( KS ) and acyltransferase ( AT ) domains , the meridamycin gene cluster is more closely related to the pladienolide BGC than to those encoding rapamycin and FK520 , the molecules to which it is often compared ( Fig . 3 ) . These examples suggest that closely related sets of protein domains can be reconfigured by evolution to yield a new scaffold that is chemically and biologically distinct . The phylogenetic trees of KS and AT domains from our data set of known BGCs revealed another unexpected finding: in spite of the structural similarity of rapamycin and FK520 , 63% of the constituent domains of their polyketide synthases ( PKSs ) cluster into entirely separate clades ( Fig . 3b , see also Figure S8 which shows that relevant bootstrap values are almost all above 90 ) . Even more remarkably , 14 out of 16 domains responsible for the biosynthesis of the sub-structure shared between these two molecules ( shown in red in Fig . 3c ) do not cluster together with the corresponding domain from the assembly line for the other molecule . This pattern of homology is consistent with a phenomenon called ‘concerted evolution’ , the homogenization of DNA sequences within a given repetitive family caused by high rates of internal recombination [14] , [15] . Given the similar sizes and architectures of the gene clusters and the structural similarity of their products , this is a much more parsimonious explanation for the patterns observed than convergent evolution of multiple similar gene clusters through successive duplication of an ancestral single-module PKS . Notably , previous phylogenetic analyses of PKS domains have also observed BGC-specific clades of PKS domains [10] , [30] , but not to the extent observed here for such closely related gene clusters: the fact that such a strong pattern is even observed for the AT domains of two different gene clusters that encode the same molecule [27] , [28] , meridamycin , shows that the underlying process may operate on very short time scales , and that recombination can remove almost all traces of independent evolution of these PKS modules . In the case of the rapamycin family , recombinations are likely to occur neutrally and have no effect on the structure of the small molecule product ( rapamycin , meridamycin and FK520 ) , whereas in other cases , single crossovers within or between gene clusters may dramatically change the modular architecture of a synthase [30] . Near-neutral changes brought about by gene conversion may occur at higher rates for some domains or domain types than for others: in the meridamycin gene clusters , no signs of gene conversion could ( yet ) be observed for KS domains , even though gene conversion manifested itself clearly when comparing the meridamycin clusters with those encoding rapamycin , FK520 and pladienolide . On the contrary , AT domain gene conversion was widespread even between the two meridamycin gene clusters . We speculate that for these BGCs , gene conversion events get fixated in the population at lower rates for KS domains because not all KS sequences work equally well for different polyketide chain lengths that occur at different points of the assembly line , so that the changes brought about by a conversion event are less neutral than for AT domains . Mapping of rapamycin family PKS sequence mutations onto the 3D structure of an AT- and KS-containing protein further supports this hypothesis ( Figure S7a ) , showing widespread sequence variability at almost every position in the AT domains , except for the residues near the substrate binding site ( Figure S7b ) . Mutations in KS domains , on the other hand , are mostly restricted to the regions in vicinity ( around the core ) of the substrate-binding site and the dimerization interface ( Figure S7c ) , suggesting their importance in influencing substrate selectivity . Concerted evolution is not peculiar to the rapamycin family ( Figure S8 ) . For the gene clusters encoding the biosynthesis of the mutually closely related macrolides erythromycin [31] , oleandomycin [32] and pikromycin [33] , BGC-specific branching appeared to occur for both KS and AT domains , similar to the pattern for rapamycin , FK520 , meridamycin and pladienolide . However , for the ansamycin antibiotics macbecin [34] , geldanamycin [35] and herbimycin [36] , and the antifungals pimaricin [37] , nystatin [38] and amphotericin [39] , BGC-specific branching occurs only for AT domains , and not for KS domains . Finally , corroborating earlier observations [40] , domains from the trans-AT PKS gene clusters encoding pederin [41] and psymberin [42] do not show any BGC-specific branching at all . We observed that certain NRPS gene clusters also show signs of concerted evolution: a clear BGC-specific branching pattern pointing to concerted evolution can be seen for the A domains and most of the C domains of the gene clusters encoding the biosynthesis of the closely related calcium-dependent lipopeptides daptomycin [23] , A54145 [43] and CDA [44] . However , the glycopeptide gene clusters encoding the biosynthesis of balhimycin [45] , teicoplanin [46] and A40926 [47] showed no such pattern at all: almost all domains cluster in groups corresponding to domains in the same positions in the assembly line . Collectively , these observations suggest that concerted evolution is a key mechanism driving the evolution of NRPS and PKS gene sequences , but the extent to which it happens depends on family-specific functional constraints as well as on the presence of other evolutionary forces acting upon a gene cluster . Our qualitative model of PKS/NRPS evolution ( Fig . 4 ) , which summarizes the interplay of concerted evolution with other evolutionary mechanisms , is relevant to PKS/NRPS engineering efforts: the highly homologous sets of domains generated by concerted evolution are more likely to be mutually interoperable than domain sets chosen at random , and might therefore be attractive building blocks for synthetic biological engineering of biosynthetic pathways . To understand more generally how PKS and NRPS BGCs evolve , we set out to measure the contributions of concerted evolution , duplication , and divergence to the evolution of all multimodular PKS and NRPS BGCs in both our known and predicted BGC data sets . We first collected and quantified 25 different features describing the nature of gene cluster sequences and the relationships among their constituent domains ( see methods for details ) . A principal component analysis ( PCA ) and hierarchical clustering using these features can distinguish many of the well-known gene cluster families from our data set of known BGCs ( Figure S9 , Fig . 5a ) . Two features in particular , the ‘internal similarity index’ and the ‘vertical evolution index’ , explain much of the variation in terms of the modes of evolution of different classes of gene clusters ( Fig . 5b ) . At the level of individual domains , we find that there are four primary mechanisms by which NRPS and PKS BGCs evolve ( Fig . 5c–f , Figure S10 ) . Firstly , gene clusters encoding glycopeptides , calcium-dependent lipopeptides and macrolides/polyethers appear to be most repetitive , pointing to a history of module duplications and/or a prominent influence of concerted evolution . The syringopeptin NRPS [48] and mycolactone PKS [49] are extreme examples of this: both are likely to have evolved recently by subsequent module duplications and concerted evolution . Secondly , we sometimes observed gradients of the internal homology p-values from the N- to C-termini of large synthases , suggesting that some gene clusters evolve to encode the synthesis of larger molecules by iterative duplication of their most N-terminal module , would have the effect of extending an intermediate NRP or PK by the addition of a new starter unit . Thirdly , a group of BGCs including the ones that encode the polyketides psymberin [42] and erythrochelin [50] show a ‘vertical’ type of evolution , in which the domains appear to evolve independently , with perhaps occasional domain swapping with related gene clusters , as has been suggested previously [40] . Finally , there are many gene clusters showing a ‘mixed’ mode of evolution , in which one or more of the above mechanisms are combined . For example , NRP siderophore gene clusters show some signs of internal recombinations , but at the same time many domains show no high mutual similarity . Like the trans-AT PKS gene clusters , they seem to have a higher tendency to recruit domains from dissimilar gene clusters . This recruitment over larger evolutionary distances appears to be a general feature of NRPS gene clusters as opposed to PKS gene clusters , and might be related to the wider range of possible substrates for NRPSs , which often require BGC-specific sub-pathways for the synthesis of a dedicated monomer [51] . The observation of so many different evolutionary mechanisms of gene cluster evolution begs the question which circumstances lead to the birth and death of BGCs over evolutionary time . Are all BGCs that are detected bioinformatically also still intact and functional , or might many of them have degenerated and entered a nonfunctional state ? The absence or presence of nonfunctional genetic units ( e . g . , pseudogenes or pseudo-gene-clusters ) is largely governed by the evolutionary population dynamics of the species . Many bacteria live in large effective population sizes and have relatively short generation times , leading to very strong purifying selection and , consequently , rigorous genome streamlining [52] . Hence , BGCs that become nonfunctional will be quickly lost in such organisms if they do not provide any evolutionary advantage . Notably , some bacteria in fact occur in smaller population sizes and/or regularly go through population bottlenecks , leading to altogether different evolutionary dynamics [53]: in such cases , a range of pseudogenized gene clusters can sometimes still be observed that have not been purged from the genome yet [54] . On the whole , however , these appear to be rather the exception than the rule [55] . Concerning the birth of new gene cluster architectures , large effective population sizes and short generation times also suggest that BGC modifications should immediately confer an evolutionary advantage in order to be maintained; on the other hand , frequent changes in population size may affect the probability of mutations to be fixated in the population [56] . Alternatively , neutral mutations could hitchhike with strongly adaptive mutations within or close to the same gene cluster . Concerning the physical growth of gene clusters , it should be noted that new enzymes may already be recruited to a biosynthetic pathway before their genes are physically recruited to the gene cluster , and such an addition to a pathway could evolve through , e . g . , positive selection acting on promiscuous enzyme activities or substrate specificities [57] . The precise reason for and evolutionary mechanism of clustering of biosynthetic genes in bacteria itself is still largely an unanswered question [58] . Our analysis of BGC evolution will enable new approaches to BGC engineering informed by the mechanisms by which BGCs evolve naturally . Our results suggest that efforts to engineer the biosynthesis of unnatural natural products could be more successful by observing the modes by which specific BGC classes evolve in nature . For example , conglomerate molecules consisting of multiple different chemical moieties could be designed by engineering BGCs consisting of novel combinations of sub-clusters . Such an effort could be guided by information taken from evolutionary comparisons , which would offer suggestions about which sub-clusters are most likely to function together , based on how often evolution has successfully forged combinations between them . Furthermore , our evolutionary analysis of NRPS and PKS gene clusters suggests that concerted evolution has created sets of domains within gene clusters that are highly homologous . These domain sets are more likely to be mutually interoperable than domain sets chosen at random , and might therefore be of great utility in future engineering efforts . Also , evolutionary strategies towards generating larger and more complex compounds could be mimicked by N-terminally extending certain types of NRPS/PKS gene clusters by duplicating and then carefully modifying the first assembly-line module . Overall , in combination with new synthetic biology techniques that may soon enable the rapid assembly of thousands of clusters from a common set of parts [59]–[61] , our results suggest a new approach for re-engaging gene cluster engineering in a manner informed by the mechanisms by which gene clusters have naturally evolved .
To remove highly similar genomes from these analyses , we used the AMPHORA [62] ( August 10th , 2010 ) dataset , which contains gene sequences from 562 organisms for 30 universally conserved genes . Genes from these organisms were compared using sequence identities based on MUSCLE [63] multiple sequence alignments . This resulted in 30 distances between each pair of organisms . The distributions of distances of all pairs were tested for normality using a Shapiro-Wilk test . An organism distance map was then built with distances defined as the mean distances of AMPHORA genes . The resulting distance map was clustered using default settings in MCL [64] , and only one member of each cluster was kept for further analyses . This left us with total of 408 organisms . To search for histidine and tryptophan biosynthetic operons , we modified ClusterFinder [8] . Pfam [65] IDs associated with the histidine biosynthesis pathway ( PF00475 , PF00815 , PF01174 , PF01502 , PF01634 , PF04864 , PF08029 , and PF08645 ) or with the tryptophan biosynthesis pathway ( PF00218 , PF00290 , PF00465 , PF00697 , PF01220 , PF01264 , PF01487 , PF04715 , and PF08501 ) were acquired from JGI IMG [66] . Trp or His operons were defined as gene clusters containing at least one of these domains with a probability >0 . 5 and containing at least two of the domains in total . Among 408 organisms searched , 350 His and 288 Trp biosynthesis operons were identified in 271 and 248 different organisms , respectively . The average number of domains per predicted gene cluster were 2 . 9 and 3 . 1 , respectively . Best matching sequence homologs of a query protein domain from a biosynthetic or primary metabolic gene cluster were obtained using MUSCLE [63] multiple sequence alignments . The distance between the organism containing the query protein domain and the organism with the best matching sequence homolog was determined based on 16S rRNA sequence similarity . Best matching sequence homologs of all protein domains that are in Pfam are included in the organism similarity histograms ( Fig . 1a ) . For each BGC , a two-dimensional array of the size corresponding to the numbers of consecutive protein domains that are in Pfam database ( rows ) and 408 selected organisms ( columns ) ( see “Comparison of HGT with primary metabolism” ) was created . The cells in the array consisted of sequence identities between a given domain from a BGC and the most homologous domain ( which is also predicted as part of a BGC ) from a given organism . Next , we calculated a Pearson product-moment correlation coefficient ( correlation coefficient ) for each possible pair of protein domains ( rows ) , resulting into a new matrix , a correlation matrix , of the size corresponding to the number of protein domains ( rows from the initial array ) in both dimensions . To take rearrangements into account , we reordered rows and columns of the correlation matrix based on hierarchical clustering of the correlation matrix in both dimensions . We then parsed linear motifs that are likely to evolve in a correlated fashion by selecting consecutive pairs of domains in this reordered correlation matrix ( consecutive fields on the first offset diagonal ) with correlation coefficient >0 . 5 . The analysis was repeated by setting the correlation coefficient cutoff to >0 . 65 and >0 . 8 . Each motif was divided into all possible sub-motifs of sizes between 2 domains and the total number of domains in a motif . To determine the significance of a ( sub ) motif occurrence , we next compared the number of ( sub ) motif occurrences to the number of all possible ( sub ) motif occurrences in all BGCs that did not pass the correlation coefficient cutoff . Pearson's χ2 test with Bonferroni correction was applied to test for statistical significance , with the null hypothesis stating that the two values are equal . We performed an all-versus-all alignment of nucleotide sequences of known and predicted BGCs using the blastn algorithm . Gene cluster sequences were divided into blocks of 1 kb , and then mapped to the most homologous blocks from other gene clusters , as well as from the same gene cluster ( to test for genomic duplications ) . 56% of the blocks ( 118 , 320 out of 212 , 176 ) did not map to any homologous regions in the same or other BGCs with >70% identity . Evolutionary events ( insertions/deletions , duplications and rearrangements ) were detected by a custom-made Python script ( Data S1 ) comparing each alignment of two-gene clusters having at least three matching blocks with >70% identity . Rearrangements were defined as an identified difference in the order of 1-kb blocks in an otherwise conserved ( piece of ) gene cluster , such as when A1-A2-A3-A4-A5 matches to B1-B4-B3-B2-B5 in an alignment of two BGCs A and B . Indels were defined as 1-kb blocks present in one gene cluster but not in the other gene cluster , such as when A1-A2-A3-A4-A5-A6 matches to B1-B2-B5-B6 in an alignment . To make these inferences more reliable , a constraint was used that the flanking regions ( of size > = 2 kb ) of each indel breakpoint must be homologous between query and hit gene cluster , and the block order must be conserved between them . Finally , duplications were defined as 1-kb blocks that had the best hit towards another block in its own gene cluster , and having a higher copy number in one gene cluster than in the other , such as when A1-A2-A3-A2-A3-A4-A5 aligns to B1-B2-B3-B4-B5 , while the mutual sequence identity between the A2 and A3 pairs is higher than between any of the A2/A3 blocks and B2 or B3 . For a given BGC pair , we first calculated sequence identities between all Pfam domain pairs of each Pfam ID , using MUSCLE [63] multiple sequence alignments . A BGC sequence similarity index was defined as the Jaccard index with the size of the intersection represented by the number of Pfam pairs whose sequence identities were higher than the best 10% alignments of all Pfam domains of the same Pfam ID . Taking into account the underlying distributions of sequence identities between all domain sequences prevented misinterpretation of simpler sequence similarity metrics ( e . g . , an absolute sequence identity threshold ) when different evolutionary rates apply to different protein families . We define structural similarity of a given BGC product pair as the Tanimoto coefficient between the two SMILES strings , using linear-path fingerprints ( FP2 ) from Open Babel [67] . Sub-clusters with known functions from experimentally characterized gene clusters were manually collected from the literature . Sub-cluster sharing between gene clusters from the training set was calculated using blastp [68] . The minimum requirement used to identify a shared sub-cluster between two BGCs was sharing either 75% of the genes with >45% average sequence identity , 50% of the genes with >50% average sequence identity , or 25% of the genes with 70% identity . To account for different modes of sequence evolution of different sub-cluster types , these values were adjusted with sub-cluster type-specific cutoffs to obtain a good match between genetic similarity and chemical similarity ( Table S3 ) . The final sub-cluster sharing network was drawn with Cytoscape [69] . To study patterns of evolution in multimodular NRPS and PKS gene clusters , a range of features was calculated describing key characteristics of these gene clusters . The first set of features was based on the topologies of intra-BGC domain similarity networks ( with protein domains and sequence similarity representing nodes and edges , respectively ) and consisted of the average clustering coefficient , average sequence similarity , graph transitivity , number of 2–4 node cliques , number of connected components in a graph with sequence similarity >50% , and average neighbor degree . We also included as features the number of different Pfam domain types in a BGC , the total number of domains in a BGC , the average number of domains per gene , and the averages and standard errors of best-matching pair sequence identities and internal BGC similarity indices . Two evolutionary indices were also added: the internal similarity index and the vertical evolution index . To obtain the internal similarity index of a gene cluster , we calculated for each of its NRPS/PKS domains the p-value of its closest blastp match inside the gene cluster , given the distribution of the percent identities of all within-gene-cluster blastp hits of all domains of that domain type in the complete set of gene clusters . The internal similarity index was then calculated from these numbers as the mean of all inverse p-values . The same inverse p-values were used for plotting the internal domain similarity across gene clusters . The vertical evolution index of a gene cluster was calculated as the average difference between the p-value of the top 10 percent identities of a domain's blastp hits to all domains from other gene clusters with the p-values of the Lin distances of the gene clusters to the host gene clusters of each of the top 10 hit domains . Consequently , gene clusters with domains with highly similar closest hits to domains in dissimilar gene clusters get a low value , while gene clusters with domains with dissimilar closest hits to domains in similar gene clusters get a high value . PCA analysis was performed with the aforementioned features as an input . Compound types were assigned using the classifications taken from the primary literature . | Bacterial secondary metabolites mediate a broad range of microbe-microbe and microbe-host interactions , and are widely used in human medicine , agriculture and manufacturing . Despite recent advances in synthetic biology , efforts to engineer their biosynthetic genes for the production of unnatural variants are frustrated by a high failure rate . In an effort to better understand what types of genetic changes are most likely to lead to successful improvements , we systematically analyzed the ways in which biosynthetic genes naturally evolve to generate new compounds . We show that large gene clusters appear to evolve through the merger of sub-clusters , which function independently , and are promising units for cluster engineering . Moreover , a subset of gene clusters evolve by concerted evolution , which generates sets of interoperable domains that may enable predictable domain swapping . Finally , many biosynthetic gene clusters evolve in family-specific modes that differ greatly from each other . Overall , this quantitative perspective on the ways in which gene clusters naturally evolve suggests novel strategies for using synthetic biology to engineer the production of unnatural metabolites . | [
"Abstract",
"Introduction",
"Results/Discussion",
"Methods"
] | [
"biotechnology",
"biochemistry",
"secondary",
"metabolism",
"genomics",
"genome",
"evolution",
"genetics",
"synthetic",
"biology",
"biosynthesis",
"biology",
"and",
"life",
"sciences",
"comparative",
"genomics",
"computational",
"biology",
"evolutionary",
"biology",
"microbiology",
"metabolism",
"microbial",
"genomics"
] | 2014 | A Systematic Computational Analysis of Biosynthetic Gene Cluster Evolution: Lessons for Engineering Biosynthesis |
DNA double-strand breaks ( DSBs ) can be repaired by homologous recombination ( HR ) , which can involve Holliday junction ( HJ ) intermediates that are ultimately resolved by nucleolytic enzymes . An N-terminal fragment of human GEN1 has recently been shown to act as a Holliday junction resolvase , but little is known about the role of GEN-1 in vivo . Holliday junction resolution signifies the completion of DNA repair , a step that may be coupled to signaling proteins that regulate cell cycle progression in response to DNA damage . Using forward genetic approaches , we identified a Caenorhabditis elegans dual function DNA double-strand break repair and DNA damage signaling protein orthologous to the human GEN1 Holliday junction resolving enzyme . GEN-1 has biochemical activities related to the human enzyme and facilitates repair of DNA double-strand breaks , but is not essential for DNA double-strand break repair during meiotic recombination . Mutational analysis reveals that the DNA damage-signaling function of GEN-1 is separable from its role in DNA repair . GEN-1 promotes germ cell cycle arrest and apoptosis via a pathway that acts in parallel to the canonical DNA damage response pathway mediated by RPA loading , CHK1 activation , and CEP-1/p53–mediated apoptosis induction . Furthermore , GEN-1 acts redundantly with the 9-1-1 complex to ensure genome stability . Our study suggests that GEN-1 might act as a dual function Holliday junction resolvase that may coordinate DNA damage signaling with a late step in DNA double-strand break repair .
The correct maintenance and duplication of genetic information is constantly challenged by genotoxic stress . DNA double-strand breaks ( DSBs ) are amongst the most deleterious lesions . DSBs can be induced by ionizing irradiation ( IR ) or caused by the stalling of DNA replication forks . In response to DSBs , cells activate conserved DNA damage checkpoint pathways that lead to DNA repair , to a transient cell cycle arrest , or to apoptosis and senescence . The full activation of DNA damage response pathways and DSB repair by homologous recombination ( HR ) depends on a series of nucleolytic processing events . Following DSB formation , broken ends are resected in a 5′ to 3′ direction to generate 3′ single-strand overhangs [1] . These tails are coated by RPA1 molecules , which in turn are thought to lead to the recruitment of the ATR checkpoint kinase [2] . This kinase , and the related kinase ATM , appear to be directly targeted to DNA double-strand breaks to act at the apex of the DNA damage signaling cascade [3] . The DNA damage specific clamp loader comprised of Rad17 bound to the four smallest RFC subunits [4] recruits a PCNA-like complex referred to as “9-1-1” complex to the dsDNA–ssDNA transition of resected DNA ends [5]–[7] . The 9-1-1 complex is needed for full ATR activation [8] , [9] . DSB repair by HR proceeds by replacing RPA1 with the RAD51 recombinase [10] , [11] . The resulting nucleoprotein filament invades an intact donor DNA to form a D-loop structure . The invading strand is extended using the intact donor strand as a template . Annealing of the 3′ single-stranded tail of the second resected DNA end to the displaced donor DNA strand ( second end capture ) , and DNA ligation lead to the formation of a double Holliday junction ( dHJ ) intermediate ( for a review , see [12] ) . This dHJ must be resolved either through cleavage by Holliday junction ( HJ ) -resolving enzymes or through “dissolution” by the combined activity of the Blooms helicase and topoisomerase III [13] , [14] . Prototypic HJ resolving enzymes are nucleases that resolve HJs by introducing two symmetrical cleavages that result in either crossover or non-crossover products , depending on which strands are cleaved . Cuts made by junction-resolving enzymes need to be perfectly symmetrical so that products can be re-ligated , thus requiring no further processing events for HJ resolution [15] , [16] . Until recently , the molecular nature of canonical HJ resolvases in animals and plants remained enigmatic despite the observation of HJ-resolving activity in cellular extracts over many years [17] , [18] . Resolving enzymes have been purified from bacteriophages , bacteria and archea but the only eukaryotic resolving enzymes that had been discovered until recently were S . cerevisiae Cce1 and S . pombe Ydc2 , both of which act in mitochondria [15] , [19] , [20] . One possible pathway of HJ resolution involves the conserved MUS81/EME1 complex , probably the principal meiotic resolution activity in fission yeast [21] , [22] , although mouse as well as budding yeast strains lacking Mus81 only have very minor meiotic phenotypes [23] , [24] . By comparison with known resolving enzymes , the in vitro properties of this complex currently appear somewhat imprecise , and more akin to flap endonuclease action [25] , yet recent evidence suggests that this complex can lead to productive HJ resolution [26] , [27] . In addition , it was recently shown that a complex between the SLX4 scaffold protein and the SLX1 nuclease can act as an HJ resolving enzyme [28]–[30] . Intriguingly , SLX4 also interacts with the XPF and MUS81 nucleases , providing a scaffold for repairing multiple DNA structures and the sequential action of SLX4/nuclease complexes on HJ might rather be described as HJ processing that nevertheless ultimately leads to HJ resolution [28]–[32] . While recent studies suggest that the SLX4 scaffold and associated nucleases may promote nuclease-dependent HJ resolution , an independent enzyme with HJ resolution activity , mammalian GEN1 , was identified in vitro via biochemical fractionation [33] . GEN1 generates symmetrical cleavage in a manner similar to the E . coli RuvC junction-resolving enzyme . In parallel the budding yeast GEN1 ortholog Yen1 was identified as a resolving enzyme using functional genomics based approaches . The biological functions of human GEN1 are unclear , and the deletion of yen1 has no obvious DNA repair defect [34] . Furthermore , it is not clear how or even if the processing of HJs is coordinated with DNA damage signaling . Recent evidence suggests that deleting the budding yeast yen1 in conjunction with mus81 leads to MMS hyper-sensitivity [34] . Also , expressing human GEN1 in fission yeast , which does not encode for a gen-1/yen1 homolog , complements the meiotic defect associated with mus81 [35] . We use the Caenorhabditis elegans germ line as a genetic system to study DNA repair and DNA damage response pathways . As part of the C . elegans life cycle invariant embryonic cell divisions occur very rapidly . Embryonic cells tolerate a relatively high level of DNA damage using error prone polymerases , possibly a result of natural selection that favours rapid embryonic divisions at the expense of genome integrity [36] . In contrast , the C . elegans germ line , which is the only proliferative tissue in adult worms , displays longer cell cycles and is much more sensitive to DNA damaging agents . The gonad contains various germ cell types arranged in a distal to proximal gradient of differentiation ( Figure 1G ) . At the distal end of the gonad cell proliferation occurs in a mitotic stem cell compartment . This compartment is followed by the transition zone where early events of prophase I , such as double strand break generation and the initiation of meiotic chromosome pairing occur . Proximal to the transition zone most germ cells are arrested in the G2 cell cycle phase and reside in meiotic pachytene , where homologous chromosomes are tightly aligned to each other as part of the synaptonemal complex . Germ cells subsequently complete meiosis and concomitantly undergo oogenesis and arrest at the metaphase I stage of meiosis before they are fertilized at the proximal end of the gonad . It takes approximately 20 hours for pachytene stage cells to mature and get fertilized , while the progression of mitotic germ cells till fertilization takes approximately 48 hours [37] , [38] . DNA damage such as IR or replication stress , leads to prolonged G2 cell cycle arrest of mitotic germ cells . In addition , late stage meiotic pachytene cells undergo apoptosis in response to DNA damaging agents [39] . DNA damage responses are mediated by components of a conserved DNA damage response pathway ( for a review see [40] ) . Upstream sensors and transducers such as the worm ATR ortholog or components of the 9-1-1 complex promote all DNA damage responses including DNA repair , cell cycle arrest and apoptosis . In contrast , downstream effectors like cep-1 , which encode the sole primordial p53-like protein of C . elegans , are only needed for IR-induced apoptosis [41] . Using unbiased genetic screening and positional cloning approaches we have cloned the C . elegans homolog of the human GEN1 HJ resolving enzyme . C . elegans gen-1 is required for repair of DNA damage-induced DSBs . Surprisingly , gen-1 mutants are defective in IR-induced cell cycle arrest and apoptosis , indicating that GEN-1 promotes DNA damage signaling . The function of GEN-1 in apoptosis induction is independent of the ATL-1 ( C . elegans ATR ) -dependent induction of the CEP-1/p53 target EGL-1 . Our results suggest that GEN-1 is a dual function protein required for the repair of DSBs as well as for DNA damage checkpoint signaling .
To uncover new genes involved in DNA damage response signaling , we chose an unbiased genetic approach and screened for C . elegans mutants hypersensitive to IR and/or defective in DNA damage-induced cell cycle arrest and apoptosis . During C . elegans development , the majority of cell divisions occur during embryogenesis . In contrast , germ cell proliferation , which commences with two germ cells at the L1 larval stage , predominates in the following three larval stages and continues in adult worms , where all somatic cells are post-mitotic , but continued germ cell proliferation results in a steady state level of ∼500 germ cells . In order to select for mutants hypersensitive to IR , worms mutagenised with ethyl methane sulphonate ( EMS ) were irradiated at the L1 stage with 60 Gy of IR . This dose of radiation does not overtly affect germ cell proliferation in wild type worms while mutants hypersensitive to IR display reduced levels of fertility ( data not shown ) . Out of 906 F2 lines screened , 3 mutations ( yp30 , yp42 and yp45 ) were recovered for the yp30 complementation group , each of which was derived from an independently mutagenised Po animal . In C . elegans , treatment of L4 larvae with IR leads to the activation of a DNA damage response checkpoint pathway that triggers apoptosis of meiotic pachytene stage germ cells , and a transient halt of mitotic germ cell proliferation leading to enlarged cells [39] . This latter phenotype results from continued cellular growth in the absence of cell division . The yp30 complementation group does not enlarge mitotic germ cells upon IR of L4 larvae , similar to the mrt-2 ( e2663 ) checkpoint mutant ( Figure 1A and 1B , Figure S1B ) [39] , and is partially defective in DNA damage-induced apoptosis ( Figure 1C ) . We did not find any further mutants , which were defective in both IR-induced cell cycle arrest and apoptosis like the yp30 complementation group ( data not shown ) . To show that yp30 germ cells do indeed fail to arrest cell cycle progression after irradiation , we stained N2 wild type and yp30 mutants with antibodies against phosphorylated tyrosine-15 CDK-1 , which serves as a G2 marker [42] , [43] . We found that wild type germ cells arrest in G2 , whereas yp30 germ cells fail to do so ( Figure 1D ) , a finding we confirmed using a YFP::Cyclin B1 fusion construct as a G2 marker ( Figure S1A ) . To clone the gene corresponding to yp30 , we followed the cell cycle arrest-defective phenotype in backcrossing , SNP-mapping and complementation experiments and positioned yp30 close to the centre of chromosome III , between dpy-17 and unc-32 , to an interval of approximately 135 , 000 base pairs ( Figure 1E , data not shown ) . Sequencing this interval in yp30 worms revealed two mutations , one in an intergenic region , and one that leads to a premature stop codon in a gene encoding for a conserved nuclease we refer to as gen-1 ( see below , Figure 1F , Figure 2A ) . yp42 and yp45 also contained the same C to T point mutation as gen-1 ( yp30 ) , but lacked the intergenic mutation found in yp30 . Sequencing of the gen-1 cDNA confirmed the predicted gen-1 cDNA sequence and the predicted intron-exon structure of gen-1 , available from Wormbase ( http://www . wormbase . org/; data not shown ) . A gen-1 deletion allele , gen-1 ( tm2940 ) ( Figure 2A ) obtained from the Japanese C . elegans knockout consortium , as well as gen-1 ( RNAi ) , similarly lead to a cell cycle arrest defect upon irradiation ( Figure S1B ) . In addition , the same phenotype was observed in gen-1 ( tm2940 ) /gen-1 ( yp30 ) trans-heterozygotes ( Figure S1B ) . Time course and dose response experiments revealed that gen-1 ( tm2940 ) , gen-1 ( yp30 ) and mrt-2 ( e2663 ) worms are equally defective in IR induced cell cycle arrest ( Figure S2 ) . Furthermore , gen-1 ( tm2940 ) is largely defective in DNA damage-induced apoptosis , similar to cep-1 ( lg12501 ) , a deletion mutant of the C . elegans p53-like gene cep-1 [41] , [44] ( Figure 1C ) . In summary , our data reveal that gen-1 is required for IR-induced apoptosis and cell cycle arrest in C . elegans germ cells . Sequence alignments suggest that GEN-1 is a member of the XPG super-family of nucleases , members of which contain two conserved domains referred to as N and I domains as part of the catalytic centre [45] ( Figure 2A , Figure S3 ) . GEN-1 contains putative catalytic residues known to be required for nuclease activity , these are aspartate 77 located in the N domain and glutamate 791 within the I domain of human XPG ( Figure S3 ) [46] . We analyzed all XPG-like genes from fungi , some invertebrates ( including other nematodes ) and vertebrates , finding that all sequences clustered within four classes of nucleases GEN1 , XPG , FEN1 and EXO1 , with high probability scores in all species except for fission yeast that does not encode for GEN1 ( Figure 2B and 2C ) . XPG is involved in nucleotide excision repair [47] , FEN1 is a flap nuclease involved in lagging strand DNA replication [48] , [49] , and EXO1 is implicated in genomic stability , telomere integrity [50] as well as DSB end resection [51] , [52] . GEN1 was first biochemically characterized based on its flap endonuclease activity in Drosophila , and named DmGEN1 ( XPG like Endonuclease-1 ) [53] . A human GEN1 N-terminal fragment was recently purified from HeLa cell extracts , and shown to have robust Holliday junction-resolving activity . Moreover an activity was also found in crude preparations of the budding yeast ortholog Yen1p [33] . The gen-1 ( yp30 ) mutation leads to the expression of a C-terminally truncated protein that does not affect the putative catalytic centre ( Figure 2A and 2D ) . In contrast , the tm2940 deletion is predicted to eliminate the majority of the I domain and is likely to be a null allele , as anti-GEN-1 antibodies detected GEN-1 protein for wild-type and gen-1 ( yp30 ) strains but not for gen-1 ( tm2940 ) ( Figure 2A and 2D , Figure S7B and S7C ) . To determine if C . elegans GEN-1 exhibits Holliday junction-resolving activity in vitro , as predicted from homology to the human GEN1 , recombinant wild type GEN-1 , GEN-1 ( yp30 ) and an E135A mutant were expressed and purified , the latter bearing a mutation in one of the putative nuclease active site residues ( Figure S4A ) . A Holliday junction-resolving enzyme should symmetrically cleave Holliday junctions and be specific for four-way DNA junctions . We tested for GEN-1 nuclease activity on two four-way DNA junctions . Jbm5 contains a 12 base pair homologous core through which the branch point can migrate [54] , and X26 contains a 26 base pair core and bears sequences unrelated to Jbm5 [33] . Using both four-way junction substrates we observed specific cleavage using GEN-1 and GEN-1 ( yp30 ) recombinant enzymes ( Figure 2E ) . Using both substrates the same cleavage pattern was observed on opposite strands as expected from symmetry ( Jbm5 , Figure 2F , data not shown ) . To confirm structural specificity towards four-way DNA junctions , we tested whether C . elegans GEN-1 showed specific nuclease activity towards a variety of other substrates , including single-stranded , blunt double-stranded DNA , a dsDNA substrate with a 3′ single-stranded overhang , and a 5′ flap structure . We observed no specific cleavage of any of these substrates with C . elegans GEN-1 ( Figure 2G , Figure S4C ) . Comparing the cleavage to that generated by the human GEN1 ( comprising amino acids 1-527 ) , we find that the major Jbm5 cleavage product resulting from incubation with human GEN1 also occurs upon incubation with the C . elegans protein ( Figure S4B ) . Human GEN1 also showed an activity towards 5′ flap structures as reported previously [33] ( Figure S4C ) . The enzymatic activity of the recombinant C . elegans enzyme is relatively low; we thus cannot exclude the possibility that C . elegans GEN-1 also shows a 5′ flap activity , albeit we did not observe such an activity in overexposed gels and multiple repeat experiments . We speculate that the low activity of recombinant C . elegans GEN-1 might be due to improper folding . Alternatively , the worm nuclease might require post-translational modifications , interacting proteins or activation by proteolytic cleavage to become fully activated as a HJ resolving enzyme , thereby preventing us from undertaking a more thorough analysis of its biochemical properties at the present time . Nevertheless , the cleavage introduced into the four-way junction by C . elegans GEN-1 as well as the orthologous relationship to human GEN1 and budding yeast Yen1p , is consistent with GEN-1 being a junction-resolving enzyme in C . elegans . A Holliday junction-resolving activity is likely to be required for meiotic recombination , and a defect in this activity is predicted to result in embryonic lethality due to random autosome segregation in meiosis [55] . We can exclude such a defect as gen-1 ( tm2940 ) worms propagate as wild type , and fail to exhibit embryonic lethality in the absence of genotoxic stress ( Figure 3A ( 0 Gy ) ) . Furthermore , we did not observe an enhanced incidence of XO males , a phenotype that would indicate defects in meiotic chromosome pairing or recombination of the X chromosome ( Table 1 ) [56] . Most C . elegans mutations of DNA damage checkpoint genes , such as hpr-17 , encoding for the Replication Factor C homolog of S . pombe Rad17 , are also considered to be required for DNA DSB repair , as the corresponding mutants are hypersensitive to IR [39] . Two assays allow for testing the IR sensitivity of cells residing in different germ line compartments . In the “L1” IR survival assay that corresponds to the screening conditions we initially used to isolate yp30 as an IR sensitive mutant , the sensitivity of mitotic germ cells is evaluated by irradiating L1 larvae and by assaying for sterility of the resulting adults . The extent of sterility is scored by counting the number of worms in the following generation . Upon irradiation of L1 larvae , gen-1 ( yp30 ) and gen-1 ( tm2940 ) mutants were equally hypersensitive to IR , similar to hus-1 ( op244 ) , mrt-2 ( e2663 ) and hpr-17 ( tm1579 ) positive control strains ( Figure 3A ) . To assess whether GEN-1 might also be required to repair DNA damage induced by methyl methane sulonate ( MMS ) treatment , we tested for MMS sensitivity in a manner analogous to the assay for radiation . MMS leads to double-strand breaks when DNA replication forks encounter alkylated bases and mutants defective in recombinational repair are MMS sensitive [57] . We found that gen-1 ( tm2940 ) and gen-1 ( yp30 ) were MMS hypersensitive ( Figure 3C ) . In contrast to various control mutants with DNA repair defects , gen-1 mutants were not hypersensitive to UV irradiation , which causes lesions predominately repaired by excision repair ( Figure 3D ) . Neither DNA cross-linking by nitrogen mustard , which is largely repaired by the DNA interstrand cross link pathway , nor hydroxyurea which slows DNA polymerase processivity by nucleotide depletion , led to hypersensitivity in gen-1 mutants ( Figure 3E and 3F ) . To corroborate our results , we also employed the L4 irradiation assay [58] . In the “L4” IR assay , the sensitivity of meiotic pachytene cells is determined by measuring survival of embryos that are produced ∼20 hours after irradiation; these embryos are derived from pachytene cells that are arrested in the G2 cell cycle stage for more than 10 hours prior to completing meiosis and oogenesis . We found that both gen-1 ( tm2940 ) and gen-1 ( RNAi ) are as IR-sensitive as the hpr-17 ( tm1579 ) deletion , whereas gen-1 ( yp30 ) pachytene germ cells were not sensitive to IR ( Figure 3B ) . A similar response profile was found in response to MMS treatment ( Figure S5A ) , while no enhanced sensitivity was found in response to UV , nitrogen mustard , or hydroxyurea ( Figure S5B and S5D ) . Thus , the gen-1 ( yp30 ) allele , which results in a C-terminally truncated protein that retains nuclease activity in vitro , elicits IR- and MMS-induced hypersensitivity for mitotic germ cells of L1 larvae , whereas a null gen-1 mutation displays additional hypersensitivity to these agents in L4 germ cells arrested in pachytene . Our results suggest that the signaling function of GEN-1 is likely conferred by the C-terminus of GEN-1 , given that the gen-1 ( yp30 ) C-terminal truncation mutants as well as the gen-1 ( tm2940 ) deletion are defective in checkpoint signaling , whereas gen-1 ( yp30 ) , which retains nuclease activity that may directly promote DNA repair in mitotic germ cells . The differential sensitivity of the gen-1 ( yp30 ) allele in L1 and L4 survival assays likely reflects the fact that checkpoint-induced cell cycle arrest contributes to the survival of mitotic germ cells to IR . Furthermore , gen-1 ( yp30 ) is only partially defective for IR-induced germ cell apoptosis ( Figure 1C ) . To test if the IR sensitivity phenotypes of gen-1 mutants correlate with persistence of DSBs , we assayed for RAD-51 foci . At doses where multiple DSBs per cell are generated , the number of persistent RAD-51 foci in mitotic germ cells of mrt-2 ( e2663 ) and both gen-1 mutants is higher as compared to wild type , indicating a DSB repair defect ( Figure 4 ) . To directly confirm whether IR leads to increased DNA double-strand breakage in gen-1 mutant worms we directly assayed for chromosome fragmentation after irradiation with 90 Gy . As shown previously [59] , 48 hours after irradiation of mitotic germ cells ( at the L4 stage ) the diakinesis chromosomes of resulting mrt-2 ( e2663 ) oocytes were fragmented . In contrast , IR-induced damage was repaired in wild type , where oocyte chromosomes appear as 6 morphologically intact condensed DAPI stained structures ( Figure 5A and 5B ) . Chromosome fragmentation for both gen-1 mutants was as strong as that observed for the mrt-2 positive control , indicating a defect in DSB repair . This chromosome fragmentation phenotype was not observed as a consequence of irradiating pachytene stage cells and observing corresponding oocytes ∼8 hours and ∼20 hours after IR ( Figure 5C ) . Given that gen-1 ( tm2940 ) and gen-1 ( yp30 ) are equally defective in repairing diakinesis chromosomes 48 hours after irradiation we consider it likely that the checkpoint functions of gen-1 ( and mrt-2 ) in mitotic germ cells contribute to DSB repair . Given that DSBs inflicted in pachytene cells are repaired in gen-1 and mrt-2 mutants while this is not the case for DSBs in mitotic germ cells there might be a stronger requirement of GEN-1 and MRT-2 for DSB repair in mitotic germ cells . We next wished to determine if gen-1 acts in a known pathway promoting the repair of DSBs . We first examined if gen-1 affects non-homologous DNA end joining . In C . elegans DNA end joining is predominantly used in somatic cells . Worms defective in DNA end joining genes such as lig-4 , cku-70 and cku-80 show a reduced pace of development upon IR of embryos [59] . We found that neither of the gen-1 mutants exhibited any such somatic developmental delay ( Figure S7A ) . The strong IR-sensitivity and the defect in checkpoint-dependent cell cycle arrest and apoptosis of gen-1 ( tm2940 ) is reminiscent of the phenotype of mutations in upstream DNA damage signaling factors such as the C . elegans 9-1-1/Replication Factor C-like complex members hus-1 and mrt-2 ( S . pombe rad1 ) . Given that mutations of genes encoding for the 9-1-1 complex lead to telomere replication defects [60] , we asked if sterility in later generation worms caused by progressive telomere attrition occurs in gen-1 ( tm2940 ) . We failed to observe such an effect , further indicating that gen-1 is not part of the mrt-2 epistasis group ( Figure S8A ) . To investigate further how GEN-1 affects DNA damage responses , we depleted gen-1 in hus-1 or mrt-2 mutant backgrounds . RNAi depletion of gen-1 in hus-1 or mrt-2 mutant strains leads to synthetic lethality ( Figure 6B ) . We confirmed this synthetic lethality by gen-1 hpr-17 double mutant analysis ( Figure S8B ) . hpr-17 encodes for the 9-1-1 clamp loader and is part of the mrt-2 epistasis group . As expected , gen-1 RNAi in a mrt-2 ( e2663 ) background led to an increased number of RAD-51 foci as compared to gen-1 RNAi in wild type worms and to the mrt-2 ( e2663 ) mutant in mitotic germ cells ( Figure S9 ) . In contrast , gen-1 ( yp30 ) , which is checkpoint-defective but encodes a protein that can promote HJ resolution in vitro , did not cause synthetic lethality when combined with an hpr-17 mutation , nor did it exacerbate the radiation hypersensitivity phenotype of hpr-17 ( Figure S8C ) . These results therefore suggest that the DNA repair function of GEN-1 may act redundantly with the 9-1-1 complex to repair DSBs occurring during normal DNA replication . We next wished to determine genetic interactions between GEN-1 , ATR and ATM PI3-like kinases , which are predicted to act upstream of the 9-1-1 complex in DNA damage signaling . Given that an atl-1 ( tm853 ) deletion leads to excessive genome instability in germ cells and concomitant sterility [61] , we could not assess the possibility of enhanced IR sensitivity in gen-1 atl-1 double mutants . In contrast , C . elegans atm-1 plays a minor role in DNA damage signaling and atm-1 ( gk186 ) results in partial defect in IR-induced cell cycle arrest and apoptosis [62] . Consistent with this notion , we found that the atm-1 ( gk186 ) deletion is not hypersensitive to IR when subjected to the L4 IR survival assay , and that the IR sensitivity is not enhanced by the gen-1 ( yp30 ) mutant ( Figure S10 ) . In contrast , the atm-1 ( gk186 ) mutant is sensitive to IR in the L1 assay , and IR sensitivity is enhanced in combination with both gen-1 ( tm2940 ) and gen-1 ( yp30 ) ( Figure 7 ) . In summary , our results suggest that gen-1 might act in parallel to atm-1 for repairing mitotic germ cells affected by DNA double-stranded breaks . Given that gen-1 encodes for a nuclease , we wanted to eliminate the possibility that GEN-1 might also be required for the processing of DSBs to generate single-stranded DNA overhangs , which would be coated by RPA1 and lead to the ATRIP-dependent activation of ATR in mammalian cells [2] . We thus tested whether IR-dependent RPA-1 loading is compromised in gen-1 ( tm2940 ) worms . We found that the sequential accumulation of RPA-1 ( green ) and RAD-51 ( red ) foci does not significantly differ between wild type and gen-1 ( tm2940 ) worms , indicating that the initial steps of DSB processing occur normally in gen-1 mutants ( Figure S6A ) . These results are corroborated by our finding that GEN-1 does not cleave double-stranded substrates or substrates with 3′ single stranded overhangs in vitro ( Figure 2G ) . To monitor the activation of the C . elegans ATL-1/ATR-mediated DNA damage checkpoint pathway in gen-1 ( tm2940 ) mutants , we analyzed the IR-induced transcriptional induction of the pro-apoptotic BH3-only domain encoding genes ced-13 and egl-1 . The induction depends on the C . elegans CEP-1 p53-like transcription factor [63] , and on upstream DNA damage response genes including atl-1 ( C . elegans ATR ) , clk-2 , hus-1 and mrt-2 [64] . egl-1 and ced-13 were induced to near-normal levels for tm2940 and yp30 alleles of gen-1 , while no induction occurred in a cep-1 ( lg12501 ) background ( Figure 6A ) . Thus , the apoptotic signaling function of GEN-1 acts in parallel to the canonical C . elegans DNA damage response pathway necessary for egl-1 induction . To further support this notion , we cytologically probed for the activation of CHK-1 , which is required for IR-induced cell cycle arrest and apoptosis in C . elegans [65] . To this end we employed an antibody against a conserved CHK-1 phosphopeptide that includes serine 345 [36] , [66] . Phosphorylation of this residue in response to DNA damage depends on ATR and ATM kinases and leads to Chk1 activation in mammals [67]–[69] and occurs in response to ATL-1/ATR activation in C . elegans [36] , [66] . CHK-1 phosphorylation is increased in response to IR in cell cycle arrested cells ( Figure 6C , top panel ) , both in wild type as well as in gen-1 mutants , further substantiating the notion that the checkpoint signaling function of GEN-1 might act in parallel to the canonical pathway . Interestingly , CHK-1 phosphorylation also occurs in the mrt-2 ( e2663 ) mutant ( Figure 6C ) . This data indicates that ATM/ATR is not fully dependent on mrt-2 , consistent with the reduction as opposed to the complete alleviation of CEP-1 dependent transcription in this mutant [64] . Thus our results suggest that gen-1 and mrt-2 act in parallel pathways needed for checkpoint signaling similar to their roles in DSB repair .
We have discovered that the deficiency of GEN-1 results in DNA damage signaling defects ( Figure 8 ) . Neither cell cycle arrest of mitotic germ cells , nor apoptosis induction of meiotic pachytene cells occurs in response to DNA damage in gen-1 mutants . These defects are as severe as those observed in known C . elegans checkpoint mutants such atl-1 , the worm ATR homolog [61] , clk-2 [70] and mutants affecting components of the C . elegans 9-1-1 complex [60] , [71] , [72] . Intriguingly , we find that the apoptosis defect conferred by a mutation in gen-1 does not result from the ATR- , CLK-2- and 9-1-1 complex-dependent activation of the primordial worm p53-like protein CEP-1 ( Figure 8 ) [41] , [44] . The two known CEP-1 target genes egl-1 and ced-13 , whose transcriptional activation confers the inhibition of the anti-apoptotic Bcl2 like protein CED-9 , are normally induced [63] . Thus , the signaling function of GEN-1 , which promotes apoptosis in meiotic germ cells , appears to be in a pathway acting in parallel or downstream of the canonical DNA damage response pathway that activates CEP-1/p53 . Analogous results have been observed for C . elegans sir-2 . 1 histone deacetylase , as well as for hyl-1 and lagr-1 ceramide synthase mutants , where CEP-1 targets are upregulated in response to DNA damage even though germ cell apoptosis fails to occur [73] , [74] . Thus , gen-1 , sir-2 . 1 , hyl-1 and lagr-1 may define components of a DNA damage response pathway that functions in parallel to the pathway , which needs CHK-1 and the 9-1-1 complex . How these pathways are integrated remains to be elucidated . We speculate that GEN-1 may facilitate DSB repair by coordinating cell cycle progression with HJ resolution . Our evidence that C . elegans GEN-1 acts as a HJ-resolving enzyme is supported by the biochemical characterization of its human and yeast orthologs [33] . Given the orthologous relationship with C . elegans GEN-1 and our biochemical evidence , it is likely that C . elegans GEN-1 can act as a HJ-resolving enzyme in vivo . The active form of GEN-1 purified form HeLa cell extracts is a C-terminal truncation [33] . We analyzed multiple preparations of full length and truncated versions of C . elegans GEN-1 but could only obtain weak nuclease activity on mobile HJ substrates . Nevertheless , this activity is lost if one of the putative active site residues was mutated , and it was specific for Holliday junction substrates . Thus , the nuclease function of C . elegans GEN-1 during DSB repair may involve HJ resolution . Future studies may refine our understanding of the substrate specificity of GEN-1 . Our results point towards the possibility that the completion of HJ resolution in response to DNA damage-induced DSBs might be monitored by GEN-1 , which might act both as a Holliday junction-resolving enzyme as well as a DNA damage signaling molecule ( Figure 8 ) . A dual function enzyme that catalyses a late step of recombination and plays a role in checkpoint signaling could provide a mechanism to suppress cell cycle progression to allow for the repair of the majority of DNA double-strand breaks before cell cycle progression resumes . The signaling function of GEN-1 is likely conferred by the C-terminus of GEN-1 . The C-terminus of GEN-1 may therefore interact with known or novel DNA damage signaling molecules that function to promote DSB repair in mitotic germ cells . At the moment we can only speculate about the nature of the RAD-51 foci that persist in gen-1 mutants . Some of these RAD-51 foci might correspond to recombination intermediates resulting from failure of specific types of checkpoint-mediated DNA repair . Alternatively , these foci might be the consequence of initial unrepaired DNA damage that result in double-strand breakage once unrepaired DNA is replicated when cells resume cell division . Given the specificity with which GEN-1 processes HJ structures in vitro , it is surprising that GEN-1 does not have any obvious function in meiotic recombination . One candidate for a C . elegans meiotic HJ-resolving enzyme might be the Him-18/SLX4/Mus312 SLX1 nuclease complex . The rate of meiotic recombination is significantly reduced in Drosophila mus312 mutants [75] , and the human SLX1/SLX4 complex has recently been shown to have HJ resolution activity in vitro [28]–[31] . Further , lack of a role for GEN-1 in meiotic crossover resolution is consistent with recent evidence that Drosophila and C . elegans him-18/slx-4 may promote meiotic Holliday junction resolution [31] , [32] . Additional proteins implicated in resolving meiotic HJ initially in fission yeast and fruit flies are Mus81 and Xpf1 , respectively [21] , [75] . Further , the combined activities of Bloom's helicase and topoisomerase III have been shown to dissolve HJ independently of canonical junction-resolving activities in vitro [13] , [14] . However , the meiotic defects of the C . elegans mus-81 , xpf-1 or him-6 Bloom's orthologs are not overtly enhanced by the gen-1 ( tm2940 ) mutation ( Simon Boulton , personal communication ) . Collectively , the absence of a meiotic defect of gen-1 together with the lack of strong synthetic effects with candidate meiotic HJ resolving enzymes , strongly suggests that C . elegans GEN-1 does not play a central role in this process . Although different species vary in their precise DNA double-strand break response strategies , and various cell types are likely to utilize different DSB repair pathways preferentially , basic regulatory complexes and processes tend to be conserved . C . elegans GEN-1 plays an essential role in responding to DSBs , but it is inert in budding yeast [34] and has apparently been lost during evolution of fission yeasts . In addition to DNA end-joining , which does not require HJ resolution , DSBs can be repaired without a HJ resolution step by DNA synthesis-dependent strand annealing [57] , [76] , [77] . We speculate that this may be related to an inherent redundancy in DNA double-strand break repair pathways in diverse organisms , and perhaps within various tissues of the same organism . Indeed , our staining for RAD-51 foci indicates that most DSBs are repaired in gen-1 mutants , likely by a combination of the above mentioned recombinational repair pathways and non-homologous end joining , but that a fraction of these breaks persists 48 hours after IR ( Figure 4 ) . Such a scenario is in line with recent data suggesting that a subset of persistent DBSs is repaired by distinct DSB repair pathways [59] , [78] . In mammals these persistent foci are associated with heterochromatin and their repair specifically requires ATM [79] , which may be consistent with the enhanced DNA damage response defects observed for atm-1;gen-1 double mutants . Thus , GEN-1 might be involved in DSB repair processes that are redundant and therefore hidden within DSB response networks in some organisms . It has recently been reported that GEN1 is absent in ovarian and colon cancer cell lines , suggesting that GEN1 is required for maintaining genome stability in human cells [80] . Thus , GEN1 might join the number of genes involved in recombinational repair such as BRCA1 , BRCA2 , and FANCJ/BACH , mutation of which is associated with cancer . Deletion of these genes does not result in cellular lethality , but affected cancer cells are uniquely sensitive towards DNA damaging agents allowing their selective eradication . Redundant mechanisms involved in resolving HJ structures might be particularly amenable to such synthetic lethal approaches . Our finding that gen-1 is synthetically lethal with mutations in known DNA damage sensors and repair proteins encoded by the 9-1-1 complex suggests one such mechanism . Overall , our results show that GEN-1 , a protein previously implicated in HJ resolution , possesses dual function that potentially couples DNA repair and DNA damage signaling .
Worms were maintained at 20°C on NGM agar plates seeded with E . coli strain OP50 as previously described [81] , unless otherwise indicated . Alleles are all described in the CGC C . elegans stock center . We generated the following strains as part of this study TG1043 gen-1 ( yp30 ) III; TG1540 gen-1 ( tm2940 ) III; TG765 cep-1 ( lg12501 ) II; TG1236 gen-1 ( yp30 ) unc-32 ( e189 ) III; TG1237 gen-1 ( yp30 ) dpy-17 ( e164 ) III; TG1233 hpr-17 ( tm1579 ) II; TG771 hus-1 ( op244 ) I; TG545 hus-1 ( op241 ) I; TG1503 hpr-17 ( tm1579 ) II: gen-1 ( tm2940 ) III; TG1502 gen-1 ( yp30 ) III , opIs76 ( CYB-1::YFP ) ; TG1064 gen-1 ( yp42 ) III; TG1060 gen-1 ( yp45 ) III; TG1565 xpg-1 ( tm1670 ) I; RB964 cku-80 ( ok861 ) III; TG190 clk-2 ( mn159 ) III; VC381 atm-1 ( gk186 ) I . DNA damage-induced apoptosis and L4 radiation hypersensitivity ( rad ) assays were performed as described [39] . For γ-irradiation a Cs137 source ( 2 . 9 Gy/min , IBL 437C , CIS Bio International ) was used . 6x-histidine tagged full length GEN-1 ( pGA343 ) was expressed in BL21 ( DE3 ) CodonPlus cells , recovered from inclusion bodies using BugBuster ( Novagen ) , solubilised in Urea buffer and purified with Ni-NTA following manufacturer's instructions ( Qiagen ) . One guinea pig was immunized ( BioGenes GmbH ) and antibodies were affinity-purified from the final bleeding using Maltose Binding Protein ( MBP ) tagged protein . N- and C-terminus GEN-1 ( fragments 1-136 and 356-434 respectively ) tagged with MBP ( pGA346 and pGA348 ) were purified using an amylose resin column ( New England BioLabs ) . For affinity purification , proteins were covalently linked to AffiGel 15 ( Bio-Rad ) . Immunostaining experiments were performed as described [74] . Primary antibodies used were rabbit anti-RAD-51 ( 1/200 dilution ) as described [82] , rat anti-RPA-1 ( 1/100 dilution ) , rabbit anti-Cdk1 ( pTyr15 ) Calbiochem , 219440 ( 1/50 dilution ) and rabbit anti-P-CHK1 ( P-CHK1 Ser 345: sc-17922 , Santa Cruz , 1/50 dilution ) . Secondary antibodies used were Cy3 labeled anti-rabbit ( Jackson Immunochemicals ) 1/1000 dilution and FITC anti-rat ( Jackson Immunochemicals ) 1/200 dilution . C . elegans full-length ORF of rpa-1 was cloned into pMAL-2c vector ( New England Biolabs ) and expressed in BL21 ( DE3 ) cells . The MBP-tagged protein was purified on an amylose resin following the manufacturer's instructions ( New England Biolabs ) and used to immunize one rat ( Eurogentec animal SAOI . 1 ) . The same purified protein was covalently linked to AffiGel-15 ( BioRad ) and used to affinity-purify antibodies from the final bleed . L1 larvae stage worms were sorted from a growing population using an 11 µm filter ( Millipore NY11 ) and treated with the indicated genotoxic agents . To test MMS and Nitrogen mustard sensitivity , worms were incubated with the indicated concentration of mutagen for 12 hours in M9 buffer . UV irradiation was performed by the XL-1000 Spectrolinker UV-C light source . 5 L1 stage worms in the P0 generation were plated onto a single plate . The number of living worms ( post the L1 stage ) present in the F1 generation within 48 hours of ( untreated ) P0 worms reaching the L4 stage was counted using a dissection microscope . For hydroxyurea ( HU ) , L1 worms were plated on 1x NGM plates supplemented with the indicated compounds and the living adult worms corresponding to the F1 generation were established similarly . Experiments were done at least in triplicate . RNAi feeding was done as described with exception of using 1 mM IPTG [58] . Recombinant proteins ( 50 nM ) were added to 2 nM ( Figure S4 ) or 5 nM ( Figure 2E–2G ) of the indicated substrates all of which were 5′ [32P]-labelled on one strand in 10 mM Tris-HCl pH 8 . 0 , 10 mM NaCl , 10 mM MgCl2 , 0 . 1 mg/ml BSA , 0 . 1 mg/ml calf thymus DNA , 1 mM DTT and 1 M NDSB 201 . Samples were incubated at 37°C for 30 min or overnight , and the reaction terminated by addition of EDTA . Cleavage products were analysed by electrophoresis in 12% polyacrylamide gels containing 8 M urea . Gels were dried and imaged using a Fuji BAS 1500 phosphorimager . For the nuclease assays , substrates were generated as described [33] , the following oligonucleotides were used , sequences are shown 5′ to 3′: Single-stranded: -1971: CGCTCTAGAGCGGCTTAGGCTTAGGCTTAGGCTTA Double-stranded ( annealing 1971 and 1972 ) : -1972: TAAGCCTAAGCCTAAGCCTAAGCCGCTCTAGAGCG 5′ Flap ( annealing A-Flap , B and C ) 3′ Flap ( annealing B-Flap , A and C ) -A-Flap: ATGTGGAAAATCTCTAGCAGGCTGCAGGTCGAC -B-Flap: CAGCAACGCAAGCTTGATGTGGAAAATCTCTAGCA -A: GGCTGCAGGTCGAC -B: CAGCAACGCAAGCTTG -C: GTCGACCTGCAGCCCAAGCTTGCGTTGCTG | Coordination of DNA repair with cell cycle progression and apoptosis is a central task of the DNA damage response machinery . A key intermediate of recombinational repair and meiotic recombination , first proposed in 1964 , involves four-stranded DNA structures . These intermediates have to be resolved upon completion of DNA repair to allow for proper chromosome segregation . Using forward genetics , we identified a Caenorhabditis elegans dual function DNA double-strand break repair and DNA damage signaling protein orthologous to the human GEN1 Holliday junction resolving enzyme . GEN-1 facilitates repair of DNA double-strand breaks , but is not essential for DNA double-strand break repair during meiotic recombination . The DNA damage signaling function of GEN-1 is separable from its role in DNA repair . Unexpectedly , GEN-1 defines a DNA damage-signaling pathway that acts in parallel to the canonical pathway mediated by CHK-1 phosphorylation and CEP-1/p53 . Thus , an enzyme that can resolve Holliday junctions may directly couple a late step in DNA repair to a pathway that regulates cell cycle progression in response to DNA damage . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"molecular",
"biology/recombination",
"genetics",
"and",
"genomics/gene",
"discovery",
"molecular",
"biology/dna",
"repair"
] | 2010 | The Caenorhabditis elegans Homolog of Gen1/Yen1 Resolvases Links DNA Damage Signaling to DNA Double-Strand Break Repair |
The effectiveness of rabies vaccination in both humans and animals is determined by the presence of virus neutralizing antibodies ( VNAs ) . The Rapid Fluorescent Focus Inhibition Test ( RFFIT ) is the method traditionally used for detection and quantification of VNAs . It is a functional in vitro test for assessing the ability of antibodies in serum to bind and prevent infection of cultured cells with rabies virus ( RABV ) . The RFFIT is a labor intensive , low throughput and semi-quantitative assay performed by trained laboratorians . It requires staining of RABV-infected cells by rabies specific fluorescent antibodies and manual quantification of fluorescent fields for titer determination . Although the quantification of fluorescent fields observed in each sample is recorded , the corresponding images are not stored or captured to be used for future analysis . To circumvent several of these disadvantages , we have developed an alternative , automated high throughput neutralization test ( HTNT ) for determination of rabies VNAs based on green fluorescent protein ( GFP ) expression by a recombinant RABV and compared with the RFFIT . The HTNT assay utilizes the recombinant RABV ERA variant expressing GFP with a nuclear localization signal ( NLS ) for efficient quantification . The HTNT is a quantitative method where the number of RABV-infected cells are determined and the images are stored for future analysis . Both RFFIT and HTNT results correlated 100% for a panel of human and animal positive and negative rabies serum samples . Although , the VNA titer values are generally agreeable , HTNT titers tend to be lower than that of RFFIT , probably due to the differences in quantification methods . Our data demonstrates the potential for HTNT assays in determination of rabies VNA titers .
Human rabies is a zoonotic disease transmitted predominantly through bites from infected animals [1] . Although , rabies is nearly 100% fatal after onset of symptoms , it is preventable by post exposure prophylaxis ( PEP ) when administered immediately and appropriately after a suspect exposure . The surrogate for protection against rabies virus ( RABV ) infection is the presence of virus neutralizing antibodies ( VNAs ) targeted against the RABV glycoprotein [2 , 3] . VNAs play important roles in preventing the invasion of RABV into peripheral nerves at the site of exposure and subsequent transport to the brain [4] . Rabies vaccination confers complete protection against the disease and can be administered either pre- or post- exposure [4] . The advisory committee on immunization practices ( ACIP ) currently recommends as pre-exposure ( Pr-E ) , a three-dose regimen administered at days 0 , 7 and 21 or 28 [5] . For PEP , ACIP recommends administration of purified rabies immunoglobulin ( RIG ) at the wound followed by four doses of vaccination at days 0 , 3 , 7 and 14 after exposure [6] . While the prophylactically administered antibodies offer immediate protection , acquired immunity followed by vaccination provides long-term memory response . The Pr-E vaccination is intended for individuals who are at high risk for RABV infection , such as laboratory personnel and animal care providers like veterinarians , wildlife rehabilitators and animal control workers [7] . In addition to Pr-E vaccination , monitoring rabies VNA titers against the virus periodically is required to determine the level of immunity against RABV [8] . Nearly 99% of the estimated annual 59 , 000 human deaths worldwide are caused by dog bites [1] . Rabies control could be achieved by vaccinating 70% of dog population , particularly free roaming dogs , in order to break the RABV infection cycle and end the circulation of virus [9 , 10] . In the United States ( U . S . ) and several other countries , circulation of canine ( dog ) RABV variant has been eliminated by comprehensive dog rabies vaccinations [11 , 12] . The importation of unvaccinated dogs to the U . S . , including a case of a RABV-infected dog by potential falsification of vaccination records , have been reported [13 , 14] . To avoid re-introduction of canine RABV variants in previously eliminated regions , the World Organization of Animal Health ( OIE ) has enforced strict guidelines for importation of pets and other domestic animals [11 , 15 , 16] . Many countries require rabies vaccination records and demonstration of rabies VNA levels in pets before travel by either rapid fluorescent focus inhibition test ( RFFIT ) or fluorescent antibody virus neutralization ( FAVN ) tests or an extended period of quarantine in the absence of titer information [14] . Developed during the 1970s , RFFIT replaced the mouse neutralization test ( MNT ) , which required demonstration of VNAs to protect infection in vivo . Due to the requirement of large numbers of mice , ethical considerations and long duration for MNT , an alternative in vitro test ( RFFIT ) was developed [17] . The RFFIT is a semi-quantitative method in which 20 microscopic fields are observed for the presence of fluorescent foci ( RABV-infected cells ) to determine rabies VNA titer . Similarly , FAVN test developed in 1998 , utilizes a modified protocol for rabies VNA titer determination and has demonstrated similar results compared to RFFIT and MNT [18] . According to ACIP , complete neutralization of RABV at the 1:5 serum dilution , which corresponds roughly to 0 . 1 International Unit ( IU ) / ml is a prerequisite for rabies protective titer in humans [5] . A minimum titer for 0 . 5 IU/ml is required as a proxy for protection according to World Health Organization ( WHO ) requirements [19] . For animals , rabies neutralization titer should be 0 . 5 IU/ml or higher as per OIE guidelines [15] . The RFFIT is a labor intensive , low throughput assay requiring skilled personnel to perform , interpret and quantify results . While the number of fluorescent foci are recorded , the fluorescent images viewed in a microscope are not stored and hence cannot be re-analyzed . Because RFFIT involves observation of 20 fields ( 40% ) and not the entire well , the choice of fields may vary with testing personnel . Considering these drawbacks , we intended to develop a high throughput and quantitative test with the ability to store and analyze the results long-term . The high throughput neutralization test ( HTNT ) described in the present study utilizes a recombinant RABV ERA variant that expresses green fluorescent protein ( GFP ) . GFP reporter viruses are used in neutralization studies for several viruses , including RABV , wherein for quantification of virus infected cells are based on GFP expression from viral genome instead of staining the cells with antibodies against viral proteins [20 , 21] . However , in order to improve automated detection and quantification in the HTNT , a nuclear localization signal ( NLS ) was added to the N-terminus of GFP [22 , 23] . The overall procedure for both RFFIT and HTNT are very similar , however in HTNT the infected cells are quantified based on the ratio of GFP positive nuclei to total nuclei , to determine percent neutralization . The high content screening ( HCS ) instrument , ArrayScan used in this study is an automated stage microscope which takes multiple images of an entire 96- or 384-cell plates to capture DAPI and GFP staining , respectively . The instrument also stores the images for future analysis . The results obtained by HTNT assay correlated 100% with RFFIT results for detection of rabies VNAs in serum samples . The rabies VNAs titers determined by HTNT based on the quantification of infected cells agreed with RFFIT values , although HTNT titers tended to be lower than that of RFFIT due to differences in quantification methods . Overall , this study demonstrates the utility of a GFP reporter-based assay for detection of rabies VNAs in an automated high throughput system .
The mouse neuroblastoma ( MNA ) and BSR ( a clone of baby hamster kidney 21 ) cells ( Centers for Disease Control and Prevention [CDC] collection ) were cultured in E-MEM media supplemented with 10% fetal bovine serum ( FBS ) , L-glutamine , essential vitamins , antibiotics ( Penicillin and Streptomycin ) and antimycotic ( Amphotericin B ) . The RABV variant challenge virus standard -11 ( CVS-11 ) was propagated in BSR cells [24] . The codon optimized Monster Green Fluorescent Protein ( hMGFP ) gene , which offers greater fluorescent intensity and lower cytotoxicity from Montastrea cavernosa , was obtained from Promega Corporation . The hMGFP gene was modified at the N-terminus by addition of NLS ( NLS-hMGFP ) for nuclear targeting and localization as previously described [24] . The NLS-hMGFP open reading frame was incorporated between the phospho- ( P ) and the matrix- ( M ) protein genes in RABV ERA genome ( Fig 1A ) . After cloning and sequence verification , the full-length recombinant viral genomic cDNA was applied for virus recovery . In brief , the BSR cells grown in a six-well plate at ~ 90% confluency were transfected using a combination of 6 plasmids: a full-length viral genomic cDNA plasmid pERA-NLS-hMGFP at 3 . 0 μg/well , and five helper plasmids of pTN at 1 . 0 μg/well , pMP at 0 . 5 μg/well , pML at 0 . 5 μg/well , pMG at 0 . 5 μg/well ( the plasmids expresses RABV encoded proteins N , P , L and G in trans ) and pNLS T7 at 1 . 0 μg/well . Seven to 10 days after transfection , the recombinant ERA-NLS-hMGFP virus was recovered and further amplified in fresh BSR cells until virus titer reached at least 107 Focus Forming Unit ( ffu ) /ml . BSR cells were seeded overnight on 12-well glass bottom tissue culture plates . The cells were infected with recombinant RABV ERA-NLS-hMGFP virus for 24 h at 37°C , fixed with 4% paraformaldehyde after infection and blocked with 10% FBS in 1X PBS for 15 minutes at room temperature ( RT ) . Mouse monoclonal antibodies against N protein ( anti-N mAb ) were diluted in 1X PBS containing 10% FBS and incubated for 30 minutes at RT , followed by three washes with 1X PBS . Alexa Flour 594 conjugated anti-mouse ( Molecular Probes ) in 1X PBS was used as secondary antibodies and incubated for 30 minutes at RT . Cell nuclei were stained with DAPI ( 4’ , 6-diamidino-2-phenylindole ) and mounted with Prolong Antifade Mounting Media ( Thermo Fisher Scientific ) . The cells were visualized using LSM 710 inverted confocal microscope ( Zeiss ) for DAPI , GFP and Alexa Flour 594 using respective filters . Human sera used as negative controls were previously confirmed by RFFIT for absence of rabies VNAs . These historic patient specimens received at the CDC Rabies Laboratory for human antemortem rabies diagnostic testing were de-identified according to an IRB approved CDC protocol #7028 . Human sera positive for rabies VNAs , previously collected from rabies vaccinees as part of Occupational Health Clinic screens were de-identified and used for validation in the HTNT and RFFIT assays . Animal sera , either from control ( unvaccinated ) or vaccinated animals , primarily dogs and cats were provided by Nancy Laboratory for rabies and wildlife , ANSES , France . The RFFIT [17] is utilized to measure the level of rabies VNAs against the RABV CVS-11 in human and animal serum samples . Eight 5-fold serial dilutions of heat-inactivated serum samples were incubated with RABV in 8-well tissue culture chamber slides for 90 min at 37°C . The titer of RABV CVS-11 virus used is 50 FFD50/0 . 1 mL . 200μL of MNA cells ( 5 X 105 cells/ml ) were then added to every well containing the serum-virus mixture , which is comprised of 50μL of serum and 100μL of the CVS-11 virus and incubated for an additional 20 hours at 37°C with 0 . 5% CO2 . Slides were then washed , fixed with acetone and stained with anti-rabies FITC ( fluorescein isothiocyanate ) immunoglobulin ( Fujirebio Diagnostics , Inc ) containing Evans blue ( 0 . 5% in PBS ) . Evans blue is a counterstain that provides a red background fluorescence to improve contrast . Twenty distinct microscopic fields per well were examined using a fluorescence microscope at X200 magnification to score the virus-infected cells ( foci ) . From the number of positive fields per well , the rabies VNA titers are mathematically calculated using the Reed-Muench calculation [25] . The endpoint neutralization titer of the test serum is converted to international units ( IU ) /ml values by calibration against the endpoint neutralization titer of the U . S . Standard Rabies Immune Globulin ( SRIG ) ( obtained from Food and Drug Administration , U . S . ) , which was measured in the same assay at 2 . 0 IU/ml . HTNT to measure rabies VNAs in human and animal sera was optimized with theRFFIT protocol as a starting point . Starting at a 1:2 . 5 dilution , eight 5-fold serial dilutions of sera were made in 96-well plates . Specifically , 40μl of serum was added to 60μl of E-MEM media and 20μl of this mixture was transferred into 80μl of media to complete the 5-fold dilutions . ERA-NLS-hMGFP ( 50 FFD50/0 . 1ml ) was pre-diluted 1:25 in E-MEM media ( 80μl ) , at a multiplicity of infection sufficient to infect 50–70% of cells , was added to the diluted sera ( 80μl ) and mixed . Media and virus only controls were included on each plate and an SRIG positive control was included at least once per run . Samples were incubated for 90 min at 37°C , followed by addition of BSR cells at a concentration of 3 . 5x105 cells/ml in equal volume to the virus and serum mixture and moved to a black plate with transparent wells ( Costar ) . All the samples and controls were run in duplicate to obtain average values . The plates were incubated for 20 h at 37°C , fixed with 4% paraformaldehyde for 15 min at RT , washed twice with PBS and incubated with 3 μM DAPI ( nuclear stain ) for 5–15 min . After three additional washes , PBS was added to each well , plates were sealed and stored at 4°C until reading on the ArrayScan reader . The ArrayScan XTI HCS reader is an automated fluorescent microscope that acquires and records images in up to six separate fluorescent channels ( ThermoFisher Scientific ) . The instrument takes multiple images starting at the center in a spiral fashion to cover entire surface of a well . The accompanying Cell Analysis Software records the size and fluorescent intensity of the imaged objects and further analysis is performed using specified criteria . In the protocol described here , the channels blue and green were used to identify DAPI-stained nuclei ( blue ) and ERA-NLS-hMGFP infected cells ( green , GFP-expressing cells ) respectively [23] . Rabies VNA titers are mathematically calculated using the Reed-Muench formula [25] . The serum end-point titer in the neutralization assay is described as the highest dilution factor with 50 percent reduction in the number of the fluorescent foci observed . The rabies VNA titers are determined using the method of Reed-Muench that calculate the difference between the logarithm of the starting dilution and the logarithm of the 50% end-point dilution ( difference of logarithms ) from the formula: [50%− ( infectivitynextbelow50%]/[ ( infectivitynextabove50% ) – ( infectivitynextbelow50% ) ]Xlogarithmofdilutionfactor The RFFIT results can be expressed as a serum titer or in international units ( IU ) . For the calculation of IU/ml , the 50% end point titer of the reference serum ( diluted to 2 IU/ml ) and that of test serum is used in the following formula: NumberofIU/ml= ( End-pointtiterofthetestserum/End-pointtiterofthereference ) X2IU/mlinthereferenceserum Total number of DAPI-stained cells and GFP-positive cells were counted using the HCS Studio Cell Analysis software . Infected cells were determined by establishing a fluorescence intensity threshold where cells with a total GFP intensity above the threshold ( determined from the instrument ) were considered GFP-positive responders . The percentage of GFP-positive responders per well was recorded and further analyzed using Microsoft Excel 2013 . Relative GFP-positive responder ( RPR ) values were then calculated by using the cell and virus only controls: Relative%responders=[ ( Sample%responders ) – ( Cellonly%responders ) ]/[ ( Virusonly%responders ) – ( Cellonly%responders ) ] The RPR were calculated for each serum dilution and the duplicates averaged . The RPRs below and above 50% were then used in the Reed-Muench method to calculate the 50% endpoint titer , in which RPR is used instead of percentage of fields used by RFFIT . Endpoint titers were converted to IU/ml as described above . The results of HTNT are compared to the traditional RFFIT , to determine sensitivity and specificity measurements using RFFIT results as the true positives and true negatives . Sensitivity was calculated by the following formula [true positive/ ( true positive + false negative ) ] . Specificity was calculated using the following formula [true negative/ ( false positive + true negative ) ] . The 95% confidence intervals were calculated using the Clopper and Pearson method . Quantitative analyses were performed to compare the titer values between the two methods . Specifically , the differences between the RFFIT and HTNT IU/ml values for positive samples were analyzed using Bland-Altman plots [26] . Bland-Altman plots are constructed by plotting the differences between the IU/ml of the two methods against the IU/ml averages of both methods . The mean difference , or bias , and limits of agreement ( +/- 1 . 96SD ) are also plotted and utilized to evaluate the systemic differences between the two methods . The Pearson coefficient and concordance correlation coefficient ( Lin’s coefficient ) were used to measure the agreement of titers between methods . All analyses were performed using Microsoft Excel , GraphPad 6 . 0 , and R 3 . 4 . 1 .
The recombinant RABV ERA virus expressing hMGFP was generated by reverse genetics as described in methods ( Fig 1A ) . The recombinant virus replicated similarly to wild type RABV in cell culture . To check the expression and localization of hMGFP , BSR cells were infected with RABV ERA-NLS-hMGFP for 24 h at 37°C , fixed and processed for confocal microscopy . The expression of nucleoprotein ( N protein ) from the RABV genome was monitored by staining with mouse anti-N monoclonal antibody ( mAb ) followed by secondary staining with goat anti-mouse IgG—Alexa Flor 594 conjugate . Fig 1B demonstrates the expression of hMGFP in infected cells , based on N protein staining . The localization of hMGFP was exclusively in the nucleus , overlapping with the nuclear stain DAPI designed to enhance fluorescent signal intensity and improve quantification . RABV ERA-NLS-hMGFP was evaluated for rabies neutralization assay by HTNT as described in methods . The steps involved in HTNT are compared and contrasted with RFFIT procedures in Table 1 . In the HTNT assay , the number of DAPI-stained nuclei denotes the total cell count , while GFP positive nuclei that co-localize with DAPI represented infected cells . Fig 2A shows representative HCS data . Around 8 , 000 cells out of 14 , 000 total cells were positive for GFP staining in virus only positive control condition demonstrating close to 50% infection . None or negligible GFP positive nuclei were observed in cell only negative control demonstrating specificity of GFP detection and quantification . The GFP expression and co-localization with DAPI observed in the virus only condition was specific to expression from RABV-infected cells , as evidenced by the complete inhibition when incubated with SRIG antibody at 1:5 and 1:25 dilutions . HTNT using the ArrayScan also detects concentration dependence of RABV VNAs in sera as illustrated by the increased infectivity of cells at higher serum dilutions . From the number of GFP positive cells in test vs controls , percent RABV infection was obtained for titer determination ( Fig 2B ) . The human sera samples were run by both HTNT and RFFIT and considered positive based on complete neutralization of RABV infection at the 1:5 dilution as recommended by the ACIP guidelines . Using the RFFIT results to categorize the samples as true positive and true negative , 100% sensitivity and specificity was obtained with HTNT ( Fig 3 ) . Of 135 human serum samples , 74 negative and 61 positive results were consistent between both assays . Similarly , for the 42 animal samples both RFFIT and HTNT had identical results ( Fig 4 ) . HTNT exhibited 100% sensitivity and specificity compared to RFFIT in the ability to completely neutralize RABV in 1:5 dilution for animal samples tested ( Fig 4 ) . As positive and negative determinations were consistent in both assays , we next compared the rabies VNA titer values of positive sera obtained from RFFIT and HTNT . The IU/ml titer values determined by RFFIT and HTNT for human sera were plotted and the Pearson coefficient was calculated ( Fig 5A ) . Statistically significant coefficient value ( r = 0 . 88 ) was observed suggesting a strong linear relationship between the RFFIT and HTNT titer values . Because the Pearson correlation coefficient only measures the linearity of the relationship but not the agreement between the two methods , we also calculated the concordance correlation coefficient ( CCC , Fig 5B ) . The CCC was moderately strong ( 0 . 77 ) , indicating that the HTNT titers replicate RFFIT titers but the IU/ml values do not match entirely . To investigate the differences between the two measurements obtained from RFFIT and HTNT analyses , we used Bland-Altman plots . In the Bland-Altman plots , the differences of the titers obtained from the two methods were plotted against the average titers ( Fig 5C ) to determine bias ( measure of the systemic difference between the two methods ) and limits of agreement . The bias value 14 . 78 indicates that the HTNT titers are on average 14 . 78 IU/ml lower than RFFIT titers . Although samples with RFFIT titers ( above 200 IU/ml ) also had the highest values in the HTNT ( above 30–60 IU/ml ) , overall HTNT titers were of lower magnitude resulting in bias . When the Bland-Altman plots were analyzed without these outliers , as expected , the bias measurement decreased to -1 . 10 , indicating that there was a systemic difference of 1 . 1 IU/ml between RFFIT and HTNT values , with samples having slightly higher HTNT titers ( Supplemental Fig 1 ) . We also compared the rabies VNA titer values ( in IU/ml ) of positive animal sera obtained from the two assays . The Pearson coefficient value was statistically significant ( r = 0 . 96 ) demonstrating a strong linear relationship between the RFFIT and HTNT titer values ( Fig 6A ) . The CCC was also higher at 0 . 97 , indicating that the HTNT titers replicate RFFIT titers ( Fig 6B ) . Similar to human sera , Bland-Altman plots for animal sera to determine the differences of the titers obtained from the two methods were plotted against the average titers ( Fig 6C ) . The bias value of 1 . 95 ( HTNT titer was lower than RFFIT by a factor of 1 . 95 IU/ml ) observed with animal samples was much less than that of human sera , partly because of the absence of high titer samples . Since 0 . 5 IU/ml titer cut-off is required for demonstration of sufficient neutralizing activity and lower titers require rabies vaccination boosters ( based on the WHO / OIE recommendations ) , we compared the titers of positive samples obtained by RFFIT and HTNT methods . The results demonstrated higher correlation for RVNA titers greater than 0 . 5 IU/ml for animal samples by both methods ( S1 Table ) . Although 100% correlation was observed for human samples based on ACIP recommended complete neutralization at 1:5 cut-off , there were differences if 0 . 5 IU/ml was considered for minimal protective titer ( S1 Table ) . A subset of animal sera ( N = 14 ) were tested three independent times on different days to measure reproducibility of both assays . Negative samples did not neutralize at the 1:5 dilution in any of the replicate runs in either HTNT or RFFIT methods . Fig 7 represents the titer of positive replicate values for both the RFFIT and HTNT assays . No significant differences in titer values were observed using either method . As SRIG was included as a control in every HTNT assay , we measured the titer value of SRIG to evaluate reproducibility across days . The median 50% endpoint dilution for SRIG replicates was 160 ( 95% Confidence interval of 128 . 9–172 . 1 ) with a coefficient of variation of 21 . 4% demonstrating the reproducibility and consistency of assay .
The presence of antibodies to bind and neutralize RABV in vitro is a proxy for determining successful vaccination against rabies . The demonstration of rabies VNAs are pre-requisites for certain personnel whose job duties possess a high risk of contracting RABV infection , such as laboratorians , veterinarians and animal handlers [7] . The traditional assays for determining VNA titers , such as RFFIT or FAVN , utilize wild type RABV and antibodies conjugated to fluorescent dyes to measure levels of RABV infection to quantitate neutralization [17 , 18] . As an alternative , reporter-based virus neutralization methods , particularly GFP-based assays are utilized widely [22 , 23 , 27 , 28] . The GFP reporter based assay utilizes recombinant virus expressing GFP under the control of viral promoter to determine and quantitate viral infected cells without the need for additional antibody staining against viral proteins or other staining methods , like crystal violet , to determine survival or clearance of infected cells . Recombinant RABV expressing a GFP reporter gene has been previously developed and utilized for the determination of VNAs [20 , 21] . The results demonstrated significant correlation of the reporter-based assay VNA titer values with RFFIT results . In addition , assays using GFP reporter-based pseudo-type viruses ( lentivirus or Vesicular stomatitis virus ) expressing either RABV or other lyssavirus G proteins displayed concordance with traditional assays [29] [30] . In this study , we utilized RABV expressing GFP reporter with an NLS at the N-terminus for nuclear localization . The nuclear targeting offers two major advantages , ( 1 ) enhances fluorescence intensity and ( 2 ) enables highly accurate and automated quantification mechanism . GFP targeted to the nucleus as opposed to being distributed throughout the cell [20 , 21] concentrates the protein to a smaller surface area . Because the shape of nucleus is more homogenous than the overall shape of cells , automated methods can be employed for more accurate quantification of cells that express GFP [23] . We used DAPI staining to count all nucleated cells in a given well and calculated the proportion of infected cells based on GFP and DAPI co-localization . Nearly 100% co-localization efficiency was achieved using NLS targeting with accurate quantification . Confocal microscopy demonstrated expression and co-localization of GFP and DAPI in the infected cells based on N protein staining ( Fig 1B ) . The ArrayScan , a high content automated fluorescent microscope , used in this assay offers several advantages in detecting and analyzing neutralization data . The automated microscope scans and captures multiple images at different wavelengths to obtain DAPI and GFP signals of an entire well . In contrast , the standard RFFIT requires scanning only a subset of 20 fields ( 40% ) from a well for anti-rabies staining . Further , HTNT does not require this additional anti-rabies staining step and relies on GFP fluorescence . In addition , the images captured by the microscope are stored for future analysis . The time required to read five samples ( with 8 dilutions ) are similar by both manual and HCS instrument . However , with the capability of reading either 96- or 384- well plates and addition of a plate stacker with the instrument , can greatly increase the high throughput capabilities , an advantage for testing large sample sizes . In our assay , an average of 13 , 000 cells are counted in every well ( based on DAPI staining ) with nearly 50% of the cells expressing GFP from RABV infection in the absence of VNAs ( Fig 2 ) . Compared to RFFIT , the numbers ( total and infected cells ) obtained by HTNT are more accurate . These numbers are used to determine the percent of infected cells in the presence and absence of VNAs , generating more exact measurements of infection in the HTNT compared to RFFIT . Although detection methods are different , HTNT and RFFIT results correlated 100% in both sensitivity and specificity from a panel of both human and animal sera ( Figs 3 and 4 ) . Complete neutralization of RABV infection at the 1:5 dilution , either based on GFP or anti-rabies detection were consistent . The correlation between HTNT and RFFIT was better at lower VNA titer values compared to higher titers , as the differences observed are amplified with high titer samples . The Bland-Altman plot determined higher bias value for human sera , primarily for samples with high titers ( above 50 IU/ml ) as exclusion reduced the difference between RFFIT and HTNT values ( S1 Fig ) . Because titers in the animal panel were on the lower range , we did not observe a significant difference in Bland-Altman bias values in titers between the two assays . As RFFIT only accounts for fluorescence in 20 fields and not on the number of fluorescent foci ( or infected cells ) in each field , it is possible to have similar number of positive fields but may exhibit differences in number of RABV infected cells . On the contrary , HTNT determines the percent of infected cells at different dilutions across the entire well which tend to be lower than the RFFIT , and hence the titer values . One of the limitations of this HTNT assay is availability of the ArrayScan or other High Content Screen instruments in rabies testing laboratories . We are currently evaluating the HTNT assay using different platforms to determine the flexibility and ability for laboratories to adopt the HTNT based on specific availability and needs . In addition , we are comparing large sero-survey results obtained from RFFIT and HTNT , in order to exemplify the advantages of HTNT . | The potency of rabies vaccine is demonstrated by the presence of virus neutralizing antibodies ( VNAs ) in serum . It is critical to evaluate immunologic status of individuals who work directly with rabies virus ( RABV ) ( laboratorians ) or at high risk of infection due to interaction with animals ( veterinarians and animal control workers ) . In addition , rabies vaccination records and demonstration of VNAs in animals are mandatory before initiating pet travel to rabies-free counties or regions . Rabies VNAs are currently determined by the rapid fluorescent focus inhibition test ( RFFIT ) and the fluorescent antibody virus neutralization ( FAVN ) test , which measure the ability of antibodies to bind and prevent infection of RABV in vitro . Both assays require staining of infected cells using anti-rabies antibodies and manual observation of infected cells by a fluorescent microscope to determine VNA titers . In this study , we have developed a GFP reporter-based high throughput neutralization test ( HTNT ) for automated quantification of infected cells . This method has the advantages of allowing investigators to analyze and store the results , and can accommodate large sample sizes . Overall , the results from HTNT exhibited 100% correlation with that of RFFIT , albeit with differences in rabies VNA titer values due to quantification methods . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [
"medicine",
"and",
"health",
"sciences",
"immune",
"physiology",
"nuclear",
"staining",
"pathology",
"and",
"laboratory",
"medicine",
"pathogens",
"immunology",
"tropical",
"diseases",
"microbiology",
"light",
"microscopy",
"green",
"fluorescent",
"protein",
"viruses",
"rabies",
"luminescent",
"proteins",
"rna",
"viruses",
"microscopy",
"neglected",
"tropical",
"diseases",
"antibodies",
"research",
"and",
"analysis",
"methods",
"rabies",
"virus",
"immune",
"system",
"proteins",
"infectious",
"diseases",
"specimen",
"preparation",
"and",
"treatment",
"staining",
"zoonoses",
"proteins",
"medical",
"microbiology",
"fluorescence",
"microscopy",
"microbial",
"pathogens",
"recombinant",
"proteins",
"biochemistry",
"dapi",
"staining",
"lyssavirus",
"cell",
"staining",
"viral",
"pathogens",
"physiology",
"biology",
"and",
"life",
"sciences",
"viral",
"diseases",
"organisms"
] | 2018 | A high throughput neutralization test based on GFP expression by recombinant rabies virus |
Although the growth factor ( GF ) signaling guiding renal branching is well characterized , the intracellular cascades mediating GF functions are poorly understood . We studied mitogen-activated protein kinase ( MAPK ) pathway specifically in the branching epithelia of developing kidney by genetically abrogating the pathway activity in mice lacking simultaneously dual-specificity protein kinases Mek1 and Mek2 . Our data show that MAPK pathway is heterogeneously activated in the subset of G1- and S-phase epithelial cells , and its tissue-specific deletion results in severe renal hypodysplasia . Consequently to the deletion of Mek1/2 , the activation of ERK1/2 in the epithelium is lost and normal branching pattern in mutant kidneys is substituted with elongation-only phenotype , in which the epithelium is largely unable to form novel branches and complex three-dimensional patterns , but able to grow without primary defects in mitosis . Cellular characterization of double mutant epithelium showed increased E-cadherin at the cell surfaces with its particular accumulation at baso-lateral locations . This indicates changes in cellular adhesion , which were revealed by electron microscopic analysis demonstrating intercellular gaps and increased extracellular space in double mutant epithelium . When challenged to form monolayer cultures , the mutant epithelial cells were impaired in spreading and displayed strong focal adhesions in addition to spiky E-cadherin . Inhibition of MAPK activity reduced paxillin phosphorylation on serine 83 while remnants of phospho-paxillin , together with another focal adhesion ( FA ) protein vinculin , were augmented at cell surface contacts . We show that MAPK activity is required for branching morphogenesis , and propose that it promotes cell cycle progression and higher cellular motility through remodeling of cellular adhesions .
Receptor tyrosine kinase ( RTK ) signaling is a key mechanism through which extracellular stimuli guide development of the kidney and many other organs , but the specific in vivo functions of intracellular cascades activated downstream of RTKs remain poorly characterized . The kidney develops as a result of classical reciprocal inductive tissue interactions between the nephron-producing metanephric mesenchyme ( MM ) , and the branching epithelium of the ureteric bud ( UB ) , a structure later giving rise to the collecting duct system of the functional organ [1] . Renal differentiation begins with the formation of UB , which invades the surrounding MM , and subsequently starts its branching . UB morphogenesis is largely instructed by the MM , which secretes growth factors such as glial cell-line derived neurotrophic factor ( GDNF ) and members of fibroblast growth factor ( FGF ) family . Their RTK receptors , namely RET and FGF receptor 2 , expressed in UB epithelial cells , regulate UB development [2] . Based on genetic and in vitro experiments , GDNF/RET signaling is required for early UB morphogenesis [3]–[5] , while the requirement for FGFR signaling appears to arise later during normal kidney development [6] , or in situations where RET signaling is absent [7] . Although the molecular basis of UB branching has been extensively studied , relatively little is known of the cellular cascades and responses regulating the formation of new branches in vivo . Binding of GDNF and FGF to their receptors activates several intracellular pathways of which phosphoinositide 3-kinase ( PI3K ) /AKT , mitogen-activated protein kinase ( MAPK ) and phospholipase Cγ ( PLCγ ) function during renal differentiation [8] . Inhibition of the PI3K pathway in kidney organ cultures suggests that primary UB formation depends on chemotactic cell motility induced by this pathway [9] , whereas similar experiments with MEK inhibitors suggest that the MAPK pathway is also required for UB morphogenesis [10] , [11] . Attempts to genetically confirm such functions are largely missing although deletion of the protein tyrosine phosphatase Pntpn11 , which positively regulates MAPK , JAK/Stat and PI3K/Akt , suggests that these intracellular cascades also mediate pivotal functions during in vivo development [12] , [13] . Mutations in specific RET docking sites known to activate certain intracellular pathways indicate that induction of PLCγ via Y1015 as well as simultaneous activation of PI3K and MAPK via Y1062 pathways are involved in renal differentiation [14]–[17] . Active cell proliferation occurs in UB tips [18] , which are the major sites for generation of new branches formed through bifurcation of an existing buds [11] . In addition to proliferation , which appears to involve transient delamination of the cells from monolayer [19] , active cell movements needing constant turnover of cellular adhesions have been implicated in UB morphogenesis [20] , [21] . MAPK pathway , which is well known cell cycle regulator , functions through the RAS-RAF-MEK-ERK cascade , but its specific requirements during different cell cycle phases are highly cell type specific . The activation of RAF kinases leads to rather linear signal transduction upon phosphorylation of dual-specificity protein kinases MEK1 and −2 , which in turn phosphorylate ERK1 and −2 ( presently their only known substrates ) [22] . ERKs have a wide variety of nuclear and cytosolic targets including cyclin D1 and focal adhesion ( FA ) scaffold protein paxillin , which also associates with MEK [23] , [24] . Either disruption of ERK/paxillin complex or lack of ERK induced phosphorylation on serine 83 abolishes cell spreading and branching morphogenesis [24] , [25] . Interestingly , paxillin and another FA protein , vinculin , are found also in adherens junctions ( AJ ) , where they associate with β-catenin to modulate adhesion at sites of cell-cell contact [26] , [27] . Vinculin stabilizes E-cadherin at AJs where it potentiates E-cadherin mechanosensory responses [28] , [29] . Here we have studied the in vivo functions of MAPK pathway during renal branching by deleting Mek1 [30] specifically in UB epithelium in Mek2 -null background [31] . As previously suggested by chemical inhibition of MAPK in whole kidney cultures [10] , [11] , our results show definitively that loss of MAPK activity specifically in the UB prevents the generation of new branches while allowing bud elongation . The MAPK pathway appears to contribute to UB branching guidance by carrying out dual functions; it regulates G1/S-phase transition during cell cycle progression , and epithelial cell adhesion through paxillin phosphorylation affecting FA and AJ dynamics .
The pattern of MAPK pathway activity was first studied in kidneys at different developmental stages . As shown before [20] , pERK1/2 localized on one side of Wolffian duct epithelium at E10 . 5 , just before UB outgrowth . A day later when the UB had branched once to form the so-called T-bud , prominent pErk1/2 staining was detected both in the epithelium and surrounding MM ( Figure 1A ) . During subsequent branching , MAPK activity was restricted to UB tip regions , in a pattern similar to Ret expression , and to early nephron progenitors in the MM and dispersed cells in the medulla ( Figure 1B–D ) . Closer examination of pERK1/2 staining revealed striking heterogeneity in MAPK activity between adjacent epithelial cells ( Figures 1A–B and 2A–F ) . In the pseudo-stratified E10 . 5 Wolffian duct epithelium [20] , the mitotic nuclei localize at the apical surface , while the S phase nuclei are found on the basal surface , and G1/2 nuclei within the middle epithelial zone ( due to interkinetic nuclear migration ) . Most pERK1/2 positive cells in E10 . 5 Wolffian duct epithelium were found in the middle and basal zones ( Figure 2A ) . Later , during the active UB branching phase , a similar pattern of pERK1/2 was maintained ( Figure 2B–F ) suggesting that MAPK pathway is activated in a subset of cells in G- and S-phases of the cell cycle . Accordingly , pulse labelling of proliferating cells with uridine analog 5-ethynyluridine ( EdU ) followed by simultaneous detection of pERK1/2 and EdU-positive cells showed co-localization in the UB tips ( Figure 2B ) . Notably , while a large fraction of cells in G- or S-phases were positive for pERK1/2 , none of the mitotic cells ( identified by their round shape and expression of phosphorylated histone H3 ) were pERK1/2 positive ( Figure 2C–F ) . This was constantly found in six distinct kidneys ( E10 . 5–13 . 5 ) accounting approximately 100 UBs , and suggests lack of MAPK activity during mitosis in UB epithelial cells . Abundant pERK1/2 in developing kidney ( Figures 1 and 2 ) together with in vitro chemical inhibition studies [10] , [11] suggested that MAPK pathway could be important for kidney morphogenesis in vivo . Conventional knockout of Mek1 is embryonic lethal while generation of a conditional allele allows its tissue-specific deletion [30] from Wolffian duct and UB lineages using Hoxb7CreGFP transgenic mice [6] . This resulted in normal looking embryonic kidney ( Figure 3A and S1A ) , similarly to Mek1 deletion in epidermal keratinocytes [32] . Since ubiquitously expressed Mek2 can phosphorylate ERK1/2 and may compensate the loss of Mek1 in Hoxb7CreGFP;Mek1F/F kidneys , we reduced the gene dosages of Mek1 and −2 in UB epithelium . Conventional deletion of Mek2 alone , which results in phenotypically normal mice [31] , or UB-specific double heterozygosity for Mek1 and –2 , had no effect on UB branching , kidney differentiation or phosphorylation of ERK1/2 ( Figure S1D–F and data not show ) . Normal UB branching pattern and renal differentiation were observed in vivo and in organ culture , even in the absence of three out of four Mek1 and −2 alleles regardless of allelic combinations ( Figures 3B and S1A–L ) . Next Mek1 was removed from UB in Mek2-/- background to examine effects on renal differentiation . Hoxb7CreGFP;Mek1F/F;Mek2-/- ( henceforth called “dko” for double knock-out ) mice were born in the expected Mendelian ratio ( data not shown ) but died within 72 h due to obvious renal defects , including severe renal hypodysplasia and sporadic hydroureters ( 16% ) ( Figure 3C , E and G ) . Histological examination showed disorganized medulla-cortex compartmentalization and few but well-differentiated glomeruli with associated tubuli and dilated epithelium ( Figure 3D–E ) . Staining with the collecting duct epithelium marker calbindin and nephron segment markers Tamm-Horsefall and Na/K ATPase indicated that the cysts originate both in collecting ducts and secondarily in nephron tubules ( Figure 3F–G and S1M–N ) . Rudimentary kidneys in dko mice suggested that UB branching , the key process by which the kidney grows in size and acquires its typical shape could be perturbed . Time lapse imaging of in vitro cultured kidneys demonstrated a remarkable reduction in formation of novel branches in UB tips; average of 10 . 5 tips in controls was reduced to 3 . 8 in dko kidneys ( Figure 4A–F , p<0 . 001 two-tailed T-test , n = 5 ) . After generation of the primary UB at the correct time and with normal morphology , the subsequent epithelial morphogenesis in dko kidneys failed to start , and the UB tips usually elongated in only one direction ( mean trunk lengths in controls: 142 . 3 µm , n = 83 , three distinct kidneys , and in dko: 195 . 9 µm , n = 18 , two distinct kidneys , two-tailed T-test , p<0 . 05; compare Figure 4A-C to D–F , S1O ) . A similar phenomenon was observed in intact dko kidneys imaged by confocal microscopy ( Figure 4G–H ) . Typically very few if any UB tips at E13 . 5 had enlarged into T-bud resembling structures , which are signs of active branching . Thin UB tips were sparsely distributed in dko kidneys ( average of 8 . 9 tips/kidney ) , leaving large areas of kidney devoid of UB branches , while in control kidneys UB epithelium was distributed over the entire cortical surface areas of kidneys ( average of 41 . 9 tips/kidney , Figure 4G–I ) . Formation of the primary UB at E11 . 5 in dko kidneys ( Figure 4D ) was surprising given the strong pERK1/2 staining in early kidneys ( Figures 2A and 1A ) . This suggested that Mek1 might not yet be deleted by Hoxb7CreGFP in early UB epithelium , or that residual MAPK activity is maintained during the initiation of renal development . MEK1 was ubiquitously expressed in developing control kidneys while specifically lost from the UB of dko kidney from E11 . 5 ( n = 5 ) onwards ( Figure S2A–B ) , in contrast to pERK1/2 , whose localization and staining intensities in dko kidneys were comparable to control kidneys at E10 . 5 ( n = 5 , data not shown ) and E11 . 5 ( n = 4 , Figure S2C–D ) , and only abolished from the UBs at E12 . 5 onwards ( Figure S2E–F and data not shown ) . Application of exogenous GDNF to kidney cultures induces extra UB formation and swelling of UB tips ( Figure S2H and [5] ) . We employed exogenous GDNF and chemical MEK-inhibition in kidney cultures to further test if the MAPK pathway is dispensable for UB outgrowth from the Wolffian duct . MEK-inhibition by UO126 dose-dependently blocked UB branching in kidney cultures ( Figure S2I and K ) mimicking the defects seen in dko kidneys ( Figure 4D–F ) and previous findings [10] , [11] . Pretreatment with UO126 followed by application of GDNF in the presence of inhibitor blocked typical GDNF responses ( Figure S2J ) , suggesting that the function of MEK1/2 cannot be overcome by RTK activation . Simultaneous UO126-inhibition and activation of RET by exogenous GDNF without UO126 pretreatment ( Figure S2L ) had the same effect . Thus , normal UB outgrowth in dko kidneys is likely due to delay in abolishing pERK1/2 activity at early stages of renal development . As GDNF/RET signaling is the key RTK regulating UB morphogenesis in the normal context [7] , we wanted to evaluate the linkage between RET and the MAPK pathway at the molecular level . Previous genetic engineering of Ret gene on the docking site known to mediate concurrent activation of MAPK and AKT cascades showed their importance for renal differentiation , but evidence for a specific requirement for MAPK activity downstream of RET was lacking [14] , [17] . To address this , we examined if known GDNF/Ret signaling targets [33] , [34] are regulated through MAPK pathway . In situ hybridization of ten GDNF/RET target genes in control and dko kidneys ( Figures 5A–E and S3 ) revealed reduction in chemokine receptor Cxcr4 and in Spry1 , a negative regulator of RTK signaling , which exerts its action at least partially by blocking MAPK pathway [35] . Downregulation of specific GDNF target genes in Mek1/2-deficient UBs suggested that MAPK pathway is an important intracellular mediator of RET signaling . Various studies have shown that RTK signaling can promote proliferation through MEK/ERK pathway [23] and that ERK1/2 regulates G1/S transition in proliferating cells [36] , while its function in G2/M transitions and M-phase remains ambiguous [37]–[39] . To reveal the cellular basis of defective UB branch formation in the absence of MAPK activity , we first examined proliferation at the onset of the morphologically distinct phenotype . Analysis of the mitotic indices at E12 . 5 showed that the percentages of pHH3+ UB epithelial cells were comparable in controls and dkos , but the amount of UB epithelium in dko kidneys was significantly reduced when quantified as total number of epithelial cells ( Figure 6A–D ) . This data indicated that in the absence of Mek1 and −2 , UB epithelial cells initially enter mitosis as efficiently as control cells suggesting that G2/M phase occurs independently of pERK1/2 . However , at E14 . 5 dko UB was almost completely devoid of pHH3 ( Figure S4A–B ) showing a gradual decrease in mitosis . The reduced overall number of mutant UB cells indicated potential problems in cell survival or impairment in other cell cycle phases . Cleaved-caspase3+ apoptotic cells were very sporadically found in UB epithelium of both control and dko kidneys ( Figure S4C–D ) showing that increased cell death is not causing reduction in cell numbers of dko UB epithelium . Apoptosis was though slightly increased in renal mesenchyme , likely due to decreased UB-numbers in dko kidneys , which leave more mesenchymal cells without induction signal . The S phase was studied by labeling the newly synthesized DNA with 5-ethynyl-2-deoxyuridine ( EdU ) . Significantly fewer UB cells were EdU+ in the dko epithelium after 1 h pulse , revealing significant reduction of cells in S phase ( Figure 6E–G ) . Since ERK is known to regulate the induction of cyclin D1 [40] , [41] , whose up-regulation is a key step for G1/S transition , and cyclin D1 at mRNA level is positively regulated by GDNF [33] , it was a potential candidate for mediating the effect of the MAPK pathway on the cell cycle . In control kidneys cyclin D1 was up-regulated in early nephron progenitors and throughout the cortical UB epithelium ( Figure S4E , G ) . In the absence of Mek1/2 , only scattered , single cyclin D1 positive cells were very rarely detected in UB ( Figure S4F , H ) , supporting the idea that MEK1/2-activated ERK functions in the G1/S transition phase during epithelial branching morphogenesis . Greatly reduced formation of new branches in dko kidneys could involve alterations in cell adhesion properties either at the cell-to-matrix contacts made by focal adhesions , or at the cell-cell contacts formed by E-cadherin based adherens junctions . Since paxillin is a direct phosphorylation ( Ser83 ) target of ERK in innermedullary collecting duct cells [25] , we first studied the effect of chemical MEK-inhibition on pPaxillin in a ureteric bud derived cell line [42] . MAPK activity was dose-dependently inhibited by UO126 , which also reduced significantly the level of pSer83 paxillin ( Figure 7A–B ) . Immunofluorescence staining showed that inhibition of MAPK activity caused reduction in pPaxillin cytosolic pools but its plasmamembraneous localization was intensified together with E-cadherin , which appeared also stronger at the cell surfaces ( Figure 7C–F , S5A–B ) . Simultaneously vinculin , another FA protein was also more pronounced in the cell surface ( Figure 7G–H′ ) . This preferential membranous localization was time dependent as after 2 h of UO126 such differences were not obvious ( data not shown ) . We next tested if genetic loss of MAPK activity could have changes in FA proteins by generating primary cell cultures from UBs isolated from control and dko kidneys . All control UBs dissected at E11 . 5–12 . 5 ( n = 10 ) produced single cell monolayers in approximately 48 h in culture [21] , [42] , [43] , but dko UBs isolated at E12 . 5 were slower in delaminating from the epithelium ( 7/7 ) and two out of seven samples failed to generate monolayers ( data not shown ) . We thus isolated UBs from E11 . 5 dko kidneys ( before pERK1/2 was lost and when the morphological phenotype was still comparable to controls ) ( Figure S2D , 4C–D ) , and found that monolayer formation was significantly improved ( S5C–D′ ) but dko cells displayed thick , spiky E-cadherin at cell surfaces ( Figure 7I , L ) . The dko cells remained tightly packed and impaired in spreading after 48 h of culture as seen by very strong and large appearance of FAs ( Figure 7J–K , M–N ) . Like FAs , AJs are constantly formed and disassembled during development and tissue homeostasis [44] . We next studied AJ molecule E-cadherin , which localized abundantly in the apical end of lateral cell walls of the control UB tips , being otherwise uniformly distributed along the lateral membranes ( Figure 8A , C ) , while in medullary UB epithelium occasional staining was also observed in the basal end of lateral membranes ( arrows in Figure S6A ) . The UB tips of dko kidneys displayed stronger overall E-cadherin staining than controls ( Figure 8B , D , asterisks in S5B ) . Additionally , E-cadherin localization had shifted from apical to more basal sides of lateral membranes , and was also accumulating in basal membranes ( Figure 8B , D ) where it was rarely observed in control UB tips . Our visual observations were confirmed by quantitative measurements: E-cadherin intensities in general and the ratios of basal-to-lateral intensities were significantly higher both in cortical and medullary UB epithelium of dko than control kidneys ( Figure S6C ) . The cytoplasmic domain of E-cadherin is linked to the cytoskeleton through a complex of proteins including p120- , β- and α-catenins [45] . We analyzed β-catenin ( Figure S6D–G″ ) and filamentous actin ( Figure S6H and J ) , which appeared normal in dko kidneys . The function of p120 and related proteins ARVCF and p0071 is to cluster and stabilize E-cadherin at AJs [46] , but they appeared normal in dko kidneys ( data not shown ) . While the shift in E-cadherin subcellular localization could also reflect general problems in apical-basal polarity , the apical markers Par3 ( data not shown ) and ZO1 localized to apical cell surface similarly in control and dko UB epithelium ( Figure S6I and K ) , indicating that apical cell polarity was maintained in the absence of MAPK activity . Staining of control and dko kidneys with tight junction marker claudin7 [47] revealed similar localization in wild type and dko UBs ( Figure S6L–M ) suggesting that MAPK activity is specifically involved in regulating E-cadherin mediated AJs . To investigate potential functional consequences of the altered FAs and E-cadherin distribution , we analyzed UB epithelium by electron microscopy ( EM ) . Control UB epithelium appeared as a uniform sheet of cells tightly apposed through cell-cell contacts on their lateral membranes , while the double mutant epithelium had increased extracellular space and many gaps between neighboring cells as seen by disintegration of the lateral membranes at several sites along their contact surface ( Figure 8E–F , S7 ) . However , the contact sites where adhesion was maintained appeared more electron dense .
By genetic studies focusing on MAPK pathway functions in renal development , we show that only one allele of either Mek1 or Mek2 is enough to support normal renal branching . Simultaneous deletion of both genes abolishes formation of new branches and complex 3D patterns and shows that this pathway is a necessary mediator of RTK signaling . The functional requirement for MEKs in UB morphogenesis is similar to that reported in skin [32] but different than what has been observed in placenta , where double heterozygosity results in embryonic lethality [48] . Similarly , Erk1;Erk2 gene dosage and protein levels are critical for survival and normal proliferation as shown by generation of double heterozygotes , most of whom die during gestation [49] . Such differences in functional requirements of MAPK activity imply tissue specific roles for the pathway components and call for investigations on their specific contributions to development and homeostasis . While RET is an important RTK regulating renal branching , its exact cellular functions ( e . g . , proliferation , migration , extracellular matrix remodeling , changes in cell shapes and adhesive properties ) and their precise contributions have remained indistinct [50] . Identification of GDNF/RET targets recently revealed several interesting candidates for mediating these functions [33] , [34] but the intracellular cascades regulating the expression changes remain to be solved . Similarly to previous chemical inhibition studies [33] , we saw that the expression of transcription factors Etv4 and -5 was normal whenever UBs were present and thus does not require MAPK activity . This is opposite to the requirement of tyrosine phosphatase Shp2 , which appears to regulate Etv4 and -5 [12] . The negative RTK regulator , Spry1 , which suppresses MAPK activity , and chemokine receptor Cxcr4 , which is involved in migration in several cell types [51] were the only genes examined whose expression was reduced in dko UB epithelium . Our results suggest that different intracellular cascades regulate expression of specific target genes , and further experimentation will reveal potential genetic interactions between RET signaling and MAPK pathway . We show that lack of MAPK activity disturbs cell proliferation by reducing the number of cells in G1/S without primarily affecting mitosis , which is in line with the finding that pERK1/2 staining was absent in the mitotic , pHH3+ cells . Reduction in markers of G1/S phase but nearly normal cell numbers in M phase suggest changes in cell cycle kinetics , so that the double mutant UB cells spend a longer time in M-phase than control cells . Such a phenomenon was demonstrated in human retinal pigment epithelia , where sustained inhibition of ERK1/2 transiently delays the cell cycle progression [37] . The fact that mutant kidneys are smaller than those in control mice , together with the reduced cell number in G1/S , supports the view that MAPK activity , through the regulation of cyclin D1 levels in UB epithelium , is required for progression from G1- to S-phase as shown previously for several cell types [52] . Consequently the defects in G1/S transition also have an effect on mitosis , which is reduced at later stages of kidney development . In epithelial remodeling and morphogenesis , during which cells move relative to each other , cell-matrix and cell-cell contacts are constantly assembled and disassembled . We observed that UB epithelium lacking MAPK activity had significantly increased E-cadherin in general and particularly extending to more baso-lateral location than in controls , indicating problems in AJ dynamics . Rapid turnover of the E-cadherin-based homophilic AJs involves its endocytosis and recycling back to the plasma membranes [53] , and defects in such processes could result in sustained cell surface localization of E-cadherin . During endocytosis , internalization of E-cadherin is initiated by tyrosine phosphorylation of E-cadherin , which induces its dissociation from p120 [54] . Normal distribution of p120 and related proteins ( p0071 and ARVCF ) in dko UBs suggests normal E-cadherin endocytosis , as increased localization of p120-family proteins at the cell surface would be expected if endocytosis was affected . However , normal appearance of p120 does not conclusively exclude changes in E-cadherin endocytosis or recycling . Growing evidence indicates active crosstalk between different adhesion sites and this can be at least partially mediated by localization of certain proteins like vinculin and paxillin to both focal adhesions and AJs [26] , [55] . Paxillin is a direct phosphorylation target of ERK-proteins and lack of pERK-induced phosphorylation on serine 83 of paxillin leads to functional defects in cell spreading and branching morphogenesis [24] , [25] . We found accordingly that inhibition of MAPK activity in a ureteric bud-derived cell line [42] reduces general pPaxillin levels but also observed a simultaneous shift in its cellular localization to the cell-cell contacts , where concomitant increases in vinculin and E-cadherin levels were also detected . Additionally , primary cells derived from the double mutant UBs were impaired in their capacity to form monolayers , remained tightly packed and displayed stronger FAs than control cells , which all argue that such regulation is functionally significant . Both paxillin and vinculin bind to β-catenin in AJs , where at least vinculin has the potential to stabilize E-cadherin and , in certain cell types , to potentiate mechanosensing [27]–[29] . Taken together with the observation that lack of MAPK activity both in vivo and in vitro intensifies E-cadherin localization on cell surfaces , and recent evidence for cross-talk between FA and AJ , we suggest that MEK1/2-activated ERK1/2 regulates cellular adhesion by phosphorylating paxillin at Ser83 , which facilitates normal FA dynamics and composition . The molecular mechanism by which lack of paxillin phosphorylation leads to increased plasmamembranous localization of itself and vinculin remains a subject of future studies , but non-receptor tyrosine kinases Src and focal adhesion kinase ( FAK ) as well the small GTPase Rac1 are shown to mediate shuttling between FAs and AJs in other cell types , and therefore are good candidates [55] , [56] . Increased E-cadherin in the dko UB epithelium intuitively suggests stronger intercellular adhesion , but electron microscopy revealed big gaps in lateral membranes of adjacent cells . However the contact sites that were maintained in the dko epithelium appeared electron dense , suggesting that they might mediate strong contacts . Therefore it is possible that the gaps in lateral membranes are actually caused by increased adhesion at the localized sites where the contacts are maintained and then such stronger contacts generate ruptures to membranes next adhesion when cells are trying to move relative to each other . In support of this , vinculin in AJs promotes high E-cadherin based adhesion strength [57] . Our findings demonstrate the importance of Mek1/Mek2 in activation of MAPK pathway during UB branching , which is largely blocked in dko kidneys . Based on our results we suggest that MAPK activity regulates cyclin D1-mediated progression of cell cycle from G1 to S , and normal cellular adhesion through phosphorylation of paxillin in FAs . Absence of MAPK activity amplifies FA proteins vinculin and remnants of pPaxillin on cell surfaces where they participate in stabilization of E-cadherin to AJs possibly to strengthen intercellular adhesion . Taken together with heterogeneous pERK1/2 localization in UB epithelium , where certain cells show high ERK-activity and others no activity at all , this may imply that MAPK activity tunes cells for higher motility and thereby together with driving proliferation promotes novel branch formation .
Mek1-floxed , Mek2-null mice and Hoxb7CreGFP mice and their genotyping by PCR has been described [6] , [31] , [32] . All mice were on mixed genetic background with contributions from C57BL6/Rcc and 129/SvEv . Embryonic staging was as described earlier [58] and all experiments were approved by Finnish Animal Care and Use Committee . NMRI or Hoxb7CreGFP mice were used for MEK-inhibition experiments . Kidneys were cultured on Trowell-type system in medium of F12:DMEM/10%FBS/Glutamax/penicillin-streptomycin , and imaged with epifluorescent microscope ( see below ) . UO126 ( Cell Signaling technologies , Inc . ) dose-dependence testing was done with 5 , 10 , 15 and 50 µM by adding only the inhibitor , or UO126 in combination with GDNF ( produced by Icosagen Ldt . , see details in Supplementary information ) either at the same time or sequentially . Already 5 and 10 µM concentrations of UO126 inhibited UB branching but upon replacement with GDNF resulted in some degree of response , while 15 µM mimicked closest the UB pattern in dko kidneys . EdU ( 25 mg/kg ) was injected intraperitoneally to pregnant females at day 12 of pregnancy . After an hour pulse the females were sacrificed and embryos collected for 4% PFA-fixation followed by standard processing for paraffin embedding ( see below ) . Embryos and tissues of indicated stages were collected and fixed with 4% PFA . Further processing for frozen and paraffin sections was according to standard procedures , the latter utilizing an automatic tissue processor ( Leica ASP 200 ) . The primary and secondary antibodies used are shown in Table S1 . HE , whole mount and frozen section immunostaining were performed as previously described [20] , [58] . Paraffin sections were gradually dehydrated followed by heat induced antigen retrieval . Sections were then blocked with 10% fetal bovine serum ( Hyclone ) for one hour at RT followed by o/n primary antibody incubation . Fluorescent detection was performed similarly to that with frozen sections and chromogenic detection was done with EnVision Detection System-HRP ( DAB ) kit ( Dako ) . EdU incorporation detection step was done by using Click-iT EdU Alexa Fluor Imaging kit ( Molecular Probes/Invitrogen ) according the manufacturer's instructions after the antigen retrieval . PFA-fixed E13 urogenital blocks were embedded in 4% low melting agarose ( NuSieve GTG , Lonza ) and 50 µm thick vibratome sections were cut ( HM650 , Microm Int . ) for whole mount in situ hybridization performed with InSituPro automate ( Intavis ) according [59] . For EM , E11 . 5 and E12 . 5 kidneys were fixed with 2% glutaraldehyde ( Fluka ) followed by 1% osmiumtetroxide ( EMS ) post-fixation and graded series of dehydration . Transitional acetone incubation was performed before gradual embedding in Epon ( TAAB Ltd . ) . Ultrathin sections ( Leica ultracut UCT ultramicrotome , Leica ) were collected on pioloform-coated single-slot copper grids and post stained with uranyl acetate and lead citrate for analyzing with transmission electron microscope ( Tecnai G2 Spirit 120 kV TEM with Veleta and Quemesa CCD cameras ) operated at 100 kV . Epifluorescence images were produced with Zeiss Imager M2 Axio ( Germany ) equipped with Zeiss AxioCam HRm camera ( Germany ) and Axio Vision 4 software . Confocal imaging was done with mutiphoton Leica TCS SP5 MP confocal microscope ( Germany ) utilizing LAS-AF software . Chromogenic immunostaining was imaged with Olympus BX61 light microscope ( Japan ) equipped with Olympus Color View Soft Imaging System camera ( Japan ) and Cell F imaging software . Quantification of mitotic indices from total of four control and four dko kidneys sectioned through to give 150 and 127 sections , respectively , and stained similarly as previously described [21] followed by counting of all UB epithelial cells in both groups at 40× magnification . EdU percentages were counted at 40× magnification from entire kidneys sectioned through ( n = 4 for both groups ) . Proliferation significances were tested with Independent samples T-test ( 1-tailed , equal variances ) . Quantification of stalk lengths , as indicated in Figure S1O , was performed from 43 trunks of three distinct control kidneys , and 18 trunks from two distinct dko kidneys using Image J program . Statistical testing was done with Independent samples T-test ( 2-tailed , unequal variances ) utilizing the SPSS software . For quantification of basal E-cadherin intensities , E13 . 5 UBs were imaged with multiphoton confocal microscope at 63× magnification with 251 . 8 nm optical section thickness . Intensity measurements were performed with Image J on 12-bit images with applied 2 pixel median filter . The average basal and lateral pixel intensities were obtained as mean grey values ( MGV ) from four tips/genotype so that 70 control cells and 97 dko cells in total were analyzed in the way indicated in Figure S6C′ . Basal-to-lateral intensity ratios were calculated by dividing basal MGV with lateral MGV and cortical and medullar UB regions were compared with Independent samples T-test ( two tailed , equal variances ) employing SPSS program after removing three outlier ratios from the control and one from the dko set . UB cells were cultured in DMEM/10%FBS/Glutamax/penicillin-streptomycin media . 1×105 cells for MEK-inhibition experiments ( 2 h ) were plated 24 h before performing the inhibition by UO126 after which they were collected in lysis buffer . For paxillin phosphorylation , UB cells were serum-starved for 4 h and followed by fetal bovine serum ( FBS ) treatment of indicated times and 15 µM UO126 was used for inhibiting MAPK activity 30 min prior plus during the induction with FBS . Western blotting was done as previously described [60] . Rabbit anti-pErk1/2 ( Cell Signaling , 1∶2000 ) , anti-ERK2 ( K-23 , Sant Cruz , 1∶1000 ) , p ( S83 ) paxillin ( ECM Biosciences , 1∶1000 ) and mouse paxillin ( ECM Biosciences , 1∶1000 ) were used with HRP-conjugated secondary antibodies to detect proteins which were visualized by Pierce ECL Western Blotting detection system ( Thermo Scientific ) and Fuji film LAS1000 . For immunofluorescence staining cells were treated with 15 µM UO126 for 4 h , then fixed with 4% PFA for 10 min and stained with anti-E-cadherin ( R&D Systems , 1∶300 ) , pPaxillin , vinculin ( Sigma , 1∶500 ) and 568-phalloidin followed by visualization of primary antibodies with corresponding Alexa-fluor secondary ABs ( 1∶400 , Jackson Immuno Research ) . Cells were imaged by Zeiss Imager M2 Axio ( see above ) . All work on animals was conducted under PHS guidelines and approved by the relevant Institutional Animal Care and Use Committees . | Development of the ureter and collecting ducts of the kidney requires extensive growth and branching of an epithelial tube , the ureteric bud . While many genes that control this process are known , the intracellular signaling pathways that underlie renal morphogenesis remain poorly understood . The cellular changes that contribute to ureteric bud morphogenesis , such as adhesion and movements , are guided by intracellular signaling triggered by stimuli at the cell surface . Mitogen-activated protein kinase ( MAPK ) pathway is known to regulate proliferation in general , but its precise functions during different cell cycle phases are debatable . Moreover , the role of MAPK activity in control of cell adhesion has been demonstrated in cultured cells , but such regulation in vivo remains to be elucidated . Here , we examine the importance of the MAPK activity in ureteric bud branching , and find that simultaneous lack of Mek1 and Mek2 genes allows elongation of the bud but specifically arrests new branch formation . We show that lack of MAPK activity leads to changes in focal adhesion molecules and E-cadherin mediated cell adhesion and delay in cell cycle progression . Our findings may help to understand the origins of certain congenital malformations in humans . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"cadherins",
"mitosis",
"developmental",
"biology",
"cell",
"division",
"cell",
"adhesion",
"adhesion",
"molecules",
"molecular",
"development",
"cell",
"migration",
"biology",
"molecular",
"cell",
"biology",
"morphogenesis",
"cell",
"differentiation",
"cyclins"
] | 2014 | Mitogen-Activated Protein Kinase (MAPK) Pathway Regulates Branching by Remodeling Epithelial Cell Adhesion |
Antivenom is the treatment of choice for snakebite , which annually kills an estimated 32 , 000 people in sub-Saharan Africa and leaves approximately 100 , 000 survivors with permanent physical disabilities that exert a considerable socioeconomic burden . Over the past two decades , the high costs of the most polyspecifically-effective antivenoms have sequentially reduced demand , commercial manufacturing incentives and production volumes that have combined to create a continent-wide vacuum of effective snakebite therapy . This was quickly filled with new , less expensive antivenoms , many of which are of untested efficacy . Some of these successfully marketed antivenoms for Africa are inappropriately manufactured with venoms from non-African snakes and are dangerously ineffective . The uncertain efficacy of available antivenoms exacerbates the complexity of designing intervention measures to reduce the burden of snakebite in sub-Saharan Africa . The objective of this study was to preclinically determine the ability of antivenoms available in Kenya to neutralise the lethal effects of venoms from the most medically important snakes in East Africa . We collected venom samples from the most medically important snakes in East Africa and determined their toxicity in a mouse model . Using a ‘gold standard’ comparison protocol , we preclinically tested the comparative venom-neutralising efficacy of four antivenoms available in Kenya with two antivenoms of clinically-proven efficacy . To explain the variant efficacies of these antivenoms we tested the IgG-venom binding characteristics of each antivenom using in vitro IgG titre , avidity and venom-protein specificity assays . We also measured the IgG concentration of each antivenom . None of the six antivenoms are preclinically effective , at the doses tested , against all of the most medically important snakes of the region . The very limited snake polyspecific efficacy of two locally available antivenoms is of concern . In vitro assays of the abilities of ‘test’ antivenom IgGs to bind venom proteins were not substantially different from that of the ‘gold standard’ antivenoms . The least effective antivenoms had the lowest IgG content/vial . Manufacture-stated preclinical efficacy statements guide decision making by physicians and antivenom purchasers in sub-Saharan Africa . This is because of the lack of both clinical data on the efficacy of most of the many antivenoms used to treat patients and independent preclinical assessment . Our preclinical efficacy assessment of antivenoms available in Kenya identifies important limitations for two of the most commonly-used antivenoms , and that no antivenom is preclinically effective against all the regionally important snakes . The potential implication to snakebite treatment is of serious concern in Kenya and elsewhere in sub-Saharan Africa , and underscores the dilemma physicians face , the need for clinical data on antivenom efficacy and the medical and societal value of establishing independent preclinical antivenom-efficacy testing facilities throughout the continent .
Snakebite annually kills over 95 , 000 people [1] residing in some of the most disadvantaged rural communities [2] , and leaves about 300 , 000 surviving victims with permanent physical disabilities and stigmatising disfigurements . Since it is the most economically-productive and educationally-vulnerable 10–30 year olds that suffer most , snakebite also poses a significant additional socioeconomic burden on these remote , already impoverished communities . Available mortality data clearly indicate that snakebite deaths are greatest in Asia , and particularly in India [1 , 3] followed by sub-Saharan Africa ( Table 1 ) . The increasing concern over the plight of sub-Saharan African snakebite victims [4 , 5 , 6] focuses upon the higher case fatality in sub-Saharan Africa than elsewhere ( Table 1 ) and upon the declining availability of effective antivenom to treat snakebite victims . The crisis in supply of effective and affordable antivenom to treat snakebite victims in sub-Saharan Africa was first reported in 2000 [7] , and has since deteriorated . Akin to the late 1990s market failure of the Behringwerke-manufactured antivenom , Sanofi Pasteur had also supplied Africa with one of the most polyspecifically-effective and widely-used antivenoms , FavAfrique , but ceased its manufacture in early 2016 after a more than a decade of commercial disincentives . This latest market failure of effective antivenom particularly affected snakebite-treatment capability in those state , private ( mostly city-based ) and charity hospitals that could afford this relatively expensive antivenom ( $140/vial; [8] ) . The SAIMR polyvalent antivenom , manufactured by the South African Vaccine Producers Pty ( SAVP ) , was also widely used and recognised to be highly effective–but outside of the Southern Africa Economic Community it has become more expensive ( $315/vial , SAVP , personal communication ) than FavAfrique was and , also because of low production volumes , become increasingly difficult to source . Cognisant presumably of potential commercial opportunities and the public health needs engendered by the snakebite-therapy vacuum in sub-Saharan Africa , several non-Africa based antivenom manufacturers have in the past two decades produced polyspecific antivenoms marketed at costs considerably lower ( $18–75 ) than the FavAfrique or SAIMR antivenoms , and supplied in vastly greater quantities [8] . Superficially , this influx of new , affordable antivenoms into sub-Saharan Africa would seem highly desirable . However , in too many cases and African countries , this has not been the case—because some of these antivenoms have proved dangerously ineffective . Thus , reports from Ghana , Chad and the Central African Republic [9 , 10 , 11] document an increased case fatality rate ( from under 2% to over 12% ) following discontinuation of effective antivenoms and introduction of replacement products . In at least one case this was because the antivenom had been manufactured from IgG purified from horses immunised with venoms from Indian snakes–instead of venoms from African snakes [12] . Antivenom efficacy is predominantly restricted to snakes whose venoms were used in manufacture [13]–because the highly snake species-specific protein composition of venom dictates an equally specific IgG response in the immunised horses/sheep . Thus , the greater the biogeographic difference between the venom/s used in antivenom manufacture and the venom injected into the snakebite patient , the weaker the efficacy of the antivenom . For this reason , antivenom manufacturers are required to preclinically test and state the snake species for which their product is effective . Fig 1 evidences another antivenom marketed specifically for Central Africa but clearly inappropriately manufactured with venoms from Asian vipers . There are very few published reports on the clinical effectiveness of the several antivenoms in current use in sub-Saharan Africa . Preclinical efficacy data is therefore the only information available to physicians and government purchasers to decide which antivenom to use/purchase . However , many , perhaps the majority , of sub-Saharan African countries do not apparently subject newly-imported antivenoms to independent preclinical efficacy and safety testing , and clinicians and purchasers per-force base clinical use/purchase decision making upon manufacture-stated efficacy statements . The above reports of antivenom ineffectiveness and rising case fatalities , seemingly throughout much of sub-Saharan Africa , demonstrate this trust can be misplaced . There is therefore an urgent need to establish independent preclinical antivenom-efficacy testing facilities and expertise in sites throughout sub-Saharan Africa . With substantive new funding , the Liverpool School of Tropical Medicine has partnered with colleagues in Kenya , Nigeria and Cameroon to form the African Snakebite Research Group and established ‘Snakebite Research and Intervention Centres’ ( SRIC ) in each of these countries . Our remit includes improving the ( i ) availability of effective snakebite treatment in rural remote hospitals in greatest need and ( ii ) access to treatment for snakebite victims . To ensure this new programme is equipped with effective antivenom , and to provide the host government with independent antivenom-efficacy information , we purchased a vial of as many different antivenom brands as available from local pharmacies and preclinically tested their efficacy against venoms of the most medically important snakes in the region . This first report from the African Snakebite Research Group emanates from Kenya-SRIC activity and demonstrates , for the first time in East Africa , that there is substantial variation in the preclinical efficacy of the available antivenoms against the lethal effects of venoms from black mambas , spitting and non-spitting cobras , puff adders and saw-scaled vipers–and that no one antivenom is preclinically effective at the doses tested , against all these life-threatening snake venoms .
The antivenoms used in this study are described in detail in Table 2 and were acquired from a commercial pharmacy in Nairobi , except the SAIMR antivenoms that were donated to the first author from expired stocks held by Public Health England and had 2012 expiry dates . We were unsuccessful in purchasing one of the ASNA antivenoms manufactured by Bharat Serums and Vaccines Ltd that we had seen in a rural hospital in Kenya . All the antivenoms are manufactured as F ( ab' ) 2 fragments of IgG , and for clarity to non-specialist readers we have used the term IgG to describe these antivenoms . The comparative preclinical efficacy of the ‘test’ antivenoms was conducted before their respective expiry dates . We were unable to purchase the SAIMR polyvalent and ECHIS CARINATUS monovalent ‘gold standard’ antivenom in Kenya and therefore used 2012-expired vials donated to us from Public Health England . SDS-PAGE profiling of these antivenoms ( Supplementary S3 Fig ) reveal that the IgG in these SAIMR antivenoms possess the same structural integrity as IgG from the ‘in date’ test antivenoms . Further validation of using these expired SAIMR antivenoms for this study is provided by the comprehensive binding of venom proteins by IgG in these antivenoms . East Africa is resident to multiple medically important vipers , elapids and colubrids . For this analysis , we selected venoms from the most relevant representative species of each genus . Venom was extracted from over four specimens of wild-caught puff adders ( Bitis arietans , Kenya ) ; saw-scaled vipers ( Echis pyramidum leakeyi , Kenya ) ; black mambas ( Dendroaspis polylepis , Tanzania ) ; Egyptian cobras ( Naja haje , Uganda ) ; black-necked spitting cobras ( N . nigricollis , Tanzania ) and red spitting cobras ( N . pallida , Kenya ) maintained in the Liverpool School of Tropical Medicine herpetarium ( a UK Home Office accredited and inspected animal research facility ) . Freshly collected venom was snap frozen , lyophilised and stored as a powder at 4°C prior to reconstitution in phosphate-buffered saline ( PBS ) . The same batches of these venoms were used for each of the analyses below to provide cross-experiment continuity . We employed routine protocols in our laboratory [14] to measure the IgG titre , avidity , venom protein-specificity and protein ( IgG ) concentration/ml antivenom to provide a detailed immunological profile of the ‘gold standard’ and ‘test’ antivenoms .
We used the WHO-recommended antivenom effective dose ( ED50 ) assay , which measures the amount of antivenom required to prevent venom-induced lethality in 50% of mice ( 5/dose group ) injected with venom/antivenom mixtures . To assess the efficacy of the SAIMR ‘gold standard’ polyvalent and monovalent antivenoms , we determined the ED50 dose of ( i ) the SAIMR ECHIS CARINATUS monovalent antivenom against only the saw-scaled viper venom and ( ii ) the SAIMR polyvalent antivenom against venoms of the black mamba , the Egyptian , red and black-necked spitting cobras and the puff adder ( Table 4 ) . We assigned the SAIMR antivenoms as the ‘gold standard’ comparators because of their sub-Saharan African clinical effectiveness [17] and because , unlike FavAfrique , these antivenoms are likely to be available for the foreseeable future . We elected a ‘gold standard comparison’ experimental design to achieve our objective instead of 24 conventional ED50 assays ( 4 ‘test’ antivenoms tested against 6 venoms ) , which would have required a minimum of 600 mice ( 25 mice/experiment ) . Instead , we tested the extent to which the ‘test’ antivenoms prevented the death of mice ( 5/dose group ) injected with the venom/antivenom mixtures at volumes equivalent to half ( 0 . 5x ) , equal ( 1x ) or , where appropriate , 2 . 5-fold more ( 2 . 5x ) the volume of the SAIMR antivenoms that impart 100% survival of the mice ( calculated by doubling the ED50 dose volume ) . Thus , 100% survival of mice injected with 0 . 5x volume of a ‘test’ antivenom indicates a higher venom-neutralising efficacy than the SAIMR ‘gold standard’ antivenoms , and a test antivenom providing less than 100% efficacy at 1x volumes would be deemed less dose-effective than the ‘gold standard’ . Failure of an antivenom to neutralise the lethal effects of a venom at 2 . 5x volumes in 100% of mice would raise serious concerns as to the potential clinical efficacy of that product . This protocol enabled us to comprehensively analyse the efficacy , compared to a gold standard , of several antivenoms against several venoms using only 40% of the number of mice had we used conventional ED50 testing ( Fig 2 ) . It is important to note that , for the sake of clarity for non-specialist readers , we have described the efficacy of the ‘test’ antivenoms relative to 0 . 5x , 1x or 2 . 5x volumes of the SAIMR antivenom that protect 100% of mice . This differs from the more conventional description as ED50 volumes ( the volume of antivenom that protects 50% of the venom/antivenom injected mice ) , and antivenom preclinical testing practitioners are referred to Supplementary S1 Table that depicts the same data in Fig 2 in the context of antivenom volumes ( μL ) and amounts ( mg ) . To ensure valid cross-comparison of antivenom-venom reactivity , we first standardised the IgG concentration of each antivenom to 5 mg IgG/ml . We next incubated serial dilutions of each antivenom with the same concentration of each venom ( Fig 3 ) . For space reasons we have excluded the graph showing the baseline reactivity of the naive control horse IgG to all the venoms . The OD readings of the antivenoms at the 1:2 , 500 dilution , in the middle of the downward slope , provide the most immunologically meaningful comparison and , for clarity , are presented as tables in each panel of Fig 3 . For example , at this dilution , the SAIMR ECHIS CARINATUS antivenom IgG shows ( i ) highest binding to the E . p . leakeyi venom , ( ii ) some cross-reactivity to other viper ( puff adder ) venom proteins , and ( iii ) near-zero binding to the four elapid venoms–results entirely consistent with an antivenom generated by immunisation with only Echis species venoms . While detectable venom-binding differences between the antivenoms exist , none of the antivenoms exhibited sufficiently poor IgG binding to venoms that account for the very poor polyspecific ED50 results of , for example , the VINS and INOSAN antivenoms ( Fig 2 ) . Thus , at the same IgG dilution ( 1:2 , 500 ) the OD values ( venom-binding ) of VINS antivenom to all the venoms was greater or equivalent to that of the more preclinically effective Sanofi Pasteur antivenom . Furthermore , INOSAN’s inefficacy against D . polylepis venom contrasted with its higher OD to this venom than the considerably more efficacious Sanofi Pasteur antivenom . Finally , the OD values of the Premium Serums & Vaccines antivenom was consistently higher to all the venoms than the more efficacious SAIMR antivenoms . The ELISA IgG titration assay demonstrated that all the ‘test’ antivenoms contained venom-binding IgG titres not dissimilar to the ‘gold standard’ SAIMR antivenoms . Thus , while we identified many discreet IgG-venom binding differences between the antivenoms , we were unable to confidently attribute any of these as being responsible for the very different venom-neutralisation efficacies of these antivenoms . We therefore next performed an assay to identify whether the antivenoms possessed IgGs of variable avidity ( binding strength ) to the six different venoms that matched their distinct venom-neutralising efficacies . Ammonium thiocyanate ( NH4SCN ) is a potent disruptor of protein-protein binding ( chaotrope ) and , by measuring the ELISA OD readings of the same concentration of IgG and venom in the presence of increasing amounts of the chaotrope , is used to test antivenom IgG-venom protein binding strength . We incubated 1:1 , 000 dilutions of each of the 5 mg/ml standardised antivenom solutions with the venoms ( in the same concentration as for the IgG titre ELISA ) and determined the OD readings after addition of 0 , 1 , 2 , 4 , 6 and 8 moles of NH4SCN ( Fig 4; we have excluded the graph showing the baseline reactivity of the naive control horse IgG to all the venoms ) . The most immunologically-informative results were gained by comparing the percentage reduction in OD values of the antivenoms without NH4SCN to that in the middle of the downward slope at 4 M NH4SCN—as illustrated by the table inserted into each panel of Fig 4 . This assay revealed that the Premium Serums & Vaccines ( panel 4F ) antivenom possesses the most consistent , and highest , cross-snake species IgG-venom binding avidity of all the antivenoms , including the ‘gold standard’ antivenoms . The INOSAN antivenom ( panel 3E ) exhibits the least consistent cross-species venom binding avidity . The chaotropic ELISA assay revealed closer links between antivenom IgG-venom binding avidity and antivenom efficacy than the IgG titre results . Thus , the notably higher IgG-binding avidity of the SAIMR polyvalent gold standard ( panel 4A ) to the B . arietans , N . haje and D . polylepis venoms , and lower avidity to the spitting cobra ( N . nigricollis , N . pallida ) venoms accurately reflect the venom-neutralising dose-efficacy of this antivenom ( Table 4 ) . Also , the substantially higher IgG avidity of the INOSAN antivenom to E . p . leakeyi venom also matched its venom-neutralising dose-efficacy profile . However , this assay did not provide data accounting for the snake-species distinct venom-neutralising efficacies of the Premium Serums & Vaccines , VINS and Sanofi Pasteur antivenoms , particularly in comparison with the superior venom-neutralising dose-efficacy of the SAIMR polyvalent antivenom against N . haje , D . polylepis and B . arietans venoms . Venoms consist of multiple distinct protein groups and the protein composition of venom is markedly snake genus/species specific–with obvious implications on the venom protein-specificity of IgGs from venom-immunised horses . Neither the IgG titre nor the IgG avidity ELISA assays are designed to examine IgG binding to specific venom proteins . We therefore next used an immunoblot assay to investigate whether the six antivenoms express differences in IgG venom protein specificities that match their distinct venom-neutralising efficacies . Fractionation of the six snake venoms by 15% SDS-PAGE revealed the numerical and molecular size diversity of the venom proteins ( Fig 5A ) , with the cobra and mamba venoms ( N . nigricollis , N . pallida , N . haje and D . polylepis ) possessing a greater abundance of the low molecular mass neurotoxins/cytotoxins than the more evenly distributed molecular mass of the enzyme-rich , haemostasis-disruptive viper venoms ( E . p . leakeyi and B . arietans ) . To determine the extent to which this wide spectrum of East African snake venom proteins are bound by IgG of the six antivenoms , we electrophoretically transferred these venom proteins to a membrane and incubated those ( under identical conditions ) with the antivenoms at 1:5 , 000 dilutions ( Fig 5 ) . For this test , we did not adjust the antivenoms to a standard 5 mg/ml concentration . This analysis demonstrated that the intensity of the IgG-venom protein binding of the ‘gold standard’ SAIMR polyvalent and SAIMR ECHIS CARINATUS antivenoms was notably greater than all the ‘test’ antivenoms ( Fig 5 ) . It is important to note that , with minor brand-specific differences , that the difference between the ‘test’ and ‘gold standard’ antivenoms in this assay related to the intensity of venom protein binding and not the protein specificity . Thus , although the intensity of the IgG-venom protein binding was lower with the ‘test’ antivenoms , the immunoblots revealed that each antivenom possessed IgGs with similar venom protein specificities as the SAIMR ‘gold standard’ antivenoms . The lack of venom-reactivity of the control , naïve horse IgG ( Fig 4B ) , evidences the venom-specificity of the antivenom IgGs . The three immune assays above identified detectable differences in the IgG titre , avidity and venom-protein specificities of the six antivenoms , but these differences were relatively minor and could not be consistently applied to the interaction of each antivenom with each venom . We were therefore unable to identify an IgG-venom binding deficiency with any of the ‘test’ antivenoms , and by inference a deficiency in their venom-immunisation protocols , that could account for the ineffectiveness/weak efficacy of some of the ‘test’ antivenoms in our preclinical assays . The substantially higher IgG-venom protein binding intensity of the SAIMR polyvalent antivenom in the immunoblot assay suggested to us that this antivenom may be formulated with a higher amount of IgG/vial than the others . Our final test was therefore to determine the protein ( IgG ) concentration ( all antivenoms are formulated as F ( ab’ ) 2 fragments of IgG ) of each antivenom because this has an obvious bearing on dose-efficacy , and because it was not stated by any manufacturer , despite it being the active component of these therapies . We used a spectrophotometric instrument ( NanoDrop ) to measure protein content of each antivenom ( in triplicate ) and the results are presented in Table 5 . We included control horse IgG of different known concentrations , to confirm the accuracy of the NanoDrop results . For the sake of completeness , we also used SDS-PAGE analysis of each antivenom to demonstrate their consistent IgG purity ( Supplementary S3 Fig ) . This analysis demonstrated the substantial inconsistency in the total IgG content of the antivenoms , and , importantly , that the VINS and INOSAN antivenoms respectively contained 19% and 28% of the IgG concentration of the SAIMR polyvalent antivenom . The IgG content of antivenom therefore exhibited the closest association to their comparative efficacy in neutralising the venoms of N . nigricollis , N . pallida and E . p . leakeyi ( Fig 2 ) . It is not however a universally-applicable explanation , because , for example , IgG content alone does not explain the superior anti-B . arietans and anti-E . p . leakeyi efficacy of Premium Serums & Vaccines ( 63 mg/ml ) antivenom over that of the Sanofi Pasteur ( 96 . 7 mg/ml ) antivenom . Assessing the antivenom efficacy data by mg antivenom did reveal that the anti-E . p . leakeyi venom efficacy of the INOSAN antivenom was equivalent to that of the SAIMR ECHIS CARINATUS ‘gold standard’ antivenom . Nevertheless , as stated above and depicted in Fig 2 , it required 2 . 5 times more the volume of the SAIMR antivenom for the INOSAN product to achieve this efficacy parity , which in the human-treatment context has cost and adverse effect implications . For completeness , we have presented the amount ( mg ) and volume ( μl ) of antivenom used for each of the ‘test’ antivenoms for each of the doses examined into Supplementary S1 Table . To facilitate comparison , we have also added the amount ( mg ) and volume ( μl ) of the calculated 2xED50 doses of the ‘gold standard’ antivenoms that provided 100% protection to the envenomed mice . This table therefore provides all the numerical data related to the preclinical assays , and interpretations of efficacy from this table is no different from that presented in Fig 2 . No matter whether the data is examined by antivenom volume in μl or amount in mg , the least polyspecifically-effective antivenoms simply do not compare well to the ‘gold standard’ and to some of the other ‘test’ antivenoms . It was important that we tested the antivenoms by volume because that is the formulation in which the antivenom is provided by the manufacturer and used by the clinician . It is near impossible to envisage a clinician calculating the dose volume of antivenom he/she is going to administer based upon the mg/ml antibody content ( and perhaps impossible because , as here , manufacturers rarely provide this information ) . Therefore , to ensure that our preclinical efficacy testing of the antivenoms is of value to clinicians and medicine-purchasing agencies in East Africa , we undertook this testing , and report the results by antivenom volume . The reality is that the efficacy of a monospecific antivenom to its homologous venom is dictated by multiple factors , including IgG concentration , titre , avidity and protein specificity , which are themselves affected by the quality of the immunising venoms , the quantitative ratio of the venoms used for immunisation , the selected adjuvant and other aspects of the immunisation and antivenom-manufacturing protocols . All these interlocking factors are made more complex , as here , by increasing the number of venoms used to manufacture polyspecific antivenom , making it very difficult/impossible to confidently assign any one factor as being primarily responsible for lack of efficacy .
The plight of snakebite victims , particularly those in sub-Saharan Africa has been the subject of considerable recent attention [see 6 , 8 , 19 , 20 , 21] and the focus of recent Wellcome Trust and the Kofi Annan Foundation sponsored international meetings to identify remedial interventions . Reports from both meetings [4 , 5] identify the urgent need for the provision of effective antivenom and preclinical efficacy testing of existing and new antivenoms to ensure this . This recommendation aligns fully with the new WHO initiative to establish a prequalification programme for African antivenom , which includes establishing venom standards for sub-Saharan Africa and using those for preclinical antivenom-efficacy testing [22] . Over the past five years , we have generated a comprehensive inventory of sub-Saharan African snake venoms of defined gene and protein composition and murine toxicity [23 , 24 , 25] . This unique resource has enabled the Kenya Snakebite Research & Intervention Centre to conduct this first preclinical assessment of the efficacy of antivenoms available for clinical use in Kenya . This is important because the market failure of the Sanofi Pasteur antivenom and the very high costs of the SAIMR antivenoms leaves many East African countries with a choice of polyspecific antivenoms restricted to brands for which there is very little/no published data on their human effectiveness . Our preclinical results illustrate that the SAIMR polyvalent antivenom is considerably more effective in neutralising the murine lethality of the Egyptian cobra ( N . haje ) , black mamba ( D . polylepis ) and puff adder ( B . arietans ) venoms than any of the ‘test’ antivenoms . The ‘test’ antivenoms exhibited a superior or equal dose-efficacy as the SAIMR polyvalent antivenom against the spitting cobra venoms ( N . nigricollis and N . pallida ) . Preclinical neutralisation of the saw-scaled viper ( E . p . leakeyi ) venom was achieved by the Premium Serums & Vaccines and INOSAN antivenoms but required 2 . 5-fold greater volumes than the SAIMR ECHIS monovalent antivenom . Perhaps the most important result of our study was that no single antivenom , at the doses tested ( see below for a detailed consideration of this assay ) , was effective in neutralising the murine lethal effects of venoms of all six medically important snakes of East Africa—despite the pan-African efficacy claims inherent in the names of many of these products . The snake species-specific dose efficacy of each of the ‘test’ and ‘gold standard’ SAIMR polyvalent antivenoms suggests that the clinical management of envenoming by these snakes with any one of these antivenoms may require distinct , snake species-specific dose regimens . In the absence of a rapid snake species diagnostic test , this is clinically problematic and will likely result in the administration of too little or too much antivenom–both highly undesirable scenarios resulting in either inefficacy or increased risk of antivenom-induced adverse effects , respectively . On a more positive note , the preclinical efficacy of the Premium Serums & Vaccines product matched that of the highly regarded , but now unavailable Sanofi Pasteur product and approached that of the expensive SAIMR polyvalent antivenom . The Premium Serums & Vaccines product , at 26% of the cost of the SAIMR antivenoms ( see Table 6 ) , was the most affordable and effective antivenom of those tested here , although we note that preclinical efficacies against the two neurotoxic snake venoms ( N . haje and D . polylepis ) were weaker than against the other four species . Our results suggest that the preclinical efficacy of the VINS and INOSAN products could possibly be substantially improved by simply increasing the amount of IgG in each vial . The INOSAN product was the most expensive of the ‘test’ antivenoms , the most preclinically effective at neutralising the saw-scaled viper venom and one of the less effective antivenoms at neutralising the lethality of other snake venoms . It is notable that the IgG avidity of this antivenom to venoms of the six snakes varied more than the other antivenoms ( Fig 4 ) , and this avidity profile matched its snake-specific venom-neutralising efficacy . This may suggest that changes to the venom-immunisation regimen or analysis of the quality of the venoms could improve the venom-binding avidity and perhaps the efficacy of this antivenom . In consideration of the clinical efficacy of the SAIMR polyvalent and ECHIS monovalent antivenoms , it would be interesting to know whether SAVP has plans to combine the venom-immunising mixtures of these products to produce a truly pan-African polyspecific antivenom . We have carefully qualified our interpretation/extrapolation of the results of our preclinical ‘gold standard’ comparison assays to the efficacy of antivenom treatment of human snakebite patients , and urge readers to be similarly cautious . Our preclinical protocol enables a rapid efficacy comparison of a matrix of six venoms and four antivenoms using the minimum number of mice , but it differs from the recommended WHO antivenom-efficacy testing protocols in that it does not provide an ED50 value for each ‘test’ antivenom against each venom . Thus for example , we are unable to state whether , or not , the polyspecifically effectiveness of the VINS and INOSAN products could attain 100% efficacy against more of the venoms by using substantially greater volumes–because of the volume constraints inherent to this murine assay . From a human-treatment perspective , the necessity to administer multiple vials of an antivenoms carries important treatment costs ( the INOSAN product was the most expensive ‘test’ antivenom ) and adverse effects issues that risk poor uptake in rural tropical regions . On a more general note , these murine preclinical antivenom-efficacy testing assays are not infallible predictors of human efficacy . A recent study in Sri Lanka has questioned the value of predicting efficacy outcomes in human patients from ED50 results [26] . Conversely , the undoubted ability of the ED50 test to discriminate between effective and ineffective antivenoms in the murine model [13 , 27] , and the effective use of ED50 data [15] to help design the dose regimen of a human antivenom clinical trial [28] suggests that while the ED50 preclinical test has inadequacies many practitioners share , it should be retained until more accurate assays are tested , validated and become available . Thus , while the results of this study identify a potentially serious therapeutic concern , one of the priorities of the Kenya , Nigerian and Cameroon Snakebite Research & Intervention Centres will be to undertake conventional ED50 tests so that the Ministries of Health can be provided with pharmacopoeia-compliant data that they can act upon to restrict human use of preclinically-ineffective antivenom . The dialogue above underscores the urgent need for more published data on the efficacy of the different antivenom brands in use in sub-Saharan Africa to treat human patients . The paucity of this information reflects the prohibitive expense , problems of recruiting sufficient patients envenomed by the myriad of venomous snake species , the lack of accurate diagnostic tools to distinguish such species-distinct envenoming , and time required to conduct full clinical trials . Until the funding and required clinical/diagnostic tools become available , we remain reliant upon clinical observation studies that have importantly identified the wide-spread use of some dangerously ineffective antivenoms [9–11] . However , only the first of these reports provided the clinically-vital information on the antivenom brand . Another objective of the recently-established African Snakebite Research Group will be to conduct surveys in many rural hospitals in Nigeria , Kenya and Cameroon experiencing high snakebite admissions . The outcomes of these surveys will include reporting antivenom availability , and assessments of the clinical outcomes of treating patients with the various brands of locally-available antivenom . In conclusion , this first report of the African Snakebite Research Group identifies a worrying differential in the preclinical efficacy of available antivenoms in Kenya and underscores the need for independent preclinical testing of antivenoms throughout sub-Saharan Africa , and the need for venom standards for all the most medically important snakes to resource this testing . There is also an urgent need for practitioners to publish data on the clinical outcomes of treating patients with brand-named antivenoms ( despite the inherent problem of such transparency ) . The widespread availability of pan-African preclinical testing and clinical observation information will substantially help to improve the effectiveness of snakebite management and thereby reduce the high case fatality currently suffered by sub-Saharan African snakebite victims . | Snakebite is one of the most under-researched , under-resourced high morbidty/high mortality NTDs , as reflected by the fact that many of the antivenoms used to treat snakebite victims in sub-Saharan Africa are of uncertain and untested efficacy . This Kenya case study is the first examination of the preclinical efficacy of all available antivenoms to neutralize the venom toxic effects of the most medically important snakes in any region of sub-Saharan Africa . Our results identify serious preclinical efficacy limitations in two of the most commonly used antivenoms , that no single antivenom is effective against all regionally important snakes and that the least effective antivenoms had the lowest IgG concentrations . It is our aim that Ministry of Health medicine-supply regulators can use this data as evidence to demand more detailed efficacy evidence from manufacturers , and to justify the establishment of national/regional preclinical testing facilities . We hope this publication will also alert physicians treating African snakebite victims to check the efficacy of antivenom in their pharmacies . We have carefully qualified the extent and limitation of the results and of our interpretation of the clinical implications thereof . | [
"Abstract",
"Introduction",
"Materials",
"and",
"methods",
"Results",
"Discussion"
] | [
"medicine",
"and",
"health",
"sciences",
"toxins",
"enzyme-linked",
"immunoassays",
"pathology",
"and",
"laboratory",
"medicine",
"tropical",
"diseases",
"geographical",
"locations",
"vertebrates",
"animals",
"toxicology",
"toxic",
"agents",
"vaccines",
"cobras",
"reptiles",
"neglected",
"tropical",
"diseases",
"infectious",
"disease",
"control",
"immunologic",
"techniques",
"africa",
"snakebite",
"research",
"and",
"analysis",
"methods",
"venoms",
"infectious",
"diseases",
"vipers",
"immunoassays",
"snakes",
"people",
"and",
"places",
"eukaryota",
"squamates",
"biology",
"and",
"life",
"sciences",
"amniotes",
"organisms"
] | 2017 | Preclinical antivenom-efficacy testing reveals potentially disturbing deficiencies of snakebite treatment capability in East Africa |
Helminth infections affect more than a third of the world’s population . Despite very broad phylogenetic differences among helminth parasite species , a systemic Th2 host immune response is typically associated with long-term helminth infections , also known as the “helminth effect” . Many investigations have been carried out to study host gene expression profiles during helminth infections . The objective of this study is to determine if there is a common transcriptomic signature characteristic of the helminth effect across multiple helminth species and tissue types . To this end , we performed a comprehensive meta-analysis of publicly available gene expression datasets . After data processing and adjusting for study-specific effects , we identified ~700 differentially expressed genes that are changed consistently during helminth infections . Functional enrichment analyses indicate that upregulated genes are predominantly involved in various immune functions , including immunomodulation , immune signaling , inflammation , pathogen recognition and antigen presentation . Down-regulated genes are mainly involved in metabolic process , with only a few of them are involved in immune regulation . This common immune gene signature confirms previous observations and indicates that the helminth effect is robust across different parasite species as well as host tissue types . To the best of our knowledge , this study is the first comprehensive meta-analysis of host transcriptome profiles during helminth infections .
Helminth infections , also known as helminthiases , are estimated to affect > 2 billion people and especially prevalent in developing countries . Despite recent progress in helminth control programs , much work remains to be done to achieve eradication [1] . To develop effective intervention strategies , it is critical to understand in detail the molecular mechanisms underlying host immune responses to these parasites . The ability of helminths to manipulate the host immune system allows the infection to persist for years without being eliminated . This helminth effect has been observed with vastly divergent parasites , including nematodes , trematodes and cestodes , and has attracted increasing attention over the past decade . Although the term “helminth” is used to describe a diverse array of phylogenetically different organisms , evolutionary pressure from the host immune system is likely to have selected for convergent phenotypes among these pathogens with immuno-modulatory capacity [2 , 3] . Indeed , host immune responses to helminthiases are typically characterized as Th2-type accompanied by general immune down-regulation [4] . Th2 responses involve the cytokines interleukin ( IL ) -3 , IL-4 , IL-5 , IL-9 , and IL-13 and additional features such as eosinophilia , goblet cell and mast cell hyperplasia , as well as alternatively activated macrophages [5] . Immunosuppression , on the other hand , is mainly mediated by regulatory T cells secreting IL-10 and TGF-beta . Helminth-mediated immunomodulation is an active process that requires live parasites and can be carried out through different mechanisms [4] . In particular , helminth parasites secrete soluble mediators that can interact with host immune cells through evolutionarily conserved chemical dialogs [6] . For instance , studies have shown that a schistosome cysteine protease induces IgE production [7] and proteins secreted from adult Nippostrongylus brasiliensis elicit strong Th2 immune responses [8] . Host immunity can be modulated at different life-cycle stages of helminths , including larvae and eggs [9 , 10] . Additionally , immunomodulation can be mediated through direct interactions between helminth surface molecules and host cells . For example , glycoconjugates and lipoproteins expressed on the surface of helminths can interact with host cells and shape immune responses [11] . Molecular mimicry is also utilized by these parasites to alter the host immune response . Homologues of anti-inflammatory molecules such as TGF-beta [12] , macrophage migration inhibitory factors [13] and SOCS-1 [14] are produced by helminths . Finally , recent studies have shown that the helminth effect may also be mediated indirectly via changes in the gut microbiota [15 , 16] . Host immune responses during helminth infections represent the outcome of eons of co-evolution between parasites and hosts . In addition to the immunomodulation mediated by helminths to prolong the infection , hosts have also evolved mechanisms to limit the damage associated with helminth infections . Host responses to helminths are characterized by a trade-off between immunity against the parasite and immunopathology caused by prolonged inflammatory responses . Indeed , the host often favors tolerance of helminth infection over complete parasite destruction to limit immune-mediated damages [5] . Studies in mouse models have shown that IL-4 deficient mice infected by schistosomes die due to excessive inflammation eight weeks post-infection , while wild-type mice enter a chronic phase of schistosomiasis [17] . Maintaining a regulated immune response through Tregs or a modified Th2 immune response is pivotal for the well-being of the host . Upon immune dysregulation , pathologies can arise as in the case of hepatic fibrosis in schistosomiasis or elephantiasis associated with lymphatic filariasis [3] . Understanding the helminth effect will not only provide enlightenment on fundamental aspects of helminthiases , but can also serve as a basis to develop new treatment strategies for immune disorders , such as asthma , allergy , inflammatory bowel disease and type 1 diabetes . Helminthiases are thought to help prevent the onset and reduce the severity of these diseases , based on observations that higher prevalence of parasitic infections in a population is associated with lower incidence of immune-related diseases [18] . Subsequent experimentations on non-obese diabetic mice infected with helminths revealed that they are able to prevent the onset of type 1 diabetes [19 , 20] . Many human studies have explored the therapeutic value of helminths to treat and prevent immune-mediated diseases , particularly with nematodes . Experimental data have shown positive results in the treatment of inflammatory bowel disease as well as multiple sclerosis [21] . The immunoregulatory environment created by helminth infections extends beyond reduction of responsiveness to parasite antigens [22] , as the infected patients are also less responsive to other antigens such as vaccines , allergens and self-antigens [23] . Global gene expression profiling using microarray or RNAseq technologies has been widely used to reveal biological and cellular processes involved in host immune responses during infection . The growing number of available datasets in public repositories such as Gene Expression Omnibus ( GEO ) [24] and ArrayExpress [25] offers excellent opportunities to perform comprehensive analyses by integrating multiple studies to identify robust gene expression signatures that would be otherwise unidentifiable in individual studies [26] . The benefits of a meta-analysis lie in the ability to limit or eliminate potential biases associated with individual studies , and to improve statistical power to enable detection of subtle but biologically meaningful variations through increased sample sizes . It has been widely used to identify patterns of gene expression in complex pathophysiological conditions , including infection and inflammation [27 , 28] , aging [29] and cancer [30] . For instance , in a landmark meta-analysis study by Jenner and Young , gene expression datasets from both in vitro and in vivo infection models were combined to reveal a common host transcriptional response against a wide array of different pathogens ( bacteria , virus and protozoa ) across different cell types [27] . To the best of our knowledge , no comprehensive meta-analysis of helminthiases has been reported . The hypothesis of our current study is that the helminth effect can be identified and characterized as a unique gene expression signature across different parasite species and host tissues , and by applying suitable meta-analysis procedures to integrate different datasets , this signature can be revealed to provide insights into the molecular mechanisms underlying the interactions between host and helminth parasites .
GEO and ArrayExpress were queried during June and August 2015 to collect datasets related to helminthiases . We focused on mouse models because our initial survey indicated that the majority of available datasets are based on this host . The following criteria were applied during data selection: ( 1 ) immune response profiling after at least one week post-infection; ( 2 ) complete raw or normalized gene expression data available; ( 3 ) presence of suitable control group; ( 4 ) minimum of four samples with two controls and two cases; and ( 5 ) in vivo studies with tissue samples from wild-type mice . As our focus is on normal immune responses , we excluded in vitro studies of particular cell types , experiments using transgenic hosts or experiments focusing on specific research subjects that may result in transcriptional changes not representative of normal infected states ( i . e . , study of gene expression after drug treatment ) . Several additional datasets were excluded due to lack of metadata or poor quality . The detailed procedures are described in Fig 1 . Individual data annotation and analysis were performed using the INVEX web-based tool [31] . In particular , probe IDs from different microarray platforms were converted to Entrez gene IDs . If more than one probe was mapped to the same gene , then the mean gene expression value of these probes was used for that gene . After probe annotation , individual datasets were log transformed followed by auto scaling . We then performed differential expression ( DE ) analysis between the control and the infected groups for individual datasets using the limma approach [32] . The resulting gene expression patterns were visually inspected using the interactive heatmap . To reduce potential study-specific batch effects , the pre-processed and normalized data sets were further subjected to the well-established ComBat procedures [33] . The method has been specially designed for adjusting batch effects in microarray expression data with small sample sizes . Firstly , genes absent in more than 80% of samples are eliminated and the remaining ones are standardized to have similar overall mean and variance; Secondly , information is pooled across genes from a batch to estimate batch effects that affect many genes such as increased level of expression and higher variability; Thirdly , these batch effects are then adjusted to obtain normalized data . To compare the sample clustering patterns before and after applying the ComBat procedures , the results were visually examined using the principal component analysis ( PCA ) . The first two principal components were then used to create a scatter plot with the colors and shapes of the data points corresponding to different studies or tissue types for visual assessment . The meta-analysis was performed using the web-based tool—INMEX [34] . Two popular meta-analysis methods were explored–the p-value combination using Fisher’s approach , and the effect size combination using the random effect model . Effect size is a standardized difference defined as the difference between group means divided by its standard deviation ( i . e . Z-score ) . In particular , Fisher’s approach combines the p values from several independent tests bearing upon the same overall hypothesis and evaluates the significance using chi-squared tests; while effect size combination takes into consideration both the direction and magnitude of gene expression changes to generate more biologically consistent results . The random effect model was chosen because there were significant cross-study heterogeneities based on the Cochrans’ Q test . An adjusted p value of 0 . 01 , based on the false discovery rate using the Benjamini–Hochberg procedure , was used to select DE genes . The merged expression data containing all DE genes was subjected to complete hierarchal clustering and explored using various interactive visualization methods offered within NetworkAnalyst [35] . Genes within the identified major clusters were subjected to over-representation analyses using Gene Ontology [36] , PANTHER classification system [37] , and Reactome pathways [38] . To characterize the relationships among those co-upregulated genes , we projected them into the mouse protein-protein interaction ( PPI ) network based on InnateDB [39] . Two subnetworks were further created for up-regulated and down-regulated clusters , respectively , for visual inspection and functional assessment .
As shown in Fig 1 , after data filtering based on the inclusion-exclusion criteria , nine studies were left , involving five different helminth parasites ( Schistosoma mansoni , S . japonicum , Fasciola hepatica , Trichinella spiralis and Nippostrongylus brasiliensis ) collected from four different tissues ( lung , intestine , liver and diaphragm ) and three mouse strains ( BALB/c , C57BL/6 and CBA ) , with a total of 55 samples [40–47] . The details of the nine datasets are provided in Table 1 . These pathogens represent two divergent trematode groups and two divergent nematode clades . The complete list of the 39 datasets together with detailed annotation is given in S1 Table . After probe annotation and filtering , we identified a total of 11 , 573 genes that are common across all nine datasets . Using an adjusted p-value cut-off of 0 . 05 , an average of 9 . 06% of genes was found to be significant . Most of the results are similar to those reported in the original publications . In some cases , however , we notice that different data normalization procedures could significantly affect the number of DE genes identified . As our goal is to identify the core immune response signature across different parasite species and tissue types , it is important to estimate the effect of these factors before meta-analysis . As shown in S1 Fig , the 55 samples are tightly clustered mainly according to the original studies . After batch effect correction using the ComBat approach , this study-specific clustering pattern mainly disappeared ( S2 Fig ) and the resulting PCA plot showed clustering based primarily on control and infection groups on the first PC ( Fig 2 ) . Another consideration is tissue-specific effects . These effects are hard to assess due to the paucity of available studies—both intestine and diaphragm are only represented by a single study each . However , the PCA plot ( Fig 2 ) clearly indicates that the overall clustering patterns are dominated by the helminth effect . We further explored the corresponding PCA loading plots and confirmed that the main genes driving the separation between control and infection groups are immune-related rather than tissue-related . We also investigated the possible effects of different mouse strains . As shown in S3 Fig , no particular clustering patterns were detected for different strains , indicating that the strain effects are unlikely to play a significant role in our meta-analysis focusing on helminth effect . S4 and S5 Figs further show the changes of two individual genes across different datasets before and after applying the ComBat procedures to help illustrate the batch effect correction procedures . Over 5 , 000 DE genes were identified with Fisher’s approach using an adjusted p value cut-off of 0 . 01 . However , close inspection indicated a significant proportion of genes showing contradictory expression changes in different studies , which may be caused by the inefficiency of Fisher’s approach when very small numbers of samples are available in individual datasets . In contrast , combining effect sizes using the random effect model produced more biologically meaningful results . Using an adjusted p-value cut-off value of 0 . 01 , we obtained 691 DE genes ( S2 Table ) . The result table for all 11 , 573 genes is provided in S3 Table . The fold changes of the top 14 immune genes across all nine studies are depicted in Fig 3 . It illustrates the strength of meta-analysis–improved statistical power for identification of consistent expression changes across different studies , and using the collective power to overcome potential bias in particular studies . For instance , the small negative fold changes of Mrc1 and Hmox1 genes in the GSE41941 dataset ( containing only two replicates ) are most likely due to study-specific bias when compared to the overall larger positive fold changes in other datasets . The hierarchical heatmap ( Fig 4 ) shows that these DE genes can be clustered into two distinctive groups , with one major group of 394 genes that are consistently up-regulated during helminth infection , and the other group of 297 consistently down-regulated genes . Among these 691 DE genes , 145 of them were uniquely identified only in the meta-analysis . To gain insights into the main functions of those genes within the two major clusters , we performed functional enrichment analyses . PANTHER GO biological process enrichment analysis of all 691 genes showed that 133 genes are involved in “immune system process” ( p-value of 2 . 32E-9 ) . Among them , 123 are up regulated . Comparing with the genes responsible for common host responses against diverse pathogens identified by Jenner and Young [27] , there is an overlap of 33 genes . The most significant overlap is found in chemokine and cytokine genes , followed by immune signalling genes . Among these genes , there are several sets of genes involved in immunological process that are overrepresented . Up-regulation of immune regulatory genes is particularly significant , with 27 genes identified as having inhibitory activity on immunological processes , such as negative regulation of B cell receptor signaling pathway ( GO:0050859 ) , negative regulation of antigen receptor-mediated signaling pathway ( GO:0050858 ) , negative regulation of interferon-gamma production ( GO:0032689 ) , etc . As for down-regulated genes , the vast majority of GO categories are related to metabolism . Further inspection of the list of up-regulated genes shows that many of them are involved in Th2- related immune responses , including the Th2 cytokines Il4 and Il33 and others ( Ccl7 and Jak3 ) . Il4 can antagonize Il1 activity by up-regulating the expression of Il1r2 , a decoy receptor for Il1 [48] . Interestingly , we observed that both Il1rn , an antagonist of Il1 , and Il1r2 are overexpressed; both genes encode proteins that inhibit activities of Il1 . In addition , the list also contains many markers of alternatively activated macrophages ( AAM ) ( Mrc1 , Retnla , Chi3l3 and Trem2 ) [49] . Unlike classically activated macrophages , AAM are induced by Il4 and Il13 produced by CD4+ Th2 cells and other cells during type 2 immune responses; AAM are characterized by the production and release of anti-inflammatory factors along with the ability to promote wound repair and extracellular matrix remodelling [49] . These main immune functional groups are discussed in detail below . Enrichment analyses showed that 56% of the down-regulated genes are involved in metabolic processes ( GO:0008152 ) . Out of 47 GO annotations , 36 of them directly relate to metabolic process . The most significant down-regulated gene is glutathione S-transferase alpha 3 or Gsta3 , with a p-value of 5 . 14E-12 and combined effect size of -4 . 85 . This gene is involved primarily in glutathione metabolism and also plays an accessory role in autophagy [50] . Interestingly , this gene is also found to be among the top up-regulated genes in mice model of type two diabetes [51] . Another down-regulated immune gene is Rorc . This gene is associated with induction of Th17 responses [52] . The main metabolic pathways affected are metabolism of lipids and lipoproteins . The group biological oxidation is also significantly down regulated . Down-regulation is also observed in four genes in the cytochrome P450 family . This family encompasses a large number of genes which play important roles in hormone synthesis and breakdown , cholesterol synthesis , vitamin D metabolism , and metabolism of xenobiotic compounds and bilirubin in the liver [53] . Overall , the magnitude of changes in expression changes is much smaller and more heterogeneous in these down-regulated genes than in genes in the up-regulated cluster . Indeed , using effect size as a reference for the magnitude of changes in gene expression , only 30% of down-regulated genes had an effect size ≥ 2 compared to 74% of up-regulated genes . The complete DE gene list is provided in S1 Table . To further characterize our data , we examined the top 50 DE genes ( Table 2 ) . Among them , 39 are up-regulated and 11 are down-regulated . The “Immune effector process” category is highly represented among these top genes . This category includes Col3a1 , Fcgr2b , Cd44 , Lat , Sash3 , Rorc , Hmox1 , Lcp1 , Anxa1 , Lpxn , Psmb8 , Themis2 and Samsn1 . Cd44 is a receptor involved in cell-cell and cell-matrix interactions that has diverse roles including maintaining tissue integrity , mediating lymphocyte activation , homing and interacting with diverse chemical factors and hormones [60]; Lat is involved in the mediation of T-cell signaling [61]; Sash3 plays an important role in the differentiation of T cells [62]; Psmb8 is a subunit of the immunoproteasome whose function notably includes the processing of antigens [63] . Interestingly , among these DE genes , only Rorc is down-regulated . This gene is associated with the pro-inflammatory phenotype and induction of Th17 responses , a type of immunity that has been associated with autoimmune diseases [52] . A previous study reported immunosuppression of Th17 response mediated by helminth infection resulted in attenuation of autoimmune disease [64] . It is interesting to note that Fcgr2b , Col3a1 , Hmox1 and Samsn1 function in immunosuppression ( GO:0050777 ) . Additionally , Lpxn is involved in negative regulation of B-cell signaling [65] and Anxa1 participates in the glucocorticoid-mediated anti-inflammatory effects [66] . Ch25h is not classified as an immune-related gene , but is associated with activation of macrophages and dendritic cells by various TLR ligands [67] . It catalyzes formation of 25-hydroxycholesterol [68] , a compound that has wide array of regulatory effects on immune cells . Emr1 and Themis2 are involved in macrophage activation [69 , 70] , while Mrc1 and Retnla over-expression suggests the presence of AAM , which are associated with Th2-type immune responses [49] . Additionally , the overexpression of Hdc , a gene in the histamine synthesis pathway [71] , suggests that activation of mast cells is associated with helminth infection , as noted previously [72] . All these changes provide evidence to support a helminth infection-related signature of Th2-type immune response . To understand the relationships among these co-regulated genes , we projected all DE genes into the mouse PPI network and obtained a densely connected subnetwork containing 253 of those DE genes and 1069 of their interacting partners ( Fig 5A ) . For up-regulated genes , a total of 48 genes directly interact with each other and form a small subnetwork with Irf8 at the center ( Fig 5B ) . Irf8 is a transcription factor that plays an important role in regulating lineage commitment and maturation of myeloid and lymphoid cells [73] . It is also associated with negative immunoregulatory functions such as inhibition of Th17 immune response [74] . This subnetwork also includes key immunoregulatory genes including Cd274 , Fcgr2b , Lpxn and Hmox1 . Cytokines and cytokine receptors include Ccl3 , Cxcl9 and Cxcr3 . It also contains Myd88 , a key regulator of Nfkb transcription factor . For the down-regulated genes , we could not identify a direct interaction subnetwork . Instead , these genes require extra interactions partners to be connected ( Fig 5C ) . The resulting subnetwork contains 51 genes with Foxo3 at the central position . Interestingly , evidence of immunosuppression can be witnessed in this subnetwork . Foxo3 is a transcription factor that can modulate the function of dendritic cells to control the magnitude of T cell immune responses [75] . It also interacts with transcription factor Foxp3 to regulate Treg cell development . The subnetwork also contains several other immune genes such as Rorc , a transcription factor for Th17 cells , and Strap , a negative regulator of Tgf-beta signaling .
Host responses to helminth infections are a well-studied subject , however , to the best of our knowledge , no comprehensive gene expression meta-analysis has been reported in this area . In this paper , we reported the results from a comprehensive meta-analysis of host immune responses in helminthiases . In particular , by integrating datasets from multiple studies across different tissue types and helminth species , we were able to identify a robust and consistent pattern of gene expression , representing the core immune signature of host response during helminth infection . Functional enrichment analyses of these genes revealed that immune-related genes are up regulated and those primarily involved in metabolic processes are mainly down regulated . Up-regulated genes are mainly involved in inflammatory responses , immune regulation and Th2 immunity . These genes and functions will serve as a basis for elucidating immune pathways targeted by helminths and facilitate future mechanistic studies to characterize the interactions between host and helminth parasites at molecular level . The hygiene hypothesis suggests that a lack of exposure to infectious agents such as parasitic helminths in early childhood can lead to an increased risk of developing chronic inflammatory , autoimmune and allergic diseases . Helminth infections , along with other microorganisms , are thought to play an important role in the proper functioning of immunoregulatory mechanisms [76] . Excessive immune reactivity in Th1 , Th2 and Th17 responses is harmful [77] . Helminth infections have been associated with reduced responses to specific parasite antigens [78 , 79] and more generally , to bystander antigens [80] . Our study shows a broad range of DE genes associated with immune responses , with a significant portion involved in modulation of immune functions . We observed a general up-regulation of genes involved in negative regulation of antigen receptor-mediated signaling . This set of genes consists of Lpxn , Thy1 , Lgals3 and Ptpn6 . Lgals3 interacts with helminth glycans [81] and plays an important role in modulation of host responses during helminth infections [82] . The common up-regulation of these inhibitory genes suggests that they might be mediators of the immunomodulatory consequences of helminth infection . Additionally , implications of helminth infections in the attenuation of immune response associated with autoimmune diseases can be witnessed in our results as well . Indeed , we observed an up-regulation of genes involved in the PD-1 pathway and down-regulation of Rorc , a key transcription factor involved in Th17 development . Similarly , we observed an up-regulation of transcription factor Irf8 , which has inhibitory activity towards the differentiation of Th17 cell [74] . It is possible that helminth infection reduces incidence of autoimmune disorders through negative regulation of Th17 response . The induction of Th2-type immune responses is mediated by multiple factors . Tissue injuries caused by parasite invasion can promote a Th2-type response for wound repair . Recent studies have shown that inflammatory responses associated with a sterile wound induced recruitment of AAM to the site of injury; depletion of these cells resulted in delayed wound repair [83] . However , an injury alone is not sufficient to maintain a Th2-type response; an adaptive response involving Th2 CD4+ T cells is required [84] . We observed up-regulation of AAM markers in addition to a set of genes involved in hemostasis , tissue repair and extracellular matrix remodelling . These changes in gene expression may be associated with responses to tissue injury caused by parasite migration or tissue invasion within the host . Helminths themselves can also promote Th2 responses by producing excretory-secretory products and expressing Th2-inducing surface molecules . Many helminth glycans can mediate immunoregulation [85] . Clec10a , a C-type lectin receptor , recognizes LacdiNac GalNacB1-4GlcNac ( LDN ) and GalNacA1-O-Thr/Ser ( Tn ) , glycan structures which are present in multiple helminth species , including S . mansoni and T . spiralis [85] . In the case of S . mansoni , Clec10a , along with CD209 and Mrc1 , recognizes soluble egg antigens and mediates Th2 responses through MHC class II antigen presentation [86] . We found both Mrc1 and Clec10a to be consistently up-regulated by helminth infections . Similarly , the beta-galactose-recognizing family of lectins , also known as galectins ( Lgals3 ) , may interact with helminth glycans , as many of them contain terminal galactose [87] . Specifically , galectin-3 has been shown to interact with LDN . In a murine in vivo study , LDN antigens induced formation of Th2-associated granulomas in liver , possibly mediated by interaction with Lgals3 due to its high up-regulation in granulomas [88] . In our data , a common up-regulation of these genes confirms the results of previous studies and indicates that these gene products may be involved in multiple host-parasite interactions . There are several limitations associated with this current meta-analysis . The first limitation is the relatively small number of datasets ( containing nine datasets and 55 samples in total ) used in this study , despite our best efforts in data collection , processing and integration using advanced statistics and visualization techniques; Secondly , the results are based on murine models which may not be able to accurately reflect the immune responses that occur in human clinical settings as shown in a recent study on chronic clinical hepatic schistosomiasis [89] . Thirdly , the results could potentially bias towards liver immune responses as four out of the nine datasets were from the liver tissue . Despite these limitations , our analyses reveal a general host immune signature to helminth infections , with constituent genes and biological processes to enable better understanding of host-parasites interactions . The signature will be further refined and validated with increasing number of gene expression studies in the future . | Many studies have been conducted to understand the immune modulatory effects in helminth infections . To determine whether there is a common transcriptomic signature characteristic of the helminth effect , we performed a comprehensive meta-analysis of publicly available gene expression datasets . The results revealed a distinct pattern of gene expression that is consistent across multiple helminth species and host tissue types , with upregulated genes dominated by those involved in immune regulation , Th2 immunity and inflammatory responses . | [
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] | [
"invertebrates",
"medicine",
"and",
"health",
"sciences",
"pathology",
"and",
"laboratory",
"medicine",
"helminths",
"gene",
"regulation",
"immunology",
"parasitic",
"diseases",
"animals",
"mathematics",
"statistics",
"(mathematics)",
"research",
"and",
"analysis",
"methods",
"mathematical",
"and",
"statistical",
"techniques",
"gene",
"expression",
"pathogenesis",
"immune",
"response",
"helminth",
"infections",
"host-pathogen",
"interactions",
"meta-analysis",
"genetics",
"biology",
"and",
"life",
"sciences",
"physical",
"sciences",
"statistical",
"methods",
"organisms"
] | 2016 | Comprehensive Transcriptome Meta-analysis to Characterize Host Immune Responses in Helminth Infections |
Flavivirus and Filovirus infections are serious epidemic threats to human populations . Multi-genome comparative analysis of these evolving pathogens affords a view of their essential , conserved sequence elements as well as progressive evolutionary changes . While phylogenetic analysis has yielded important insights , the growing number of available genomic sequences makes comparisons between hundreds of viral strains challenging . We report here a new approach for the comparative analysis of these hemorrhagic fever viruses that can superimpose an unlimited number of one-on-one alignments to identify important features within genomes of interest . We have adapted EvoPrinter alignment algorithms for the rapid comparative analysis of Flavivirus or Filovirus sequences including Zika and Ebola strains . The user can input a full genome or partial viral sequence and then view either individual comparisons or generate color-coded readouts that superimpose hundreds of one-on-one alignments to identify unique or shared identity SNPs that reveal ancestral relationships between strains . The user can also opt to select a database genome in order to access a library of pre-aligned genomes of either 1 , 094 Flaviviruses or 460 Filoviruses for rapid comparative analysis with all database entries or a select subset . Using EvoPrinter search and alignment programs , we show the following: 1 ) superimposing alignment data from many related strains identifies lineage identity SNPs , which enable the assessment of sublineage complexity within viral outbreaks; 2 ) whole-genome SNP profile screens uncover novel Dengue2 and Zika recombinant strains and their parental lineages; 3 ) differential SNP profiling identifies host cell A-to-I hyper-editing within Ebola and Marburg viruses , and 4 ) hundreds of superimposed one-on-one Ebola genome alignments highlight ultra-conserved regulatory sequences , invariant amino acid codons and evolutionarily variable protein-encoding domains within a single genome . EvoPrinter allows for the assessment of lineage complexity within Flavivirus or Filovirus outbreaks , identification of recombinant strains , highlights sequences that have undergone host cell A-to-I editing , and identifies unique input and database SNPs within highly conserved sequences . EvoPrinter’s ability to superimpose alignment data from hundreds of strains onto a single genome has allowed us to identify unique Zika virus sublineages that are currently spreading in South , Central and North America , the Caribbean , and in China . This new set of integrated alignment programs should serve as a useful addition to existing tools for the comparative analysis of these viruses .
Flaviruses , including Dengue , Yellow Fever , Japanese Encephalitis and West Nile viruses , are significant public-health pathogens responsible for wide-spread epidemics . Recently , another member of this genus , Zika virus ( ZIKV ) , has emerged as a global public health threat ( reviewed in [1] . Two major ZIKV lineages have been recognized: an African lineage first detected in the Uganda Zika forest in 1947 , and an Asian lineage , first isolated in South East Asia during the 1950s , that has since spread to the Americas ( for review , [2 , 3] ) . Phylogenetic analysis has revealed that both the African and Asian lineages can be further divided into distinct sublineages or groups [4 , 5] . Recent studies have also shown that ongoing epidemics are accompanied by the continued diversification of viral sequences via accumulation of base substitutions and recombinant exchanges between related sub-groups [3 , 6 , 7] . Members of the Flavivirus genus have been grouped based on their vectors ( reviewed in [8] ) . Mosquito-borne human pathogens include ZIKV , Yellow Fever virus , four Dengue virus species , St . Louis and Japanese encephalitis viruses , and West Nile virus , along with other highly diverse less-characterized groups for review , [8] . Although mosquitos are considered the primary vector for ZIKV transmission , recent studies have identified human to human transmission via sexual contact [9] . Analysis of Filovirus human outbreaks during the last 49 years , from the initial 1967 Marburg virus outbreak in Germany through the most recent 2014–15 Ebola virus epidemic in West Africa and in the Congo , indicates that these pathogens will continue to pose serious public health risks ( reviewed in [10–12] . Ebola virus species involved in these outbreaks and other non-human infections include the Zaire , Sudan , Taï Forest , Reston and Bundibugyo species , with the Zaire strains responsible for the most extensive human outbreak [13 , 14] . Likewise , multiple Marburg outbreaks have occurred in Kenya , the Congo , Angola , Uganda and South Africa ( for review , [11 , 15] . Studies indicate that each Filovirus genus may have its own particular transmission cycle that includes non-human primates , bats , rodents , domestic ruminants , mosquitoes and ticks ( reviewed in [16] ) . While bats are considered the primary reservoir for many of these viruses [17 , 18] , studies on humans that survive acute Ebola/Zaire infections reveal the presence of persistent active virus within immune-privileged or tissue sanctuary sites [19] . Phylogenetic analyses of both Ebola and Marburg strains responsible for human and non-human primate hemorrhagic fevers reveal that genetically identifiable strains from distinct lineages are associated with individual outbreaks; during these outbreaks , evolving sublineages have emerged [20–27] . For example , sequence analysis of Ebola isolates collected during the 2014–2015 West African Zaire/Makona outbreak has revealed the presence of multiple distinct sublineages that can be temporally traced to an initial Guinea strain that diversified during its spread into Liberia and Sierra Leone [13 , 28–31] . The availability of hundreds of Flavivirus and Filovirus genomic sequences is an important resource for acquiring insights into the evolution of these pathogens [32 , 33] . Using current web-accessed alignment tools , when multiple viral genomes are compared , alignments are often difficult to visually assimilate given the large size of their readouts . For example , a ClustalW alignment [34] of 14 ZIKV strains produces a 51-page readout . In addition , web-accessed alignment programs restrict the number of viral isolates that can be compared in an individual alignment . To circumvent these limitations , we have developed a multi-genome alignment method that can superimpose hundreds of one-on-one alignments to reveal sequence polymorphisms and conservation as they exist within a sequence of interest [35 , 36] . Individual one-on-one input:database alignments can also be accessed directly from the input-centric readouts . The combined EvoPrinter/Clustal alignment algorithms described here access databases of hundreds of Flavivirus or Filovirus genomes , allowing the user to input a full or partial viral sequence to initiate a comparative analysis . EvoPrint readouts identify sequences shared by all selected strains , in addition to highlighting ( through color-coding ) unique base substitutions and those shared by subsets of database entries . EvoPrinter databases currently contains 1 , 094 Flavivirus entries including 148 ZIKV strains and 460 Filovirus genomes with 393 Zaire isolates from the recent West African Ebola outbreak . To demonstrate the utility of these comparative tools , we show how 1 ) alignment readouts highlight unique bases in both the input and database sequences; 2 ) multiple sublineages are identified within ongoing Florida , Dominican Republic , Puerto Rico , and Brazil ZIKV outbreaks; 3 ) SNP analysis of other ZIKV strains also reveals different Central American , Caribbean and Chinese sublineages; 4 ) novel Dengue2 and Zika recombinant viruses and their parental lineages were identified using differential SNP pattern screens; 5 ) SNP patterns differentiate between Ebola/Zaire sublineages; 6 ) host cell A-to-I hyper-editing within Ebola and Marburg genomes is identified by SNP profiling and 7 ) inter-species multi-genome Ebola virus alignments can identify ultra-conserved sequences .
Genomic sequences were curated from the NCBI/Genbank database [32] , and additional information about virus strains was obtained from the Virus Pathogen Database [33] . To ensure that duplicate genomes do not interfere in the identification of uniquely shared sequences among different strains , redundant entries ( detected by BLAST or Evoprinter alignments ) were excluded . Database genome names contain the following information: species , NCBI designation , country of origin and year of isolation . When available , additional information is included in the names , such as lineage assignments , group designations and/or serotypes [14 , 25 , 37–44] . A lineage represents a set of genomes that differ from others within a species by a unique assemblage of sequence polymorphisms when compared to other species members . Different lineages are often marked by greater than 50 unique lineage-specific base differences . In addition to FASTA formatted sequences , each entry was formatted for enhanced-BLAT ( eBLAT ) alignments to speed initial database searches [36 , 45] . For eBLAT alignments , each genome was indexed into non-overlapping 11-mers , 9-mers and 6-mers and used to generate independent BLAT alignments that are superimposed to produce an eBLAT readout [36] . As of April 2017 , the Flavivirus EvoPrinter database contains 1 , 094 non-redundant genomes that include the following: 574 Dengue ( groups 1–4 ) ; 37 St . Louis Encephalitis; 115 West Nile; 110 Japanese Encephalitis; 70 Yellow Fever; 148 Zika; 8 Aroa-related; 7 Edgehill-related; 3 Entebbe-related; 3 Natya-related; 2 Spondweni-related; 12 Yaounde-related; 14 Insect-specific; 4 No Known Vector; 5 Seabird Tick-associated; and 8 Tick-borne genomes . Flavivirus groupings correspond to those previously described [8 , 46 , 47] . Databases will be updated when new genomes are submitted to NCBI . The Filovirus database currently consists of 460 genomes that include 66 Marburg strains and 393 Ebola ( 371 Zaire , 10 Sudan , 7 Reston , 4 Bundybuygo , and 1 Taï Forest ) isolates . Also included in the database is a single Cuevavirus genus strain , Lloviu Cuevavirus , isolated from European cave bats [48 , 49] . To initiate the comparative analysis of a user-provided sequence , an eBLAT search is performed to identify database genomes that closely match the input sequence [36] . User-supplied sequences can range from 100 bases to complete genomes . Once the eBLAT search identifies the input species , one-on-one Clustal alignments using the alignment algorithms developed by [34] are generated between the input sequence and the intra-species database genomes . Although BLAT alignments are significantly faster than Clustal comparisons , aligning bases at or near sequence ends are often missed due to insufficient K-mer alignment target lengths . Pairwise alignments are then converted to distinguish between aligning bases ( upper case ) and non-aligning bases ( lower case ) within the input sequence for each comparison [45] . This input-centric format allows for the superimposition of alignment data from an unlimited number of pairwise comparisons [35 , 36] . In addition , holding one-on-one alignment data in memory instead of multi-genome alignments allows for user-customized comparisons . To achieve higher throughput volumes and processing speeds , we wrote a Java-based program that employs multithread parallel processing [50] to generate pairwise alignments concurrently . By random allocation of 144 computational threads , database search and alignment processing speeds are significantly enhanced using a Hewlett Packard 2 . 5GHz/512 GB RAM; 4 socket , 18-core processor server operating with the RedHat Enterprise Linux 6 operating system . User-provided Flavivirus sequences ( including full genomes ) are automatically aligned to all intra-species database genomes and , to speed up processing times , alignments to the larger 18 kb Filovirus genomes are done incrementally , with the initial alignment round to the top ten eBLAT scoring Filovirus genomes . Additional database genomes can then be added to include strains of interest . From the genome selection tree , the user can select genomes for single or multi-genome comparisons with the input sequence . The genome selection page orders the one-on-one alignments , based on the number of base mismatches with the input sequence ( least to most ) . The selection page allows the user to 1 ) view individual alignments with the input sequence by selecting the genome of interest; 2 ) view multi-genome superimposed alignments of all or a selected subset of genomes in order to either identify shared or conserved sequences via an EvoPrint readout or to highlight sequence differences by generating an EvoDifference print readout , and 3 ) initiate inter-species alignments . By moving back and forth between the genome selection page and alignment readouts , the user can quickly add or remove viral strains from the comparative analysis . Sequence differences in multi-genome EvoDifference print readouts are color-coded to highlight base differences that are 1 ) unique to the input , 2 ) differ in only one of the database genomes , or 3 ) differ in two or more of the database entries ( Fig 1 ) . While sequence identity among the aligning genomes is indicated by gray-colored text in the EvoDifference print , conserved sequences within an EvoPrint readout are denoted by black text ( Fig 2 ) and less conserved sequences are shown in gray font highlighted in green . In addition , bases that are unique to the input sequence and not present in any of the database genomes included in the EvoPrint readout are highlighted in red . The start and stop translation codons of open reading frames are highlighted when included in the alignment . For Flaviviruses , protein boundaries for the processed polyprotein are annotated ( positions taken from the Virus Pathogen Resource [51] ) . Sidebars to the right of the readouts delineate protein encoding ORFs . Sequence lines in both EvoDifference and EvoPrint readouts can be expanded to view the alignment details for each of the database genomes and , by selecting a virus strain listed in the readout , the user can view its one-on-one alignment with the input sequence . Amino acid alignments can also be viewed from one-on-one ORF alignments , to allow the user to assess whether nucleotide changes result in different encoded amino acids . A tutorial that details these alignment steps is available at the Flavivirus or Filovirus EvoPrinter websites via the EvoPrinter homepage ( https://evoprinter . ninds . nih . gov/evoprintprogramHD/evphd . html ) . As an alternative to a user-provided sequence analysis , a database genome can be selected as the input reference sequence for either individual or multi-genome alignments . EvoPrinter keeps a library of one-on-one alignment data between all Flavivirus database entries and a separate library for Filovirus database alignments that can be accessed for rapid comparative analysis . As with the user supplied input sequence search , database alignments are ordered on the genome selection page based on numbers of base mismatches compared to the input and individual alignments can be viewed by clicking on the database genomes . To resolve different lineages and/or sublineages , the user should select ten or more genomes that have similar mismatch numbers with the input reference sequence and generate a multi-genome EvoDifference print . On the genome selection page , the bracketed numbers after the database name represent the number of base mismatches with the input sequence . In the readout , bases in black text indicate two or more database mismatches , and when these multi-genome differences are identical in two or more strains they frequently represent lineage or sublineage identity SNPs . In other words , when multiple strains have the same base substitutions these SNPs can be considered markers of lineage progression . By expanding readout lines that contain multiple mismatches , sublineages can be differentiated by their uniquely shared differences with the input ( see Figs 3 and 4 ) . One-on-one EvoDifference print SNP patterns can be used to identify recombinant viruses and their parental lineages . Virus strains that are closely related to the input sequence , as revealed by low mismatch numbers ( listed after the database genome name on the Genome Selection Tree ) , usually have randomly distributed base differences throughout their pairwise alignments with the input sequence . Discontinuity in mismatch scores between related database genomes , as seen by a sudden jump in score values , are often due to one of two reasons . First , a higher score can indicate a sublineage difference and in this case , the increased SNPs are randomly distributed throughout the alignment . Second , the higher score could indicate a recombinant exchange , and in this case , a cluster of high-density SNPs ( a recombinant fragment from a more divergent minor parent ) would be flanked by regions of lower SNP densities ( from the major parent ) . Alternatively , if the recombinant is aligned with a member of the minor parental lineage , a significantly reduced low-SNP density region ( corresponding to the above high SNP density cluster ) is flanked by regions of higher SNP densities ( from the major parent ) . To identify members of the minor parental lineage , the database search is repeated using the region of the input sequence that generates the high SNP density cluster along with flanking sequences of the putative recombinant strain . If members of the minor parental lineage are present in the database , they will likely have the lowest mismatch numbers when compared to the other database genomes . By repeating the initial search using the complete or nearly complete recombinant genome and then comparing one-on-one alignments with members of both parental lineages , the genomic region that generates the high SNP density when aligned to a major parental lineage strain ( Figs 5A and 6A ) will show near identity within the corresponding region when aligned to a minor parental strain ( Figs 5B and 6B ) . If a member of the minor parental strain is detected first , the members of the major parental lineage can be identified in database searches by using the low SNP density region plus its flanking higher SNP density regions and examining high mismatch scoring strains . Differential SNP patterning can also be used to identify recombinant strains that are decedents of multiple rounds of recombinant exchanges with different partners . For example , if not all of the high SNP density clusters observed in the recombinant / major parental lineage alignment have corresponding “SNP clearings” when aligned to a member of the minor parental lineage , then the recombinant strain most likely is a mosaic of different recombination events involving multiple partners . To confirm putative recombinants , we recommend that additional recombinant detection programs be employed such as the Recombination Detection Program [52] . Both one-on-one and multi-genome EvoDifference prints of related Ebola or Marburg strains can be used to identify genomic sequences that have undergone A-to-I editing by host cell adenosine deaminases . When the conversion occurs within the replicative template of Filoviruses , the inosines are read as guanine residues resulting in T/U -> C substitutions in the negative stranded RNA genome ( for review , [53] ) . In both one-on-one and multi-genome EvoDifference prints , hyper-editing appears as clusters of T or C unique substitutions depending on whether the editing occurred in the input sequence or database genome .
Resolving sublineages during a viral outbreak or epidemic facilitates the identification of the genetic heterogeneity among viral isolates , identifies the spread of related strains to different countries , and allows for the detection of recombinant variants . Based on phylogenetic analysis , previous studies have identified major ZIKV groups: two African groups , consisting of West and East African sublineages [3] and a diverse Asian/Western hemisphere lineage ( for review , [54] ) . The West African group contains isolates primarily from Senegal and Cote-d’Ivoire , while the East African sublineages can be further resolved into isolates from Uganda and the Central African Republic . EvoDifference print readouts can be used to highlight sequence differences among related and evolutionary distant ZIKV strains ( Fig 1 ) . Using the capsid , pre-membrane and envelope encoding region from the Zika_KU321639 . 1_Brazil_2015 strain as the reference input sequence , one-on-one alignments with twenty-nine ZIKV database genomes were selected to identify 1 ) bases that are unique to the input , 2 ) bases that differed in only one of the database genomes , 3 ) sequences that differed in two or more database genomes , and 4 ) sequences shared by the input and all selected database genomes ( Fig 1A ) . Alignment details and the color-coded names of the database isolates included in the comparative analysis can be viewed by selecting line numbers ( Fig 1B ) . In this example , sequence line number 975 was expanded to highlight SNPs that are unique to the input or database genomes and shared SNPs . The expanded sequence line also highlights the greater SNP density of the more divergent African isolates ( located below the horizontal line ) when compared to the Asian isolates ( above the line ) . Database genomes are ordered by their total number of base differences when aligned to the input sequence ( least to most ) . The genome ranking and base differences are also part of the database selection page . Differentially shared SNP patterns among multiple ZIKV isolates can be used to resolve individual sublineages . For example , when 525 bases of the Zika_KF383118 . 1_Senegal_2001 NS5 coding region are used to generate an EvoDifference print with database genomes from different African sublineages , their base differences with the input Senegal isolate or SNP profiles resolve different sub-groups ( S1 Fig ) , that correspond to previously described sublineages [3 , 4 , 55–57] . Phylogenetic footprinting , identifying evolutionary conserved sequence elements using multi-genome alignment protocols , has become an important tool for resolving essential genomic information [35 , 58 , 59] . A significant advantage of EvoPrinter is the ability to rapidly change the cumulative evolutionary divergence stringency of a multi-genome comparison . By moving between the genome selection page and the EvoPrint readout , one can quickly add or remove viral strains from the analysis to reveal different levels of conservation of essential elements , as they exist within genomes of interest . For example , to identify previous characterized Ebola virus conserved transcriptional start and stop regulatory elements ( for review [60 , 61] ) , we generated a multi-genome EvoPrint of the Zaire_lin6_Kissidougou_GIN_C15_KJ660346 . 2_2014 strain that included 271 non-redundant genomes from 3 Ebola species ( 269 Zaire , 1 Bundibugyo and 1 Taï Forest ) ( Fig 2 ) . In addition to resolving transcriptional regulatory elements that flank each of the seven Ebola virus genes , the divergence stringency of the EvoPrint is sufficient to highlight essential amino acid codons by revealing their less-conserved wobble positions and identify the transcription editing site within the GP gene ( Fig 2 ) . The EvoPrint also delineates less-conserved intergenic regions and the evolutionarily variable GP mucin-like domain encoding region [62] ( Fig 2 ) . As with the Filoviruses , near-base resolution of essential information is obtained with Flaviviruses . A multi-genome EvoPrint was generated using the YellowFever_GQ379162 . 1_Peru_2007 NS3 encoding region as the input reference sequence , comparing it with 15 South American and African Yellow Fever strains selected from the Yellow Fever database ( S2 Fig ) . Together the 15 strains provide a cumulative evolutionary divergence sufficient to resolve essential bases , as evident from the less conserved codon wobble positions ( S2A Fig ) . Flavivirus SNP differences can also be accessed by expanding readout lines of multi-genome EvoPrints . The shared SNP profiles of different Yellow Fever Virus sub-groups ( S2B Fig ) . correspond to previously identified phylogenetic tree groupings [63] . Shared SNPs that highlight differences between groups of viruses serve as ancestry informative markers for identifying sublineages ( for review , [64] ) . We call these identity SNPs ( ID-SNPs ) , since they represent lineage markers for descendants of an earlier parental strain and multiple shared ID-SNPs , or profiles can be used to resolve different sublineages and illuminate ancestral relationships among ZIKV strains during spreading epidemics . Most ID-SNPs highlight differences between a sublineage and all other strains outside of the sublineage that have maintained the same ancestral base at those nucleotide positions ( Figs 3 and 4 ) . Phylogenetic tree comparisons of Asian/Oceania strains have revealed that the South American epidemic ( first identified in Brazil ) derives from a distinct sublineage that arose from an outbreak in French Polynesia in 2013 [4 , 55–57] ( for review [4 , 65 , 66] ) . Our SNP profiles of Brazilian isolates reveal that they can be further divided into at least four different subgroups based on non-overlapping ID-SNP patterns shared among 20 isolates ( Fig 3 and S3 Fig ) . For example , when the Zika_KX447510 . 1_FrenchPolynesia_2014 strain is used as the input reference genome and aligned to 13 Brazilian isolates , 3 subgroups ( Br1-3 ) ( each represented by multiple isolates ) were distinguished by 22 ID-SNPs that are positioned throughout the genome ( Fig 3 ) . When isolates from China , Ecuador , Florida , Dominican Republic , Puerto Rico , Suriname and French Guiana are included in the analysis , all five of the Florida isolates , all of the Ecuador , and two of three Dominican Republic strains share ID-SNPs with the first Brazilian subgroup ( Br1 ) but not with the Br3 subgroup ( Fig 3 ) . The second Brazil sublineage ( Br2 ) shares ID-SNPs with Florida isolates and with the Puerto Rico strains but not with Br1 or Br3 ( Fig 3 ) . The alignment also reveals that the Puerto Rico , Suriname , French Guiana and a single Dominican Republic isolate share ID-SNPs with the third Br3 Brazil subgroup but not with the Br1 sublineage . In addition , while isolates from Florida and Puerto Rico represent two distinct subgroups , the ID-SNP patterns of isolates from the Dominican Republic reveal that one isolate is related to the Puerto Rico subgroup while the other two share ID-SNPs with the Florida subgroup ( Fig 3 ) . Interestingly , pairwise alignments between the Dominican Republic isolate that is related to the Puerto Rico subgroup , the Zika_KX766028 . 1_DominicanRepublic_2016 strain , and any of the China Ch2 sublineage members reveal near identity , suggesting that the Ch2 sublineage may have originated from the Caribbean ( Fig 3 and S4 Fig ) . This possibility is further strengthened by the observation the China Ch2 strains share many ID-SNPs with isolates from Puerto Rico , Dominican Republic , Suriname , French Guinea , and members of the Brazil Br3 sublineage . In addition , these observations are in agreement with Zhang et . al . , who report the presence of highly diversified ZIKVs that have been most likely imported into China [67] . Comparative analysis of isolates from the recent southern Florida outbreak identify ancestral ID-SNPs that together suggest a progressive evolutionary divergence away from other related strains and other members of the Asian lineage . For example , an EvoDifference print of the Zika_KX832731 . 1_Florida_2016 isolate with 71 other Asian/Oceanian/Western hemisphere strains ( both related and evolutionarily distant ) revealed ID-SNPs that are shared among Florida and Dominican Republic isolates while all other strains have maintained the same ancestral base at those positions ( Fig 4 ) . Our analysis also identified ancestral ID-SNPs that are restricted to just a subset Florida and Dominican Republic strains and ID-SNPs that only distinguish a subset of Florida isolates from all other Asian lineage strains . Taken together , the different subgroups indicate that progressive , multi-generation base substitutions at different genomic positions are playing a significant role in ZIKV divergence . In addition , the multi-genome analysis demonstrated that the KX832731_Florida strain has recently acquired three unique SNPs that are not shared by any of the other Asian/Oceanian/South American strains ( two of the three unique SNPs are red highlighted in Fig 4 ) . We have also used ID-SNP profiles to search for additional Western Hemisphere sublineages by examining pair-wise alignments of South/Central American and Caribbean isolates . Our screen identified two Central American sublineages , differentiated from the Brazil Br1-4 subgroups by combinations of 15 ID-SNPs ( S3 Fig ) . These subgroups contained isolates from Mexico , Guatemala , Honduras , Panama and Columbia . Strains from Mexico fall into either the first or second central American group . Our comparative analysis also revealed that the single Martinique isolate , Zika_KU647676 . 1_Martinique_2015 , most likely originated from a Mexican strain as it differs from the Zika_KU922960 . 1_Mexico_2016 isolate by only 4 bases . To examine sublineage heterogeneity among Asian and Southeast Asian ZIKV strains , we searched for ID-SNPs that group isolates from different locations . As indicated above , our SNP pattern screen revealed two Chinese subgroups that are differentiated by 31 ID-SNPs ( Fig 3 and S4 Fig ) . Using Zika_KU955589 . 1_China_2016 as the input reference genome , our multi-genome analysis revealed that the Chinese Ch2 subgroup shares many ID-SNPs with Western hemisphere isolates , while the first China subgroup ( Ch1 ) constitutes a distinct ( perhaps older ) Asian sublineage ( Fig 3 and S4 Fig ) . The French Polynesian strains share six ID-SNPs with the Ch1 subgroup and the Tonga strain shares eight ID-SNPs , suggesting that strains from Tonga and French Polynesia may be evolutionarily positioned between the Chinese Ch1 sublineage and Western hemisphere isolates ( S4 Fig ) . Genomic diversity among Flaviviruses is driven in part by homologous recombination between related strains , with their recombinant exchanges occurring in both protein encoding and noncoding sequences [3 , 7 , 68 , 69] . Alignment programs that scan for changes in sequence homology within multiple genomes and methods that examine differential phylogenetic clustering using genomic sub-regions have been used to identify recombinants and locate approximate recombinant fragment boundaries [70 , 71] . Evoprinter screens can also identify recombinants and resolve the approximate boundaries of their recombining fragments within parental lineages . By examining a previously characterized Dengue2 recombinant , we show how SNP profiling can be used to identify recombinant strains and their parental sublineages ( S5 Fig ) . Phylogenetic tree clustering analysis of the Dengue2_AF100466 . 2_Venezuela_ 1990 ( Mara4 ) strain with other Dengue2 genomes revealed that Mara4 is the recombinant progeny of two distinct Dengue2 sublineages [71] . Differential phylogenetic clustering analysis revealed that the first ~500 bases of Mara4 are nearly identical to Dengue2 strains from Thailand , while the remaining genome is related to American strains [71] . Side-by-side EvoDifference SNP profile comparisons of the Mara4 recombinant with members of the parental sublineages ( from Thailand and Jamaica; S5A and S5B Fig , respectively ) demonstrate that the 5’ recombinant fragment originated from the minor parental Thailand sub-group ( boxed region in S5 Fig ) . Note that , by convention , the strain that produces the highest SNP density within the recombinant region when aligned to the recombinant strain is designated as the major parental lineage , while the minor parental sub-group shares identity or near identity with the recombinant within the boundaries of the recombinant fragment . In this example , the differing parental SNP pattern boundaries are located at positions 594 ( major parent ) and 600 ( minor parent ) , indicating that the recombinant exchange most likely occurred between bases 595 and 599 ( S5 Fig ) . One advantage of the genome SNP profiling is that recombinants and their parental lineages can be identified by differental SNP patterning . For example , Fig 5 identifies a novel Dengue2 recombinant strain . Side-by-side comparisons of SNP profiles generated from one-on-one EvoDifference prints of the Dengue2_GQ398269 . 1_PuertoRico_1994 strain with another Puerto Rico strain ( Dengue2_KF955363 . 1_PuertoRico_1986 ) and with a New Guinea isolate ( Dengue2_AF038403 . 1_NewGuinea_1988 ) –Fig 5A and 5B , respectively—revealed that the Puerto Rico_GQ398269 . 1 strain is the resultant progeny of a recombinant exchange between a member of a Puerto Rican subgroup ( major parental sublineage ) and a New Guinea sub-group member ( minor parental sublineage ) ( Fig 5 ) . The abrupt SNP density pattern change within the recombinant Puerto Rico/New Guinea strain alignment delineates an ~2 , 100 base region ( spanning the NS2B and NS3 protein encoding sequences ) that is identical in both the recombinant and New Guinea genomes ( Fig 5B ) . Note that the higher density SNP cluster in the Puerto Rico ( major parent ) –recombinant strain SNP profile alignment corresponds to the region of identity shared between the recombinant and the minor parental strain ( Fig 5A and 5B ) . Using SNP profiling , we have sought evidence of recombination within Asian and African ZIKV lineages . Our initial screen of the China Ch-1 sublineage isolates revealed that many are nearly identical , however , the SNP profile generated when the Zika_KU963796 . 1_China_2016 strain was aligned to Zika_KU866423 . 1_China_2016 identified two genomic regions that have significantly higher SNP densities when compared to flanking sequences ( Fig 6A ) . Further analysis that included other Asian strains revealed that when the KU866423 . 1 strain was aligned to a Cambodian isolate , Zika_JN860885 . 1_Cambodia_2010 , their genomes are identical in the same two regions that displayed higher density SNP clustering in the above KU866423 . 1—KU963796 . 1 comparison , but differ significantly in sequences flanking these regions ( Fig 6 ) . The matching genomic positions that have converse high SNP density vs . sequence identity reveals that the KU866423 . 1 strain is the recombinant progeny of the two separate parental sublineages , one from China ( major parent ) and the other from Cambodia ( minor parent ) . The one-on-one SNP pattern comparisons also revealed that the recombinant strain is the product of two genomic exchanges , with one occurring in sequences that code for NS3 , NS4A and 2K proteins , while the other in-frame exchange occurred within the 3’ end of the NS5 coding region . Notably , when members of the parental lineages are aligned , their SNP profiles do not reveal any significant changes in SNP densities that would flag these as recombinant strains ( Fig 6C ) . African lineage ZIKV recombinant strains have been described previously [3] . Consistent with these observations , EvoDifference prints of available African ZKIV strains have identified multiple one-on-one alignments that display significant changes in SNP densities within different regions of their polyprotein encoding sequences ( Fig 7 ) . For example , using the Zika_KF383119 . 1_Senegal_2001 as the input reference genome and examining other African strains , significant changes were identified in SNP densities within different genomic regions . Our initial multi-genome comparisons identified a 139-base region within the NS5 coding region that significantly differs from sequences within the original 1947 Uganda Zika forest sentinel monkey isolate and two other strains from Senegal and the Central African Republic ( S6A Fig ) . Expanding the readout lines revealed that the Uganda and Senegal isolates are identical to adjacent but non-overlapping portions of the KF383119 . 1_Senegal reference sequence , while a Central African Republic strain shares many of the sequence differences of both the Uganda and Senegal isolates ( S6B Fig ) . Examination of other African strains also revealed SNP clustering within this region and other significant changes in SNP densities outside of the NS5 coding region . Similar to the China/Cambodia recombinant , many of the high-to-low-to-high SNP density changes indicate multiple recombination exchanges have occurred within these viruses ( Fig 7 ) . For example , alignment of the KF383119 . 1_Senegal with Zika_KF383118 . 1_Senegal strains identified three additional clusters of sequence differences; most notably , a putative recombinant fragment that spans the capsid and envelope encoding sequence ( Fig 7B ) . Also note , the SNP cluster in panel A ( that spans the NS5 encoding sequence ) and the high density SNP cluster within the same genomic region shown in panel B were adjacent but non-overlapping ( also highlighted in S6B Fig . pdf ) . High density SNP clusters were also identified in an EvoDifference print of KF383118 . 1_Senegal and the LC002520 . 1_Uganda ( Fig 7C ) , with the NS5 SNP cluster expanded to include both NS5 high density SNP clusters ( Fig 7A and 7B ) . The juxtaposition of high and low SNP densities within the one-on-one comparisons highlight putative recombinant exchanges , with one of the aligning strains most likely belonging to the major parental sublineage ( Fig 7A–7C ) . An EvoDifference print of the African KF383119 . 1 strain with an evolutionarily distant African strain , Zika_KF383116 . 1_Senegal_1968 , shows extensive divergence throughout their coding sequences , with the exception of the centrally located NS3 encoding sequence ( bases 5227 to 5556 ) ( Fig 7D ) . In addition , pairwise alignments with other African strains uncovered evidence of additional African recombinant exchanges . For example , one-on-one SNP profiles of KF383119 . 1 or KF383116 . 1 strains with another highly divergent Senegal strain , Zika_KF383120 . 1_Senegal_2000 revealed multiple significant changes in SNP densities ( Fig 7E and 7F , respectively ) . Although the KF383120 . 1 strain is considered to be inactive , given the presence of an internal in-frame stop codon [3] , recent phylogenetic analysis reveals that the KF383120 . 1 strain belongs to a distinct African sublineage that includes other closely related functional strains [4] . Filovirus database genomes are grouped according to their species and lineage designations [23 , 25 , 26 , 37 , 72] . A comparative analysis of the Zaire species identified seven lineages that make up three major groups: 1 ) the Kikwit ( lin1 ) , Gabon ( lin2 ) and Mayinga ( lin3 ) isolates taken together fall into a related group; 2 ) the Ilembe ( lin4 ) , Luebo ( lin5 ) and Boende ( lin7 ) together fall into a second group , and 3 ) the recent Zaire/Makona West African isolates ( lin6 ) represent a more divergent lineage ( Fig 8 ) , in agreement with recent lineage designations [72] . Alignments of the other Ebola species revealed two Bundibugyo lineages , four Reston , six Sudan lineages and one Taï Forest ( the sole sequence in this species ) . Consistent with previous studies , EvoDifference prints identified nine Marburg lineages [22 , 73] . As an example of using SNP patterning to resolve different Filovirus lineages , we show how a multi-genome EvoDifference print of the Zaire_lin1_Kikwit_AY354458 . 1_1995 GP gene mucin-like domain encoding sequence [62] with other Zaire reference strains can identify different lineage-specific SNP patterns ( Fig 8 ) . By selecting a readout line number ( line 7425 in this case ) , sequence differences are revealed among the different aligning lineage pairs , allowing an assessment of which bases conform to the input sequence and which are different and unique to a single lineage or shared by various other lineages ( Fig 8B ) . Interestingly , most of the SNPs within the mucin domain are T/U->C substitutions and may be the result of host cell A-to-I RNA editing ( discussed below ) . Phylogenetic analyses coupled with retrospective epidemiological studies of the recent West African Ebola/Zaire outbreak revealed that the epidemic started in Guinea and spread to Sierra Leone and Liberia [28 , 30 , 74 , 75] . During its rapid spread , base substitutions were identified that distinguished between early and late isolates [74] , reviewed by [12] . To highlight the ability of EvoPrinter to identify subgroups , we illustrate how a multi-genome EvoDifference print , using the early isolate Gueckedou_GIN_C05_KJ660348 . 2_2014 genome as the input reference sequence , identified two subgroups within the Ebola/Zaire outbreak [marked by two identity SNPs at position 13 , 856 ( A->G ) and position 15 , 660 ( T->C ) ] , one represented by the Gueckedou subgroup ( Guinea-1a ) , and a larger subgroup represented by the majority of Ebola/Zaire strains ( S7A Fig ) . The accumulation of SNPs in Guinea-1a strains from Coyah and other locations illustrates the persistence of this early lineage over the course of the epidemic . The second identity SNP at position 15 , 660 ( T->C ) , reinforces the hypothesis that the Coyah isolates , plus an isolate from Liberia , are part of the same early sublineage [12 , 29 , 30 , 31] . Using a strain isolated during the later phase of the epidemic as the input reference sequence , identity SNPs were identified in the Sierra Leone-Guinea-3 sublineage [74 , 76] ( S7B Fig ) . Our analysis revealed an identity SNP at position 10218 ( A->G ) , that marks isolates from Sierra Leone and Guinea . In addition , several Sierra Leone members of this subgroup are closely related to the reference sequence , while others are more distantly related , as seen by the presence of many SNP differences with the reference genome . Many sublineage A strains , described by [31] contain both an adenosine nucleotide at position 10218 and an additional A->G substution , at position 10273 . Zaire strains with a G at position 10218 include the following: 1 ) all early sublineage Guinea-1a , 2 ) all Liberia strains , indicating their early origin during the course of the epidemic , 3 ) many Sierra Leone and Guinea isolates , both closely or distantly related to the reference sequence , and 4 ) a group of isolates from Guinea that contained an additional marker at position 10248 ( T->C ) ( S7 Fig ) . Host-cell adenosine deaminases that act on RNA ( ADARs ) modify RNAs by converting adenosine bases to inosines ( for review , [53] ) . When ADARs edit a Filovirus replicative template , the viral polymerase interprets inosines as guanines , resulting in the negative stranded RNA genome having a cytosine instead of an uracil base at the modified or edited position . ADAR editing has been detected in both Marburg and Ebola isolates ( for review , [44] ) . Whole-genome EvoDifference prints of related Marburg strains have revealed multiple T/U -> C base substitution clusters within non-coding regions and in protein encoding sequences of individual strains ( Fig 9 and S8 Fig ) . For example , the intra-lineage comparison of the Marburg_lin2_Popp_Cercopithecus_Human_Z29337 . 1_1967 strain with the Marburg_lin2_LakeVictoria_GQ433353 . 1_2011 isolate identified a cluster of T -> C base differences within the NP gene 3’UTR and flanking VP35 intragenic region ( Fig 9A ) . All 40 base differences in the 556 base non-coding region ( bases 2 , 282 through 2 , 838 ) were identified as T/U -> C substitutions in the Lake Victoria strain . Examination of Marburg strains revealed examples of putative T/U -> C base editing in the VP35 and VP40 ORFs . Our analysis identified two strains with clusters of T/U -> C within the VP35 coding sequences . The first , found in Marburg_lin3_Ang_KM2601523 . 1 is illustrated in S8A Fig . The twenty-three T/U -> C base substitutions span 289 bases , 19 of which are in the intragenic region between the NP and VP35 coding regions; four additional substitutions are within the VP35 ORF and two of these resulted in nonsynonymous amino acid changes . The second example of A to I editing within the VP35 ORF was found in Marburg_lin9_Kenya_EU500826 . 1_1987 ( S8B Fig ) . The T/U -> C substitutions resulted in 5 amino acid changes . Although overlapping in distribution , the substitutions in these lin3 and lin9 strains occurred at different positions . We also identified evidence of T/U -> C substitutions in VP40 coding sequences . Within Marburg_lin9_Kenya_EU500828 . 1 , fourteen T/U -> C base substitutions were identified ( S8C Fig ) ; 13 bases fell within the VP40 coding sequence . These substitutions resulted in four nonsynonymous amino acid changes . A second example identified in the VP40 coding sequence of Marburg_lin9_Kenya EU500826 . 1_1987 is illustrated in S8D Fig . These substitutions resulted in three amino acid changes . Our search of Ebola genomes also uncovered clusters of T/U -> C base substitutions in Zaire and Bundibugyo strains . Within three Zaire_lin6_Port Loko_2015 isolates , we found identical T/U -> C patterns in the 3’ UTR of their NP genes ( Fig 9B ) . The fact that the 3 isolates have the same T/U -> C substitutions indicates that the editing most likely occurred in a previous generation of this subgroup and was not a product of in vitro cell culture passage . In the Bundibugyo lineage , evidence of A-to-I editing was detected within the GP Mucin domain encoding sequence of the Bundibugyo_lin1_Uga_FJ217161 . 1_2008 strain ( Fig 9C ) . Prior to full genomic sequencing , the GP gene from this Bundibugyo strain was sequenced from a patient serum-derived PCR product [21] , indicating that the putative editing occurred in vivo . Of note , four of the 12 T/U -> C substitutions in the Bundibugyo_lin1_Uga_FJ217161 . 1_2008 mucin-like domain encoding sequence result in amino acid codon changes , suggesting that A-to-I editing may contribute to the antigenic diversity of the Filovirus spike proteins . The methodology and databases described here represent a new set of alignment tools for the rapid comparative analysis of a Flavivirus or Filovirus sequence . By superimposing alignment data of either one or up to hundreds of strains onto the user’s input sequence , uninterrupted readouts enable the following; 1 ) surveillance of lineage complexity within viral outbreaks , 2 ) the identification of unique base substitutions within the input sequence and/or database genomes , 3 ) the identification of recombinant strains , and 4 ) superimposed alignments highlight conserved sequence elements and allow for the identification of viral genomes that have been modified by host cell editing . EvoPrinter should not be considered a stand-alone application for the analysis of Flavi or Filovirus evolution . We recommend that it’s search algorithms be used in conjunction with other tools that employ different sets of comparative analysis stratagies . For example , while EvoPrinter resolves sublineage markers and isolate-specific SNPs , other phylogenetic analysis programs provide information concerning lineage progression and diversification ( e . g . [12–14] ) . Our strategy of detecting recombinants , using differential SNP patterns , is also complementary to other tools such as the multi-genome Recombination Detection Program that identifies recombinant fragments in graphic readouts [52] . When used together with these other tools , EvoPrinter should prove to be an important addition for the genetic surveillance of these evolving pathogens . | Flaviviruses , including Zika and Dengue viruses , and Filoviruses , including Ebola and Marburg viruses , are significant global public health threats . Genetic surveillance of viral isolates provides important insights into the origin of outbreaks , reveals lineage heterogeneity and diversification , and facilitates identification of novel recombinant strains and host cell modified viral genomes . We report the development of EvoPrinter , a web-accessed alignment tool for the rapid comparative analysis of viral genomes . EvoPrinter superimposes alignment data from multiple pairwise comparisons onto a single reference sequence of interest , to reveal both similarities and differences detected in hundreds of selected viral isolates . Evoprinter databases provide easy access to hundreds of non-redundant Flavivirus and Filovirus genomes . allowing the user to distinguish between sublineage identity SNPs and unique strain-specific SNPs , thus facilitating analysis of the history of viral diversification during an epidemic . EvoPrinter also proves useful in identifying recombinant strains and their parental lineages and detecting host-cell genomic editing . EvoPrinter should serve as a useful addition to existing tools for the comparative analysis of these viruses . | [
"Abstract",
"Introduction",
"Materials",
"and",
"methods",
"Results",
"and",
"discussion"
] | [
"medicine",
"and",
"health",
"sciences",
"pathology",
"and",
"laboratory",
"medicine",
"pathogens",
"microbiology",
"viruses",
"genomic",
"databases",
"filoviruses",
"rna",
"viruses",
"genome",
"analysis",
"microbial",
"genomics",
"research",
"and",
"analysis",
"methods",
"sequence",
"analysis",
"viral",
"genomics",
"sequence",
"alignment",
"bioinformatics",
"medical",
"microbiology",
"microbial",
"pathogens",
"biological",
"databases",
"comparative",
"genomics",
"sequence",
"databases",
"ebola",
"virus",
"flaviviruses",
"virology",
"database",
"and",
"informatics",
"methods",
"viral",
"pathogens",
"genetics",
"biology",
"and",
"life",
"sciences",
"genomics",
"computational",
"biology",
"hemorrhagic",
"fever",
"viruses",
"organisms",
"zika",
"virus"
] | 2017 | Flavivirus and Filovirus EvoPrinters: New alignment tools for the comparative analysis of viral evolution |
Several recent large clinical trials evaluated HIV vaccine candidates that were based on recombinant adenovirus serotype 5 ( rAd-5 ) vectors expressing HIV-derived antigens . These vaccines primarily elicited T-cell responses , which are known to be critical for controlling HIV infection . In the current study , we present a meta-analysis of epitope mapping data from 177 participants in three clinical trials that tested two different HIV vaccines: MRKAd-5 HIV and VRC-HIVAD014-00VP . We characterized the population-level epitope responses in these trials by generating population-based epitope maps , and also designed such maps using a large cohort of 372 naturally infected individuals . We used these maps to address several questions: ( 1 ) Are vaccine-induced responses randomly distributed across vaccine inserts , or do they cluster into immunodominant epitope hotspots ? ( 2 ) Are the immunodominance patterns observed for these two vaccines in three vaccine trials different from one another ? ( 3 ) Do vaccine-induced hotspots overlap with epitope hotspots induced by chronic natural infection with HIV-1 ? ( 4 ) Do immunodominant hotspots target evolutionarily conserved regions of the HIV genome ? ( 5 ) Can epitope prediction methods be used to identify these hotspots ? We found that vaccine responses clustered into epitope hotspots in all three vaccine trials and some of these hotspots were not observed in chronic natural infection . We also found significant differences between the immunodominance patterns generated in each trial , even comparing two trials that tested the same vaccine in different populations . Some of the vaccine-induced immunodominant hotspots were located in highly variable regions of the HIV genome , and this was more evident for the MRKAd-5 HIV vaccine . Finally , we found that epitope prediction methods can partially predict the location of vaccine-induced epitope hotspots . Our findings have implications for vaccine design and suggest a framework by which different vaccine candidates can be compared in early phases of evaluation .
The HIV epidemic is a major global health challenge leading to more than 1 . 8 million deaths annually , and despite significant efforts the search for an efficacious and safe vaccine continues . Many different formulations of candidate HIV vaccines have been proposed and tested in recent years [1] . One of the leading approaches in this field focuses on vaccines that are primarily designed to elicit CD8+ T-cell responses that have been shown to be critical for controlling HIV infection [1]–[5] . These vaccines are comprised of vectored immunogens that use a modified virus ( e . g . adenovirus or poxvirus ) from which specific HIV genes are expressed . While several adenovirus types are currently being studied including rAd-35 and rAd-26 [6]–[8] , only rAd-5 based HIV vaccines have been extensively tested to date . rAd-5 was chosen as a vaccine vector because previous work showed that it was both safe and highly immunogenic , eliciting vaccine-induced T-cell responses in 77% of the vaccinees [1] , [9] . In the current study , we analyze epitope mapping data from two candidate rAd-5 HIV immunogens that were tested in human clinical trials . The MRKAd-5 HIV-1 gag/pol/nef vaccine was a multivalent rAd-5 vaccine that contained clade B gag/pol/nef HIV inserts and was tested in both a phase I trial ( Merck16 ) [10] and a phase IIb trial ( HVTN 502/Step ) [11] . The VRC-HIVAD014-00VP was a multiclade , multivalent recombinant rAd-5 vaccine that contained a clade B gag-pol insert as well as envelope inserts from the three major HIV clades ( A , B and C ) , and was tested in a phase I trial ( HVTN 054 ) [12] . The HVTN 502/Step phase IIb trial was halted after an interim analysis showed that the tested vaccine did not reduce the rate of HIV-1 incidence nor reduce plasma viremia after infection [9] , [11] . Considerable work has been conducted to identify potential reasons for the vaccine's failure . Preliminary analysis suggested an interaction between Ad-5 neutralizing antibody ( nAb ) titers and vaccine efficacy , but subsequent analyses failed to find a significant difference [13] , [14] . A separate hypothesis was that the Merck vaccine , while highly immunogenic , induced only narrow responses generating a median of ≤1 T-cell response per participant [9] . Here , we sought to characterize the epitope responses generated by these two vaccines on a population level and used these epitope maps to address several questions: ( 1 ) Are vaccine-induced responses randomly distributed across vaccine inserts , or do they cluster into immunodominant epitope hotspots ? ( 2 ) Are the immunodominance pattern observed in these three vaccine trials ( two of which tested the same Merck vaccine product ) different from each other ? ( 3 ) Do vaccine-induced hotspots overlap with epitope hotspots induced by natural infection with HIV-1 ? ( 4 ) Do immunodominant hotspots target evolutionarily conserved regions of the HIV genome ? ( 5 ) Can epitope prediction methods be used to identify these hotspots ? We found that vaccine-induced responses tended to cluster into immunodominant epitope hotspots in all three vaccine trials , and some of these hotspots were not observed in a chronic natural infection cohort . Comparing the hotspots induced in each trial , we found statistically significant differences between the patterns induced by Merck and VRC HIV vaccines , but also between the Merck16 phase I trial and the HVTN 502/Step trial that tested the same vaccine product in different populations . Some of the immunodominant hotspots targeted were from highly variable regions of the HIV genome , and this was most evident for the Merck vaccine . In addition , we showed that epitope prediction methods can partially predict the location of epitope hotspots and presented statistical tests for subsequently comparing these hotspots across different vaccine products and trials . Taken together our findings suggest that rAd-5 vector vaccines generate a clear immunodominance pattern on a population level , with many participants targeting similar areas . Specifically , they point to a small subset of regions within the HIV vaccine inserts that are highly immunogenic . These hotspots can be characterized experimentally with a relatively small number of participants ( n<100 ) , and , where there is knowledge about the importance of the hotspots for potential vaccine protection , may be used as novel immunogenicity endpoints for comparing candidate vaccine products and regimens [15] in phase I/II vaccine trials . Specifically , coupled with recent efforts to identify regions of the HIV genome to which responses are protective and non-protective [16]–[18] , the identification of vaccine-induced epitope hotspots , may allow scoring different vaccine candidates based on their ability to generate responses to these protective regions . Furthermore , hotspots may potentially be used to define novel population-based biomarkers for assessment as immunological correlates of risk and protection in phase IIb/III efficacy trials [19] , [20] .
To assess vaccine immunogenicity , we analyzed epitope mapping data from 177 vaccine recipients from three HIV-1 rAd-5 T-cell based vaccine clinical trials: Merck16 , HVTN 054 and HVTN 502/Step , and compared the response patterns of each trial to those of 372 persons with chronic HIV infection ( Table 1 ) [3]; constituting the three largest vaccine-induced epitope mapping studies to date . Epitope mapping was performed using IFN- γ ELISpot assays with sets of overlapping peptides , as detailed in Table 1 . Responses were mapped down to the level of a single reactive K-mer peptide ( K = 9–22 ) . Specifically , high-resolution mapping was performed using 9 mers in both Merck16 and HVTN 054 , 15 mers in HVTN 502/Step , and 15–20 mers in the natural infection cohort . Our goal was to characterize HIV-1 vaccine-induced CD8+ T-cell responses on a population level . We therefore used the epitope mapping data of each cohort to create population-based epitope maps by tallying the number of responses that were observed for each position along a given HIV-1 protein . Counts were then normalized to provide population-based detection frequencies ( Figure 1 ) . Maps were generated using conservative estimates of the number of epitope responses for each individual by considering consecutive positive K-mers as a single epitope response ( see Materials and Methods ) . These maps were also used to compare immune response patterns between recipients of the different vaccines as well as to persons with natural infection with HIV-1 , focusing on the location the most immunodominant hotspots in each trial . The Merck16 and HVTN 502/Step trials used an identical rAd-5 gag/pol/nef vaccine developed by Merck Laboratories . The HVTN 054 trial administered a rAd-5 gag-pol/envA/envB/envC vaccine developed by the NIH Vaccine Research Center . While the two vaccines contained different immunogens , their gag and pol inserts were both clade-B isolates that were extremely similar to one another [10] , [12] , and both were based on an rAd-5 backbone . An epitope hotspot is typically defined as a public immunodominant region that contains several epitopes that are presented by different HLA alleles and is targeted by many individuals . In this study we defined hotspots statistically as sets of contiguous sites along a protein that were targeted more frequently than under the null hypothesis of equal targeting frequencies for all sites . Using permutation tests on the location of epitopes for each participant , we found that epitope responses in all three vaccine trials clustered in immunodominant hotspots ( p-values ranging from 0 . 0001–0 . 07 in all vaccine inserts ) ( Table 2 ) . Most vaccine inserts contained at least four statistically significant hotspots ( Figure 2 ) ; these were also observed among naturally infected persons [3] ( Figure 2 ) . We next sought to compare the location of these epitope hotspots among the groups by developing a statistical test that was based on the targeting frequency of the most dominant hotspot for each cohort ( see Materials and Methods for “targeting frequency” calculation ) . Using a permutation test , we ascertained if the maximal difference in targeting frequency was higher than expected under equal frequencies . We found significant differences in Pol between HVTN 502/Step and HVTN 054 ( p = 0 . 045 , Figure 2h , j ) and between HVTN 502/Step and Merck16 ( p = 0 . 0041 , Figure 2h , i ) . We also used a permutation test to ascertain if the sum of differences in targeting frequencies was higher than expected under equal frequencies . We found significant differences in Gag and Pol between HVTN 502/Step and HVTN 054 ( Gag: p = 0 . 045 , Figure 2a , c; Pol p = 0 . 0007 , Figure 2h , j ) , and in Nef and Pol between HVTN 502/Step and Merck16 ( Nef p = 0 . 0028 , Figure 2e , f; Pol: p = 0 . 05 , Figure 2h , i ) . We then compared the frequency of HLA alleles in these cohorts , which may bias responses towards specific epitope hotspots targeted by different HLA alleles . In comparing these distributions we found no evidence for significant differences between the three trials ( Fisher's exact test p = 0 . 43 , Table 3 , Figure 3 ) . To further address this hypothesis , we used an HLA matching strategy to identify HLA strata that are comparable between HVTN Step/502 and Merck16 ( see methods for details ) . We then recomputed the two statistical tests for hotspot differences described above accounting for HLA strata . After correcting for HLA we found no evidence for statistical differences in hotspot locations in Gag , but the differences in Nef and Pol remained significant ( Table 4 ) . This suggests that it is unlikely that all of the differences in the immunodominant hotspots observed in these two trials are a result of differences in the HLA distributions of trial participants . We compared the vaccine-induced immunodominant hotspots to those elicited from natural infection to determine how similar the T-cell responses are between these different populations . It has been previously shown that T-cell responses during natural infection are of higher magnitude and breadth than those resulting from HIV vaccination [3] , [4] , [21] , [22] , and that on a population level , almost all regions of a given HIV-1 protein are targeted by a T-cell epitopes [3] . We therefore asked two questions: ( 1 ) what is the correlation between the vaccine-induced and natural infection-induced epitope maps; and ( 2 ) are there epitope hotspots that are targeted following vaccination that are not frequently targeted during natural infection ? We first computed the Spearman correlation between vaccine-induced maps and natural infection maps . We found that natural infection response patterns to Gag were positively correlated with the patterns induced by the vaccines in HVTN 502/Step ( r = 0 . 42 , p<10−5 ) , Merck 16 ( r = 0 . 42 , p<10−5 ) and HVTN 054 ( r = 0 . 40 , p<10−5 ) ( Figure 2a–d ) . However , natural infection response patterns to Nef were not significantly correlated with response patterns for HVTN 502/Step ( r = −0 . 07 , p = 0 . 29 ) and Merck16 ( r = 0 . 10 , p = 0 . 15 ) ( Figure 2 e–g ) . Similarly , no or very weak correlations were found between the natural infection responses to Pol and those elicited by HVTN 502/Step ( r = −0 . 03 , p = 0 . 34 ) , Merck16 ( r = −0 . 07 , p = 0 . 048 ) , or HVTN 054 ( r = 0 . 069 , p = 0 . 042 ) ( Figure 2h–k ) . We then asked if any of the vaccine-induced hotspots targeted areas that were not frequently targeted in natural infection , i . e . if they introduced any novel immunodominant hotspots . We identified several hotspots in both Gag ( Figure 2a , d ) and Nef ( Figure 2e , g ) that were targeted by HVTN 502/Step vaccine recipients , but not targeted in chronic natural infection . In order to assess the relationship between targeted hotpots and evolutionary conservation , we computed HLA targeting efficiency scores for each population-based epitope map . The HLA targeting efficiency score is defined by the Spearman correlation coefficient between binding scores and conservation scores for amino acids along a given protein [23] . A positive score indicates preferential binding to conserved sites along the protein and a negative score indicates preferential binding to variable regions . In a previous study , we found that the HLA alleles preferentially targeted the conserved regions of pathogens and self proteins [23] . In accordance with those findings , the HLA targeting efficiency scores of the natural infection epitope maps were positive for both Gag and Nef ( Table 5 ) . Surprisingly , some of the efficiency scores of the vaccine induced maps were negative , indicating preferential binding to variable regions . For example , the efficiency scores of Gag were negative in all vaccine trials , but were positive in the natural infection cohort . To further characterize this phenomenon , we overlaid the conservation scores of each site of each protein over the population-based epitope maps . Indeed , we found that most epitope hotspots in both Gag and Nef contained highly variable sites ( Figure 4a–b ) . We then compared the conservation scores of hotpots vs . other sites for each of these epitope maps ( Figure 4c ) . For Merck16 , the targeted sites of Nef had significantly higher conservation scores than non-targeted , but an opposite trend was observed for HVTN 502/Step . Motivated by the need to improve sampling designs for expensive immunological endpoint experiments , which require both large quantities of PBMCs and numerous procedures , we sought to determine if HLA binding predictors can be used to identify epitope hotspots . Several recent benchmarks have shown that HLA binding predictors are highly accurate and can also be used to predict binding to HLA alleles that have not been experimentally characterized , building upon other alleles for which experimental data are available [24] , [25] . Here we developed a population-based approach in which we pool predictions for all HLA alleles into one prediction map . The approach is based on the observation of immunological hotspots targeted by many different HLA alleles . Population maps were generated by tallying the number of predicted 9 mers that straddled each position along a vaccine insert given the HLA alleles of the trial participants . Predicted binders were defined as 9 mers for which the predicted IC50 value was below a threshold δ ( δ = 50 nM , 150 nM , and 500 nM ) . We compared the measured population-based epitope maps to the prediction maps ( Figure 5a–b ) . We found that the predicted maps were significantly correlated to the measured maps for Gag , Pol and Nef but less so for Env ( r values ranging from 0 . 07 for Env to 0 . 5 for Gag , Figure 5c ) . In almost all cases the immunodominant measured peaks were identified in the prediction maps . However , predicted maps contain several additional peaks that were not detected experimentally . This is not surprising due to the fact that the predicted maps are based solely on HLA binding , which is only one determinant of immunodominance , and does not take into account proteasomal processing , epitope half-life , TAP transport and expression on cell surface , TCR avidity , and other intrinsic properties of individual T cells .
In this study , we analyzed epitope mapping data from three independent HIV vaccine clinical trials that evaluated two distinct rAd5-based vaccines , and made comparisons to a large natural infection cohort . Unlike responses in natural infection , which were measured using consensus clade B peptides and based on PBMC samples from varying timepoints following infection , vaccine-induced responses were measured using vaccine-matched peptides for HVTN Step and Merck16 and clade-B peptides ( highly similar to the vaccine insert ) for HVTN 054; in all vaccine trials , responses were assayed at a fixed timepoint following vaccination . These factors significantly increase our confidence in the vaccine-induced epitope maps , since responses were not lost due to mismatches between the tested peptides and the autologous epitope or to viral escape via mutations in and around epitopes . We presented statistical tests for identifying and comparing epitope hotspots across different vaccine products and trials . We found that both vaccines induced highly significant epitope hotspots . To our knowledge , this is the first study to utilize epitope hotspots for the comparison of different vaccine trials and products by the immunodominance patterns that they induce . We found that the immunodominance hierarchies generated by each vaccine were distinct from one another . The differences between the two Merck vaccine trials are surprising , since they used the same vaccine . We further showed that some of the vaccine-induced hotspots were not frequently observed in a large cohort of chronic naturally infected individuals . While we found no evidence for significant differences in the HLA frequencies between the three trials ( Table 3 ) , and found that some of the differences between the immunodominance patterns between HVTN 502/Step and Merck16 were still significant after HLA stratification ( Nef and Pol , Table 4 ) , there are several other potential explanations for the differences in the immunodominance patterns that were observed here . First , there is lack of power to detect the full landscape of immunodominant responses in each trial . This is partially supported by the finding that correlations between predicted and measured epitope maps were stronger for some vaccine inserts when considering only high-affinity predicted binders ( Figure 5c ) , and also by differences found between HVTN 502/Step and Merck16 . We note however that the data analyzed here included the three largest T-cell epitope mapping studies performed in HIV-1 vaccine trials , and as such are the best existing data currently available . A second potential explanation is that slight changes in the immunogen can lead to large differences in the immunodominance patterns that they induce . While both the Merck vaccine and the VRC vaccines were based on a rAd-5 backbone , they had several important differences: ( 1 ) the VRC product included a Gag-Pol fusion in a single insert and the Merck product contained a separate vector for Gag and Pol; ( 2 ) the VRC vaccine also included Env inserts which could have provided epitope competition for MHC binding; and ( 3 ) vector design – the GenVec rAd5 used in HVTN 054 was E4 and E3 partially deleted , and these regions were not deleted in the Merck rAd5 vector . E4 is required to produce Ad structural proteins such as hexon , which can activate cell signaling processes that can affect the proteasome and the “inflammasome . ” Accordingly , the most immunodominant hotspot in HVTN 502/Step ( targeted by more than 40% of participants ) was not observed in HVTN 054 . A similar effect , albeit for antibodies , was recently reported for the RV144 vaccine regimen in which the replacement of the C-terminus of gp120 with a gD tag modified the antibody immunogenicity pattern induced by this immunogen [26] . Another potential explanation is differences in the epitope mapping strategies used in each study . While Merck16 and HVTN 502/Step were mapped using vaccine matched peptide sets , HVTN 054 and the natural infection cohort were mapped using consensus B peptides . Merck16 was mapped with vaccine matched 9 mers in the Merck laboratories , while HVTN 502/Step and HVTN 054 were mapped with 15 mers in the HVTN laboratory , and the natural infection cohort was mapped with 15–20 mers in a third laboratory . The differences observed between epitope hotspots in natural infection to those induced by vaccination could also be due to changes in the immunodominance patterns between acute and chronic infection [5] . Since some of the vaccine-induced hotspots were in more variable regions of the HIV genome , they may not be chronic immunodominant hotspots due to T-cell escape . It may therefore be important to compare responses in acute infection to vaccine-induced responses . While it is impossible to tease out which of these factors ( or combination thereof ) were responsible for these differences , we believe they highlight the importance of conducting additional studies to unravel the underlying factors that influence the immunodominance patterns in a vaccine setting . We note that the epitope maps described in this paper were based on epitope prevalence and not on the actual magnitudes of the T-cell responses as measured by ELISPOT . Therefore , some of these may be hotspots of low-magnitude responses . We found that the average ELISPOT response measured in Step was 432 SFC/M . Furthermore , the average response of the two most prevalent peaks in HVTN 502/Step Gag ( Table 2 ) was 534 SFC/M and 593 SFC/M , accordingly . This suggests that prevalent hotspots were also magnitude hotspots . An analysis of the evolutionary conservation of vaccine-induced hotspots showed that some hotspots were directed against highly variable sites in the HIV genome , in which the virus can readily tolerate a variety of mutations that may allow escape from immune recognition ( see also [27] ) . A consequence of this finding is that breakthrough infections in these trials are likely to lead to early post-acquisition mutations which do not incur significant fitness cost . Indeed , sieve analysis of breakthrough infections in the HVTN Step trial found evidence for T-cell sieve effects in both Gag and Nef [28] . Finally , we showed that epitope prediction methods can be used to predict the location of epitope hotspots in vaccine trials . Importantly , predicted epitope maps tended to overshoot – predicting additional hotspots that were not seen in the empirical epitope mapping , and only rarely missed a measured hotspot . This suggests two uses of epitope prediction methods for potentially improving epitope mapping protocols in clinical trials . First , given the HLA frequencies of the target population , prediction models can be used to identity epitope hotspots that can contribute to the scoring and comparing of candidate vaccines as outlined below . Second , binding predictors may potentially be incorporated into the epitope mapping protocol itself , allowing a more focused investigation of epitopes that is tailored for each individual based on their HLA alleles , thereby reducing the number of tests required for epitope mapping . However , while the correlations between predicted and experimental epitope maps were encouraging , additional research and validation studies are needed to develop new epitope mapping algorithms that combine epitope predictions with direct epitope measurements; such algorithms should be shown to be at least as specific and sensitive as current epitope mapping protocols before they merit use . Taken together , our findings demonstrate that vaccines can generate a clear immunodominance pattern on a population level . Specifically , they suggest that by fine-mapping the immune responses in early Phase I or Phase IIa trials one may obtain a complete set of the regions that are likely to be targeted by the specific vaccine candidate , and those regions can then be further analyzed in terms of their functional importance , evolutionary conservation and any other biological property of interest to determine if targeting these regions is likely to provide any functional effect on HIV acquisition or replication capacity . A recent report comparing T-cell responses to Gag in HIV controllers vs . non-controllers , excluding individuals with protective HLA alleles , found that while the breadth and magnitude of responses in both groups were comparable , responses in the controllers were more cross-reactive and of higher avidity than those in the non-controllers [17] . Another study identified peptides that had a “protective ratio” by comparing the viral loads of responders and non-responders to each peptide [16] . Similarly , Dinges et al . reported that T-cell responses were better predictors of HIV disease progression than HLA alleles [18] . These data point to the possibility of defining an importance function that can be used to weight different positions within a vaccine insert . Combining such a weighting function with experimental epitope mapping data can provide a powerful tool to assess and compare different candidate HIV vaccines in early stages of their development [15] , [19] , [20] .
We analyzed data from three HIV-1 vaccine clinical trials that administered immuonogens based on a replication defective rAd-5 vaccine vector into which several HIV proteins were inserted . We also analyzed data from a natural infection HIV-1 cohort . Merck16 – a phase I trial of the MRK rAd-5 HIV-1 gag/pol/nef vaccine developed by Merck Research Laboratories that enrolled 259 participants [10] , [27] . The vaccine was a 1∶1∶1 mixture of rAd-5 constructs containing HIV-1 clade B gag , pol , and nef which were inserted into the E1 region of the rAd-5 backbone . HVTN 054 – a phase I trial of an Ad5 Gag/Pol/EnvA/EnvB/EnvC vaccine developed by the Vaccine Research Center that enrolled 48 participants . The product contained a mixture of 3∶1∶1∶1 E1- , partially E3- , and E4- deleted rAd5 constructs expressing a gag-pol fusion gene from HIV-1 subtype B and env genes ( from subtypes A , B , and C ) from the E1 region of the rAd-5 backbone [12] . Epitope mapping was performed on samples obtained 4 or 12 weeks after vaccination . HVTN 502/Step – a phase IIb trial of the MRK rAd-5 gag/pol/nef vaccine given at months 0 , 1 and 6 that enrolled 3 , 000 participants . The trial was unblinded after an interim analysis found that vaccine recipients had a higher risk of infection as compared to placebo recipients [11] . Epitope mapping was performed on samples obtained 4 weeks after the second vaccination . Natural infection cohort – 372 HIV-1 clade B chronically infected subjects were recruited from four hospitals in the Boston area and at the Queen Elizabeth Hospital in Barbados , as previously described [3] . Briefly , this cohort included predominantly chronically infected participants , some of which were undergoing anti-retroviral treatment . All data analyzed in this study were de-identified , and the study was approved by the HVTN review committee . Epitope mapping was performed using a group testing approach [2] , [29] in which T-cell responses are measured using an IFN-γ ELISPOT assay as previously described . Briefly , peptides representing Gag , Pol , Nef and Env were tested in pools , and peptides contained in positive pools were further tested individually . Responses to individual peptides were considered positive if they were at least threefold above the average of at least six negative control wells ( containing the peptide diluent ) and ≥50 spot forming cells ( SFC ) /106 PBMC . Responses were first measured to master or matrix pools of overlapping 9 mer ( Merck16 ) , 15 mer ( HVTN 054 , HVTN 502/Step ) or 15–20 mer ( natural infection cohort ) peptides , each containing 40–100 peptides that span the vaccine immunogens . Positive responses were then further tested using minipools containing 5–10 peptides . The reactive 9 mers/15 mers were identified by testing each individual 15 mer from all reactive pools . Epitope mapping of Merck16 was performed by Merck Laboratories and included 72 participants . Responses were mapped using vaccine-matched 9 mer peptides that spanned all three immunogens with consecutive peptides overlapping by 5 amino acids . Positive responses were defined as responses that were over 50 SFC and three times higher than background responses , as relatively high backgrounds were observed in this study ( see [27] for further details ) . Epitope mapping of HVTN 054 was performed by the HVTN laboratory and included 29 participants . Responses were measured using 15 mer peptides that spanned a consensus HIV-1 clade B ( conB ) strain that closely resembles but does not match the vaccine strain . Consecutive peptides had an overlap of 11 amino acids . Positive pool responses were defined as responses that were over 50 SFC and were two times higher than background . Positive 15 mer responses were further de-convoluted to identify the optimal epitope , based on sample availability . Epitope mapping of HVTN 502/Step was performed by the HVTN laboratory and included 71 participants . Responses were measured using vaccine matched 15 mer peptides with an overlap of 11 amino acids between each consecutive pair of peptides . Positive responses were defined as responses that were over 50 SFC and were three times higher than background . Epitope mapping of the natural infection cohort ( n = 372 ) was performed across the entire HIV-1 genome using conB overlapping 15–20 mers , with an overlap of 10 amino acids between adjacent peptides . Positive responses were defined as responses that were 4 times background levels and higher than 50 SFC . Four-digit HLA class I typing was performed on all trial participants for whom we had epitope mapping data . HLA typing was not available for the natural infection cohort . Figure 3 presents the distributions of HLA class-I alleles for the participants within these trials , and includes a statistical comparison of these distributions . Population-based epitope maps were generated for each vaccine insert by tallying the number of reactive 15 mers that included a given site across all study participants . The frequency of response was calculated by dividing the number of responses at a given site by the number of individuals who had any positive response to the given vaccine insert . In order to obtain a conservative estimate of the response frequencies , given two consecutive positive 15 mer responses for a participant , we counted these as a single epitope by only tallying sites that were part of the overlap of the two peptides ( typically 11 amino acids ) . Similarly , three and four consecutive peptides were considered as two distinct epitope responses , and five consecutive responses were counted as three distinct epitopes . We also generated predicted population-based epitope maps , which were based on using HLA binding predictors . Specifically , we used the ADT algorithm , a structure-based epitope prediction method to predict epitopes for the HLA alleles of each vaccine cohort [30] . For these maps , we only considered predicted 9 mer epitopes . Similar to the experimentally measured maps , we identified all predicted epitopes for each vaccine insert and then tallied the number of reactive 9 mers that were predicted for each site along the protein , normalizing by the number of individuals in the cohort . For each clinical trial , predicted population-based maps were weighted by the HLA distribution of the trial participants . HLA binders were defined using a binding threshold on the IC50 value . We created maps using 3 thresholds: conservative ( 50 nM , strong binders ) , moderate ( 100 nM ) and permissive ( 500 nM , weak binders ) . Epitope hotspots were defined as sets of contiguous sites that were targeted more frequently than under the null hypothesis of equal targeting frequencies for all sites . To assess the significance of such hotspots , we compared the targeting frequency of experimentally measured hotspots to those obtained from a null distribution in which the same number of epitopes were drawn at random from a uniform distribution for a given vaccine insert . Using 10 , 000 random realizations , p-values were computed for each hotspot as follows: hotspots were sorted in decreasing order of frequency and for each hotspot we computed the probability of obtaining a hotspot with similar or higher frequency under the null . We compared the first hotspot ( highest peak ) to the corresponding first hotspot in the random maps . Subsequent comparisons were done for all other peaks in descending frequency . Two testing procedures were developed to test whether two epitope maps differ . These procedures compare the difference of two experimentally measured maps to those obtained from maps obtained by randomizing the cohort assignment of participants . If trial 1 had n1 participants and trial 2 had n2 participants , we randomly divided the n1+ n2 participants into two sets of size n1 and n2 and computed a population epitope map for these two randomly assigned sets . The first test statistic is based on the maximal difference in targeting frequency between two maps . Specifically , we computed the maximal difference between random maps for 10 , 000 pairs of maps and calculate p-values by comparing the frequency of obtaining maximal difference in frequency that was equal or larger than that obtained between the two observed maps . The second test statistic was based on the sum of absolute differences between two maps . Specifically , for each pair of maps we computed the absolute sum of differences in frequency . P-values are computed by comparing the set of differences between 10 , 000 pairs of random maps to the one obtained from the experimentally measured maps . Conservation of sites across the HIV genome was computed using Shannon Entropy . Specifically , we used the LANL HIV entropy scores ( computed using http://www . hiv . lanl . gov/content/sequence/NEWALIGN/align . html ) . Sites with low entropy are highly conserved , so negative scores were used for computing correlations with conservation , and for visualization . The HLA targeting efficiency score is the Spearman correlation coefficient between binding scores and conservation scores for amino acids along a given protein . In Hertz et al . [23] , these scores were computed for each HLA separately , and were based on predicted epitopes . Here , we computed vaccine targeting efficiency scores , which compute the correlation between experimentally measured population-based epitope maps and evolutionary conservation . A positive score indicates preferential targeting of conserved regions , and a negative score indicates preferential targeting of variable regions . To compare the conservation scores of epitope hotpots vs . non-targeted sites , we defined epitope hotspots as sites that were targeted by more than 15% of the participants that had an epitope response to the given protein . Non-targeted sites were sites for which no epitope responses were detected . | The HIV epidemic is a major global health challenge leading to more than 1 . 8 million deaths annually , and despite significant efforts , the search for an efficacious and safe vaccine continues . Several candidate vaccines were designed to elicit CD8+ T-cell responses and were based on using recombinant Adenovirus serotype 5 ( rAd-5 ) vector that expresses HIV-derived antigens . While none of these vaccines had protective effects , they provide an opportunity to study vaccine-induced T-cell responses on a population level . Here , we analyze data from the three largest epitope mapping studies performed in three clinical trials testing two rAd-5 vaccines . We find that vaccine-induced responses tend to cluster in “epitope hotspots” and that these hotspots are different for each vaccine and more surprisingly in two different vaccine trials testing the same vaccine . We also compared vaccine-induced hotspots to those elicited by natural infection and found that some of the vaccine-induced hotspots are not observed in natural infection . Finally , we show that epitope prediction methods can be useful for predicting vaccine induced hotspots based on participants HLA alleles . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"medicine",
"clinical",
"research",
"design",
"clinical",
"immunology",
"clinical",
"trials",
"meta-analyses",
"immune",
"response"
] | 2013 | HIV-1 Vaccine-Induced T-Cell Reponses Cluster in Epitope Hotspots that Differ from Those Induced in Natural Infection with HIV-1 |
Analyses of gene expression profiles in evolutionarily diverse organisms have revealed a role for microRNAs in tuning tissue-specific gene expression . Here , we show that the relatively abundant and constitutively expressed miR-58 family of microRNAs sharply defines the tissue-specific expression of the broadly transcribed gene encoding PMK-2 p38 MAPK in Caenorhabditis elegans . Whereas PMK-2 functions redundantly with PMK-1 in the nervous system to regulate neuronal development and behavioral responses to pathogenic bacteria , the miR-58 , miR-80 , miR-81 , and miR-82 microRNAs function redundantly to destabilize pmk-2 mRNA in non-neuronal cells with switch-like potency . Our data suggest a role for the miR-58 family in the establishment of neuronal-specific gene expression in C . elegans , and support a more general role for microRNAs in the establishment of tissue-specific gene expression .
Since the initial genetic identification of microRNAs in Caenorhabditis elegans [1 , 2] , biochemical cloning methods and computational approaches have identified hundreds of microRNAs [3 , 4] , though genetic analysis has defined functional roles for relatively few of these [5 , 6] . A single microRNA , miR-58 , constitutes nearly half of all microRNAs in C . elegans , with constitutive expression in non-neuronal tissues through all developmental stages [7 , 8] . Differential RNA binding protein ( RBP ) immunoprecipitation with subsequent mRNA and protein quantification analyses ( RIP-chip-SRM ) has indicated the presence of hundreds of miR-58 targets [9] . Whereas deletion of miR-58 does not cause any apparent defects , a strain carrying deletions in the miR-58 family , comprised of miR-58 and the homologous microRNAs miR-80 , miR-81 , and miR-82 , exhibits multiple mutant phenotypes , including defects in size , locomotion , and reproductive egg-laying [6] . Three C . elegans genes with homology to mammalian p38 MAPK—pmk-1 , pmk-2 , and pmk-3—are in a single operon along with an additional upstream gene , islo-1 ( Fig . 1A ) . PMK-1 and PMK-2 are highly homologous , sharing a 62% amino acid sequence identity and have the signature TGY motif found in the activation loop of p38 MAPKs [10] . PMK-1 regulates innate immunity in the intestine of C . elegans and is activated by a MAPK signaling cassette composed of p38 MAPK kinase SEK-1 and the MAPKKK NSY-1 , homologous to mammalian MKK3/6 and ASK1 , respectively [11 , 12] . Functioning upstream of NSY-1 and required for activation of PMK-1 in C . elegans is TIR-1 , a conserved Toll-Interleukin-1 Receptor domain adaptor protein orthologous to mammalian SARM [13 , 14] . TIR-1-NSY-1-SEK-1 functions in the nervous system to regulate the specification of neuronal asymmetry in the AWC neuron pair [15–17] , reproductive egg-laying behavior , and the upregulation of serotonin biosynthesis in the ADF chemosensory neurons in response to infection by Pseudomonas aeruginosa [12] , but the MAPK targeted in the nervous system for these processes has not been defined , with pmk-1 loss-of-function not affecting these neuronal phenotypes . Here , we show that PMK-2 functions redundantly with PMK-1 in the nervous system of C . elegans to regulate development and behavioral responses to pathogenic bacteria , whereas PMK-1 alone functions in the intestine to regulate innate immunity . We observe distinct tissue expression patterns for the co-operonic pmk-1 and pmk-2 genes; in contrast to the ubiquitous expression pattern of PMK-1 , PMK-2 is largely restricted to the nervous system . Tissue-specific expression of PMK-2 is established by the miR-58 family , which switches off expression of PMK-2 in non-neuronal tissues . Our data suggest a role for the relatively abundant miR-58 microRNA in the establishment of tissue-specific gene expression in C . elegans .
We used mutant alleles of pmk-1 and pmk-2 to confirm that PMK-1 alone is required for expression of an intestinal reporter for p38 MAPK activity and innate immunity to infection by P . aeruginosa in the intestine . The agIs219 reporter transgene contains the green fluorescent protein ( GFP ) gene fused to the promoter of the PMK-1-regulated gene T24B8 . 5 , which is predicted to encode a peptide homologous to ShK toxin peptides , and serves as an in vivo readout of p38 MAPK activity in the intestine [12] . Expression of agIs219 in the intestine is extremely diminished in the pmk-1 ( km25 ) mutant ( Fig . 1B ) . In contrast , expression of agIs219 is unchanged in pmk-2 ( qd284 ) and pmk-2 ( qd287 ) mutant animals ( Fig . 1B ) . Similarly , in contrast to the enhanced susceptibility to P . aeruginosa observed in pmk-1 ( km25 ) mutant animals , pmk-2 ( qd284 ) and pmk-2 ( qd287 ) mutant animals display a normal innate immune response to P . aeruginosa ( Fig . 1C ) . To evaluate the roles of PMK-1 and PMK-2 in mediating the activities of the TIR-1-NSY-1-SEK-1 signaling module in the nervous system , we utilized two assays of neuronal signaling processes that are dependent on TIR-1-NSY-1-SEK-1 . First , the establishment of asymmetry in the AWC neurons during development is a stochastic process for which expression of a transgenic reporter , str-2::GFP , in one AWC neuron or the other serves as a readout [17] . TIR-1-NSY-1-SEK-1 signaling represses the expression of str-2::GFP in one AWC neuron such that in a wild type animal , only one AWC neuron expresses str-2::GFP . Strains carrying loss-of-function mutations in the TIR-1-NSY-1-SEK-1 signaling module express str-2::GFP in both AWC neurons [15–17] . We observed expression of str-2::GFP in only one AWC neuron in pmk-1 and pmk-2 loss-of-function single mutants ( Fig . 1D ) . In contrast , using two strains carrying loss-of-function mutations in both genes—pmk-2 ( qd279 qd171 ) pmk-1 ( km25 ) and pmk-2 ( qd280 qd171 ) pmk-1 ( km25 ) —we observed expression of str-2::GFP in both AWC neurons , similar to what is observed in the sek-1 ( km4 ) mutant ( Fig . 1D ) . A second process that is dependent on TIR-1-NSY-1-SEK-1 activity in the nervous system is the increased expression of tph-1 , which encodes the serotonin biosynthetic enzyme tryptophan hydroxylase , in the ADF chemosensory neurons upon exposure of C . elegans to pathogenic P . aeruginosa [12 , 18] . Upregulation of serotonin levels in the ADF neuron pair has been implicated in aversive learning behavior to pathogenic bacteria [18] . We observed that PMK-1 and PMK-2 also function redundantly in the P . aeruginosa-induced expression of a Ptph-1::GFP transgene reporter in the ADF neurons ( Fig . 1E ) . These data establish that while PMK-1 function alone is required to regulate intestinal innate immunity , PMK-1 and PMK-2 function redundantly in the nervous system downstream of the TIR-1-NSY-SEK-1 signaling module in neuronal developmental and pathogen-dependent responses ( Fig . 1F ) . We reasoned that the genetic redundancy of pmk-1 and pmk-2 in neurons , but not in the intestine , might be the result of differences in tissue expression of these genes . We proceeded to examine the tissue expression patterns for PMK-1 and PMK-2 by constructing a translational reporter for the pmk operon , qdEx101 , consisting of upstream promoter sequence and the entire length of the operon with a GFP tag engineered onto the C-terminus of PMK-2 and a mCherry tag engineered onto the C-terminus of PMK-1 ( Fig . 2A ) . We observed broad expression of PMK-1 in multiple tissue types , whereas expression of PMK-2 was mostly restricted to the nervous system; PMK-2 expression was detected in head , body , and tail ganglia as well as the nerve ring and ventral nerve cord ( Fig . 2B ) . We also observed faint and diffuse expression of PMK-2 in the distal tip cell and spermatheca at levels much lower than observed in the nervous system . Importantly , PMK-2 expression was excluded from the intestine , where PMK-1 functions solely in innate immunity . These data suggest that the tissue-specific genetic redundancy of p38 MAPK signaling in C . elegans is a consequence of distinct tissue expression patterns of the co-transcribed pmk-1 and pmk-2 genes and implicate post-transcriptional mechanisms in the tissue-specific regulation of pmk-2 expression . Insight into the post-transcriptional regulatory mechanism underlying the restricted tissue expression of PMK-2 came from a genetic screen for suppressors of the immunocompromised phenotype of a pmk-1 deletion mutant [19] , in which we serendipitously isolated a gain-of-function mutant of pmk-2 , qd171 , containing a 913 bp insertion/184 bp deletion located in the 3’UTR of pmk-2 ( Fig . 1A ) . The starting strain for the screen carried the km25 deletion allele of pmk-1 ( Fig . 1A ) and the agIs219 integrated GFP reporter transgene [12 , 19] ( Fig . 3A ) . The pmk-2 ( qd171 ) deletion suppressed the diminished intestinal GFP expression from the agIs219 reporter transgene ( Fig . 3A ) and the enhanced pathogen susceptibility of the pmk-1 ( km25 ) mutant ( Fig . 3B ) . RNAi of sek-1 and RNAi of pmk-2 , reverted the pmk-2 ( qd171 ) pmk-1 ( km25 ) pathogen resistance and agIs219 intestinal GFP expression phenotypes associated with suppression of pmk-1 loss-of-function ( Fig . 3A-3B ) . These data suggest that the ability of the pmk-2 ( qd171 ) mutation to suppress pmk-1 loss-of-function is dependent on PMK-2 . Because the location of the qd171 insertion/deletion in the 3’UTR of pmk-2 might be anticipated to influence mRNA stability , we proceeded to measure levels of pmk-2 mRNA in the pmk-2 ( qd171 ) pmk-1 ( km25 ) mutant . We detected a 5 . 7-fold increase in pmk-2 mRNA levels in the pmk-2 ( qd171 ) pmk-1 ( km25 ) mutant compared to wild type worms ( Fig . 3C ) . The effect of increased mRNA levels due to the qd171 mutation is specific to pmk-2 and not a general characteristic of the pmk operon , as pmk-3 mRNA levels did not change in the pmk-2 ( qd171 ) pmk-1 ( km25 ) mutant relative to wild type . We next determined levels of activated PMK-2 protein in the pmk-2 ( qd171 ) pmk-1 ( km25 ) mutant using an antibody that recognizes the dually phosphorylated TGY motif in the activation domain of activated mammalian p38 MAPK and that is cross-reactive with C . elegans PMK-1 and PMK-2 [11] . Corroborating the increase in mRNA levels , we detected at least a 5 . 7-fold increase in activated PMK-2 protein levels in the pmk-2 ( qd171 ) pmk-1 ( km25 ) mutant ( Fig . 3D ) . To verify that the increase in levels of pmk-2 mRNA and activated protein was due to the 3’UTR deletion of pmk-2 in the pmk-2 ( qd171 ) pmk-1 ( km25 ) mutant , we used CRISPR-Cas-9-mediated genome editing to engineer the wild type N2 strain to carry a deletion in the 3’UTR of pmk-2 . We obtained a 144 bp deletion , qd305 , in the 3’UTR of pmk-2 ( S1A Fig . ) . We proceeded to measure levels of pmk-2 mRNA and activated protein in the pmk-2 ( qd305 ) mutant and detected a 7 . 4-fold increase in pmk-2 mRNA levels compared to wild type worms ( S1B Fig . ) and a corresponding increase in activated PMK-2 protein levels ( S1C Fig . ) . Taken together , these data suggest that a deletion in the 3’UTR of the gene encoding PMK-2 p38 MAPK can suppress the immunodeficient phenotype conferred by loss-of-function of pmk-1 through increased stability of pmk-2 mRNA and levels of activated PMK-2 protein . We next sought to define the cis-regulatory determinants of the pmk-2 3’UTR that function to repress PMK-2 expression . The region deleted in the qd171 allele of pmk-2 contains the most distal polyadenylation signal ( PAS ) used in 3’-end formation of pmk-2 mRNA . We performed 3’ RACE on pmk-2 mRNA isolated from wild type and pmk-2 ( qd171 ) pmk-1 ( km25 ) mutant animals to determine the effect of the qd171 mutation on the length of the 3’UTR of pmk-2 mRNA . Sequencing of the 3’ RACE products revealed use of a more proximal PAS leading to a 206 bp truncation in the 3’UTR of pmk-2 mRNA in the pmk-2 ( qd171 ) pmk-1 ( km25 ) mutant ( Fig . 4A ) . We examined the truncated region for cis-regulatory elements conserved among Caenorhabditis species and identified three seed match sites for the miR-58/80-82 family of microRNAs residing in this region absent from the 3’UTR of pmk-2 mRNA in the pmk-2 ( qd171 ) pmk-1 ( km25 ) mutant ( Fig . 4A ) . These miR-58/80-82 seed match sites are also absent in the pmk-2 ( qd305 ) mutant ( S1A Fig . ) . The miR-58/80-82 family consists of miR-58 , miR-80 , miR-81 , miR-82 , and miR-1834 . Mutants carrying deletions in mir-58 , mir-80 , and mir-81-82 [5] were used to assess whether the miR-58/80-82 family functions to repress the expression of pmk-2 . Loss of any individual mir-58/80-82 family member had no effect on pmk-2 mRNA levels relative to wild type ( Fig . 4B ) . However , loss of both mir-58 and mir-80 resulted in a 3-fold increase in pmk-2 mRNA levels relative to wild type , and loss of mir-58/80-82 led to an even further increase in pmk-2 mRNA to a level 6 . 3-fold greater than wild type ( Fig . 4B ) , without altering levels of pmk-3 mRNA . Corroborating the mRNA analysis , we observed at least similar increases in activated PMK-2 protein levels in both mir-80; mir-58 and mir-80; mir-58; mir-81-82 mutants , but not in other mutants ( Fig . 4C ) . These data suggest that the miR-58/80-82 family acts redundantly to destabilize pmk-2 mRNA . The observation that the pmk-2 ( qd171 ) mutation suppresses the loss of pmk-1 function in the intestine suggested that not only are PMK-2 levels increased in the pmk-2 ( qd171 ) pmk-1 ( km25 ) mutant , but that the site of expression had also changed . To determine the spatial effect on expression of releasing pmk-2 from regulation by the miR-58/80-82 family , we engineered a new pmk operon translational reporter , qdEx102 , which carries mutations in the second and third miR-58/80-82 seed match sites [20] in the 3’UTR of pmk-2 ( Fig . 4A ) . Expression of PMK-1 was unchanged , whereas expression of PMK-2 was detected in many additional cell types including the intestine , body wall muscle , pharyngeal muscle , hypodermis , and vulva ( Fig . 2C ) . Expression of PMK-2 in the distal tip cell and spermatheca , which was faint and diffuse with miR-58/80-82 regulation intact ( Fig . 2B ) , was markedly elevated ( Fig . 2C ) . Similar misexpression of pmk-2 was observed when we crossed qdEx101 ( pmk operon reporter with miR-58/80-82 seed match sites intact ) into the mir-80; mir-58; mir-81-82 mutant ( Fig . 2D ) . These data suggest that the miR-58/80-82 family of microRNAs restricts expression of PMK-2 p38 MAPK through the post-transcriptional destabilization of pmk-2 mRNA in non-neuronal tissues ( Fig . 4D ) .
Our data on the tissue-specific genetic redundancy of p38 MAPK signaling in C . elegans define a role for the relatively abundant and constitutively expressed miR-58/80-82 family of microRNAs in establishing the tissue-specific expression of PMK-2 p38 MAPK . We show that although the pmk-2 gene is widely transcribed , PMK-2 protein is found nearly exclusively in the nervous system of C . elegans where PMK-2 functions redundantly with PMK-1 to regulate neuronal development and behavioral responses to pathogenic bacteria . We determined that the tissue-specific expression of PMK-2 is dependent on cis-regulatory sequences found within its 3’UTR and demonstrated that the miR-58/80-82 family is required to switch off expression of PMK-2 in non-neuronal tissues post-transcriptionally , thereby establishing its tissue-specific expression in C . elegans . Genome wide analyses of microRNA expression in C . elegans using transgenes containing the putative promoters of microRNAs fused to the gene encoding green fluorescent protein have been reported previously [8 , 21] . mir-58 was inferred to be expressed in the intestine , hypodermis , pharynx , spermatheca , excretory canal , and excretory cell soma [8] . Expression of mir-58 was not observed in the nervous system . mir-80 was shown to be expressed in the posterior intestine , head and body wall muscle , uterus , vulva , distal tip cells , excretory cells , dorsal nerve cord , and amphid neurons; mir-81 was shown to be expressed weakly in head neurons; and mir-82 was observed to be expressed in pharyngeal muscle , spermatheca , and a subset of both the ventral nerve cord and the amphid neurons [21] . The tissues in which these microRNAs are reportedly present are consistent with the absence of PMK-2 expression we observe in these tissues when miR-58/80-82 regulation is intact ( Fig . 2B ) . Additionally , these data are consistent with the tissues that misexpress PMK-2 when miR-58/80-82 regulation is disabled ( Fig . 2C-2D ) . Taken together with our data , these reported expression data on the miR-58/80-82 family suggest that the tissue-specific expression of PMK-2 is a consequence of distinct tissue-expression patterns of the corresponding microRNAs that target pmk-2 . MicroRNAs have been implicated to function both as a backup to reinforce transcriptional gene programs and as an instructive signal to shape gene expression patterns [22] . Expression analyses of microRNAs and their putative target transcripts in Drosophila and mammals showed largely nonoverlapping expression patterns , suggesting that microRNAs function secondarily to transcriptional mechanisms to reinforce promoter-defined spatial expression patterns of their targets at the post-transcriptional level [23–25] . Studies in zebrafish revealed overlapping expression patterns and observed that levels of transcripts are often lower in cells expressing the microRNA , leading to the hypothesis that microRNAs may also play a more active role in conjunction with transcriptional mechanisms to shape gene expression patterns of their targets [26] . Supporting this hypothesis is the genetic analysis of muscle microRNAs miR-1 and miR-133 in zebrafish , which were shown to refine the expression levels of their target transcripts in muscle [27] . We show nonoverlapping expression of PMK-2 and miR-58/80-82 , established not by the promoter of these genes , but rather complete destabilization of pmk-2 mRNA by miR-58/80-82 , suggesting that the activity of microRNAs can define specific patterns of gene expression in different cell types . Promoter and enhancer elements commonly direct tissue-specific gene expression in animals , and thus a role for miR-58 family-mediated post-transcriptional regulation in establishing the tissue expression pattern of pmk-2 might be somewhat unexpected . For pmk-2 , microRNA-mediated regulation may in part be a consequence of the co-transcription of both pmk-2 and pmk-1 from the same operon . Genes co-transcribed in eukaryotic operons need not share a similarity in function and are generally not thought to be expressed in specific tissues [28] . Our data suggest that microRNA-mediated regulation may contribute to differential tissue expression patterns of genes co-transcribed from the same operon in C . elegans . MicroRNAs have been shown to establish cell fate through specific cell type expression during C . elegans development [29–31] . For example , the lsy-6 microRNA regulates the development of left/right asymmetry through the specific repression of its target in one of two ASE chemosensory neurons [29] . In contrast to the highly selective expression and function of the lsy-6 microRNA in a pair of chemosensory neurons , the miR-58 family functions in a large number of cells and tissues to restrict expression of PMK-2 in non-neuronal tissues at all developmental stages . The systematic identification of microRNAs of C . elegans by deep sequencing determined that miR-58 is the most abundant microRNA , accounting for nearly half of all microRNAs present in C . elegans at all developmental stages [7] . The abundance of miR-58 , along with the presence of multiple miR-58 binding sites in the 3’UTR of pmk-2 , may contribute to the large increase of pmk-2 mRNA and protein levels when miR-58 family regulation is inhibited . The approximately six-fold difference we observe in pmk-2 mRNA levels and comparable difference in PMK-2 protein levels reflect quantitation from whole worm lysates in which PMK-2 is detectably expressed in the nervous system . Considering this and the lack of PMK-2 protein observed in non-neuronal tissues when miR-58 targeting is intact ( i . e . wild type conditions ) , the magnitude of PMK-2 repression in non-neuronal tissues is likely much greater than six-fold . The switch-like “off” state of PMK-2 expression imposed by the miR-58/80-82 family in non-neuronal tissues in regulating the spatial expression of PMK-2 is reminiscent of the magnitude of target repression exhibited by lin-4 and let-7 microRNAs in the temporal control of developmental timing [2 , 32] . In addition , our data corroborate prior phenotypic analysis suggestive of redundancy among members of the miR-58 family [5 , 6] , demonstrating redundant roles for miR-58 , miR-80 , miR-81 , and miR-82 in the destabilization of pmk-2 mRNA and corresponding repression of activated PMK-2 protein levels . The broad and constitutive expression of mir-58/80-82 suggests a more general housekeeping role for this microRNA family in the establishment and maintenance of tissue-specific gene expression by repressing the expression of neuronal-specific genes in non-neuronal tissues . Supporting this hypothesis is the genome-wide analysis of tissue-specific gene expression in C . elegans , which revealed enrichment for miR-58 binding sites among neuronal genes [33] . We speculate that the defects in size , locomotion , and egg-laying behavior observed in the mir-58; mir-80; mir-81-82 mutant are due to the cumulative misexpression of miR-58/80-82 target genes in non-neuronal tissues , as pmk-2 loss-of-function alone cannot suppress these defects ( D . J . P . and D . H . K . , unpublished observations ) . The characterization of microRNAs expressed in specific tissues of mice revealed the presence of a single microRNA and/or microRNA family in high abundance [34] , raising the possibility that such microRNAs might function to establish tissue expression patterns . Our data on the regulation of PMK-2 tissue expression by the miR-58 family provide genetic evidence for this hypothesis and point to a more general role for the highly abundant miR-58 family in the maintenance of tissue-specific gene expression .
All C . elegans strains were maintained and propagated on E . coli OP50 as described previously [35] . N2 was the wild-type strain . The following mutations were used in this study: LGI: kyIs140[str-2::GFP , lin-15 ( + ) ] LGIII: agIs219[PT24B8 . 5::GFP , Pttx-3::GFP] , tir-1 ( qd4 ) , mir-80 ( nDf53 ) LGIV: mir-58 ( n4640 ) , pmk-2 ( qd284 ) , pmk-2 ( qd307 ) , pmk-2 ( qd287 ) , pmk-2 ( qd279 ) , pmk-2 ( qd280 ) , pmk-2 ( qd305 ) , pmk-2 ( qd171 ) , pmk-1 ( km25 ) LGX: mir-81-82 ( nDf54 ) , sek-1 ( km4 ) , nIs145[Ptph-1::GFP , lin-15 ( + ) ] Extrachromosomal arrays: qdEx101[Poperon::islo-1::pmk-3::pmk-2::GFP::pmk-1::mCherry] , qdEx102[Poperon::islo-1::pmk-3::pmk-2mut::GFP::pmk-1::mCherry] A list of all strains used in this study is provided in the Supporting Information ( S1 Table ) . Cultures of P . aeruginosa PA14 were grown in Luria-Bertani ( LB ) broth overnight at 37°C . Five microliters of the overnight culture was used to seed 35-mm slow-kill assay plates ( 1 . 7% agar , 0 . 35% peptone , 0 . 3% sodium chloride , 5 μg/ml cholesterol , 1 mM calcium chloride , 1 mM magnesium sulfate , 25 mM potassium phosphate ) containing 50 μg/ml 5-fluorodeoxyuridine ( FUdR ) , used to prevent eggs from hatching . The culture was seeded in the middle of the plates and was not spread to the edges , meaning the resulting lawn would be “small , ” allowing for behavioral avoidance . The seeded plates were incubated overnight at 37°C and then overnight at room temperature . Roughly 40 L4 larval stage worms were placed onto a plate containing the prepared P . aeruginosa bacterial lawn with four plates per strain . The assay was carried out at 25°C . Plates were checked at the indicated times and worms that did not respond to a gentle prod from a platinum wire were scored as dead . Worms that crawled off of the plate or burrowed were censored . Adult worms ( specifically , 12–16 hr post L4 larval stage at 25°C ) were transferred to plates containing either a lawn of non-pathogenic E . coli OP50 or pathogenic P . aeruginosa PA14 and incubated at 25°C for 6 hr at which time the worms were immobilized with 50 mM sodium azide and mounted on 2% agarose pads . Immobilized worms were viewed using an AxioImager Z1 fluorescence microscope ( Carl Zeiss AG , Oberkochen , Germany ) with an EC Plan-Neofluar 40x/1 . 3 Oil DIC objective and the focal plane with the strongest GFP signal in the ADF neuron was used to take a 30 ms exposure picture with an AxioCam HRm camera . The images were analyzed in Fiji [36] , where the ADF neuron was located and the pixel intensity values were examined . The maximum pixel intensity value in the ADF neuron was used as the Ptph-1::GFP fluorescence value for each worm . In each experiment , 7–10 worms of the indicated genotype were imaged for each condition ( OP50 , PA14 ) . For the experiment shown in Fig . 1E , one outlier was identified ( in the tir-1 ( qd4 ) mutant , P . aeruginosa exposure dataset ) using Grubbs’ test and excluded from the graph and analysis . Statistical analyses of changes in fluorescence were performed in Prism 5 ( GraphPad Software , Inc . , La Jolla , CA ) using a two-way ANOVA and Bonferroni post-test . The qd284 , qd287 , and qd307 alleles of pmk-2 were isolated by CRISPR-Cas9-mediated genome editing as described previously [37] . Two separate pmk-2 single guide RNA ( sgRNA ) expression vectors derived from pUC57 were constructed following the published protocol . Both sgRNAs target sequences located in what corresponds to the first exon of the pmk-2 transcript . The target sequences were chosen to contain restriction enzyme recognition sites to facilitate the screening for mutations . Germline transformation was performed as described previously [38] using the following plasmids: 50 ng/μl Peft-3::Cas9-SV40 NLS::tbb-2 3’UTR , 45 ng/μl PU6::pmk-2 sgRNA , 5 ng/μl pCFJ104[Pmyo-3::mCherry] . Screening for mutations was performed as outlined [37] . To confirm that a frameshift in the first exon of pmk-2 results in a null allele , the qd284 mutation was crossed into the mir-80; mir-58; mir-81-82 mutant and levels of activated PMK-2 protein were determined . Activated PMK-2 protein was not detected in the mir-80; mir-58 pmk-2 ( qd284 ) ; mir-81-82 mutant , indicating that these mutations are null ( S2 Fig . ) . The qd171 allele of pmk-2 was isolated from a screen for suppressors of the enhanced susceptibility to pathogen ( Esp ) phenotype conferred by the km25 loss-of-function allele of pmk-1 . pmk-1 ( km25 ) mutant L4 larvae carrying the agIs219 reporter transgene for p38 activity in the intestine were mutagenized with ethyl methanesulfonate ( EMS ) as described previously [39] . Synchronized F2 progeny from roughly 35 , 000 mutagenized genomes were screened for an increase in expression of GFP from the agIs219 reporter transgene using a dissecting microscope equipped to detect GFP fluorescence . Mutants with increased expression of GFP were tested individually for the ability to suppress the Esp phenotype conferred by loss of pmk-1 function . Mutants that suppressed the Esp phenotype conferred by loss of pmk-1 function were then subsequently screened for suppression of the diminished expression of pmk-1 target genes conferred by loss of pmk-1 function . Single-nucleotide polymorphism ( SNP ) -based mapping using the C . elegans isolate CB4856 was performed as described previously [40 , 41] . One of the suppressors from this screen , qd171 , mapped to a region on LGIV containing the pmk operon . Sequence determination of pmk-1 and pmk-2 revealed that qd171 was an allele of pmk-2 . The qd305 deletion allele of pmk-2 was isolated by CRISPR-Cas9-mediated genome editing as described previously [37] . Three separate pmk-2 sgRNA expression vectors derived from pUC57 were constructed following the published protocol . All three sgRNAs target sequences located within or directly downstream of the 3’UTR of pmk-2 . Germline transformation was performed as described previously [38] using the following plasmids: 50 ng/μl Peft-3::Cas9-SV40 NLS::tbb-2 3’UTR , 50 ng/μl of each of the three PU6::pmk-2 3’UTR sgRNA , 5 ng/μl pCFJ104[Pmyo-3::mCherry] . Screening for deletions was performed by PCR analysis of F1 transgenic animals . The qd279 and qd280 alleles of pmk-2 were isolated from a screen for suppressors of the pmk-2 ( qd171 ) suppressor phenotype of the pmk-1 ( km25 ) loss-of-function phenotype . pmk-2 ( qd171 ) pmk-1 ( km25 ) mutant L4 larvae carrying the agIs219 reporter transgene for p38 activity in the intestine were mutagenized with EMS as described previously [39] . Synchronized F2 progeny from roughly 14 , 000 mutagenized genomes were screened for a decrease in expression of GFP from the agIs219 reporter transgene using a dissecting microscope equipped to detect GFP fluorescence . The sequence of pmk-2 was determined in mutants with diminished GFP expression from the agIs219 transgene . We identified three alleles of pmk-2: qd279 and qd280 ( which are missense mutations in conserved residues and were utilized in this study ) , as well as qd281 . Hypochlorite-synchronized populations of L4 larval stage worms were flash-frozen in liquid nitrogen and stored at -80°C until RNA extraction using TRI reagent ( Ambion , Life Technologies , Thermo Fisher Scientific , Inc . , Waltham , MA ) . Strain MT15563 , carrying mutations in mir-58 , 80-82 , grew slower than the other strains and therefore was harvested ~10 hr after the wild type strain . For 3’ RACE experiments , cDNA was prepared using the FirstChoice RLM-RACE Kit ( Ambion ) . Two successive rounds of PCR were performed using nested primers specific to pmk-2 and an adaptor at the 3’-end . Sequence determination was performed by direct Sanger sequencing of 3’ RACE products . miR-58/80-82 seed match sites were identified using TargetScan [42] . For quantitative RT-PCR experiments , cDNA was prepared with the RETROscript Kit ( Ambion ) using oligo dT primers . qRT-PCR was performed with a Mastercycler Realplex ( Eppendorf AG , Hamburg , Germany ) with SYBR Green detection ( Roche Diagnostics Corp . , Indianapolis , IN ) in triplicate 20 μl reactions . pmk-1 , pmk-2 , and pmk-3 mRNA levels were normalized to the control gene snb-1 . Fold change relative to wild type was determined using the Pfaffl method [43] . Sequences of primers used for qRT-PCR: pmk-1 , tgaatgatgatgtaagggcaga and cttcctcttcgtcagcaaatg; pmk-2 , caagtgttacgtgggctcaa and cgagaatcttgacttcgcatc; pmk-3 , gtatcgaagcaacgggaaac and tggaccacatggttttgaga; snb-1 , ccggataagaccatcttgacg and gacgacttcatcaacctgagc . For the experiments with pmk-1 ( km25 ) , pmk-2 ( qd171 ) pmk-1 ( km25 ) , and pmk-2 ( qd279 qd171 ) pmk-1 ( km25 ) strains with RNAi treatment ( Fig . 3D ) , mixed stage populations of worms subjected to RNAi for multiple generations were used for Western analysis . For experiments with mir-58/80-82 family microRNA deletion strains ( Fig . 4C and S2 Fig . ) and pmk-2 ( qd305 ) ( S1C Fig . ) , hypochlorite-synchronized populations of L4 larval worms were used for Western analysis . Strain MT15563 , carrying mutations in mir-58 , 80-82 , grew slower than the other strains and therefore was harvested ~10 hr after the wild type strain . Worms were collected , washed twice with M9 , incubated in M9 while rotating at 20°C to clear the gut of bacteria , washed again twice with M9 , and then pelleted . An equal volume of 2x lysis buffer ( 4% SDS , 100 mM Tris HCl pH 6 . 8 , and 20% glycerol ) was added to the worm pellets , which were then boiled for 15 minutes with occasional vortexing and then centrifuged to pellet the debris . The protein concentration of the lysates ( supernatant from the previous step ) was determined using the BCA Protein Assay Kit ( Pierce , Thermo Fisher Scientific , Inc . , Waltham , MA ) . For each sample , 50 μg of total protein was separated on a 10% SDS-PAGE gel ( Bio-Rad Laboratories , Inc . , Hercules , CA ) and then transferred to a nitrocellulose membrane ( GE Healthcare , Little Chalfont , United Kingdom ) . Blots were blocked with TBST supplemented with 5% skim milk power and then probed with either a 1:1 , 000 dilution of rabbit anti-ACTIVE p38 MAPK pAb ( Promega Corp . , Madison , WI ) , which recognizes the dually phosphorylated TGY motif of activated p38 MAPK , or a 1:10 , 000 dilution of mouse anti-β-tubulin ( E7 Developmental Hybridoma Bank , Iowa City , IA ) in TBST with 5% skim milk powder . Horseradish peroxidase ( HRP ) -conjugated anti-rabbit and anti-mouse IgG secondary antibodies ( Cell Signaling Technology , Inc . , Danvers , MA ) were used followed by detection with ECL reagents ( GE Healthcare ) . Feeding RNAi was performed as previously described [44] . NGM agar plates supplemented with 2 mM isopropyl-β-D-thiogalactopyranoside ( IPTG ) and 25 μg/ml of carbenicillin were seeded with E . coli HT115 bacterial cultures carrying the control plasmid pPD129 . 36 ( Ligation number L4440 ) or carrying specific plasmids derived from pPD129 . 36 designed to target either sek-1 or pmk-2 for RNAi . For pathogenesis assays and visualization of GFP expression from the agIs219 reporter , 3–6 L4 larval stage worms were placed onto seeded RNAi plates and their progeny were assayed . For Western analysis , ~10 L4 larval stage worms were placed onto seeded RNAi plates and the progeny of these worms were harvested before being deprived of food . The pmk operon translational reporter was constructed through ligation of overlapping PCR amplicons by homologous recombination in Saccharomyces cerevisiae strain FY2 as previously described [45] . A ~20 . 3 kb region containing the pmk operon and ~4 . 4 kb of upstream sequence was amplified from fosmid 34bC01 in adjacent fragments with at least 50 bp of overlap between fragments . For the pmk operon translational reporter carrying mutations in the 3’UTR of pmk-2 , 90 bp reverse complement primers containing the desired mutations were designed and used to amplify the 3’UTR of pmk-2 . The gene encoding GFP was amplified from pPD95 . 75 ( Addgene , Cambridge , MA ) . The gene encoding mCherry was amplified from pCFJ90 ( Addgene ) . The PCR amplicons were transformed into S . cerevisiae strain FY2 along with destination vector pNP30 ( gift of N . Paquin , pNP30 is a pRS426-derived plasmid compatible with MosSCI integration at locus ttTi5605 on LGII ) digested with XhoI and AvrII ( New England Biosciences , Inc . , Ipswich , MA ) . Yeast DNA was extracted with phenol-chloroform and transformed into DH5-α electrocompetent cells ( Protein Express , Inc . , Cincinnati , OH ) . Plasmids were prepped using a Miniprep Kit ( Qiagen N . V . , Venlo , Netherlands ) and their sequences were verified . Germline transformations were performed as described previously [38] . To visualize expression of GFP from the agIs219 reporter transgene , adult worms ( 16–24 hr post L4 larval stage at 20°C ) were picked over to an unseeded NGM agar plate and immobilized with 50 mM sodium azide . Worms were viewed using a Stereo V12 fluorescence microscope ( Zeiss ) and pictures were taken with an AxioCam MRc camera . To visualize expression of GFP and mCherry from the pmk operon translational reporter , L4 larval stage transgenic animals were immobilized with 50 mM sodium azide and mounted on 2% agarose pads . Worms were viewed using a LSM510 confocal microscope ( Zeiss ) . Stacks of confocal images were acquired and processed in Fiji to obtain maximum projections [36] . All images were prepared in Photoshop ( Adobe Systems , Inc . , San Jose , CA ) . | MicroRNAs are small , noncoding RNAs that act post-transcriptionally to inhibit expression of their target mRNAs . Gene expression studies of microRNAs and their target transcripts in diverse organisms have suggested that microRNAs may function to shape patterns of tissue expression . In this paper , we show that the miR-58/80-82 family of microRNAs , which accounts for roughly half of all C . elegans microRNAs at all developmental stages , defines the spatial expression pattern of PMK-2 p38 MAPK . While the pmk-2 gene is broadly transcribed , its tissue-specific expression is established by the redundant activities of miR-58 , miR-80 , miR-81 , and miR-82 , which switch off expression of PMK-2 through destabilization of pmk-2 mRNA in non-neuronal tissues . Our data suggest a housekeeping role for the miR-58/80-82 family in establishing and maintaining neuronal patterns of gene expression in C . elegans , and supports a more general role for microRNAs in establishing patterns of tissue expression . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [] | 2015 | Tissue Expression Pattern of PMK-2 p38 MAPK Is Established by the miR-58 Family in C. elegans |
Many species rely on olfaction to navigate towards food sources or mates . Olfactory navigation is a challenging task since odor environments are typically turbulent . While time-averaged odor concentration varies smoothly with the distance to the source , instaneous concentrations are intermittent and obtaining stable averages takes longer than the typical intervals between animals’ navigation decisions . How to effectively sample from the odor distribution to determine sampling location is the focus in this article . To investigate which sampling strategies are most informative about the location of an odor source , we recorded three naturalistic stimuli with planar lased-induced fluorescence and used an information-theoretic approach to quantify the information that different sampling strategies provide about sampling location . Specifically , we compared multiple sampling strategies based on a fixed number of coding bits for encoding the olfactory stimulus . When the coding bits were all allocated to representing odor concentration at a single sensor , information rapidly saturated . Using the same number of coding bits in two sensors provides more information , as does coding multiple samples at different times . When accumulating multiple samples at a fixed location , the temporal sequence does not yield a large amount of information and can be averaged with minimal loss . Furthermore , we show that histogram-equalization is not the most efficient way to use coding bits when using the olfactory sample to determine location .
Diverse species throughout the animal kingdom use olfactory cues for navigation tasks critical to survival , including locating food sources and mating partners . However , olfactory navigation is not simple: odorants are often volatile and carried on rapidly changing currents , resulting in spatiotemporal distributions that are turbulent thereby defeating simple strategies such as gradient detection . Consequently , recent efforts at understanding olfactory navigation have focused on identifying the viable computational strategies for making navigation decisions [1 , 2] . Here we focus on the most basic aspect of this process: how odor samples are encoded in the first place . Since sensory resources are finite , tradeoffs are inevitable . For example , resources may be allocated to encoding individual samples of odor concentration at a fine level of detail , or alternatively , to encoding multiple samples , either in space or in time , but at a coarser resolution for concentration . In this study , we investigate the implications of these and related tradeoffs , using the tools of information theory . Specifically , we compare an array of sampling and encoding strategies , asking to what extent they capture information about location within an olfactory environment . There are several aspects of the statistics of an odor plume that can give clues as to the location of the source [3–7] . For example , the mean concentration varies smoothly in lateral and longitudinal directions . However , animals do not base their navigation decisions on mean concentration , as the time it takes to obtain stable estimates of mean concentration exceeds the typical time taken by animals to make navigation decisions [8–10] . Other olfactory features that have been proposed as useful for navigation decisions include the time between odor encounters [11–13] and intermittency ( the probability of the odor concentration above threshold ) [4] . However , as for mean concentration , obtaining stable estimates of these quantities takes more time than animals typically use for navigation decisions . Hence averaged quantities—even if aided by other sensory inputs—are probably not used to guide navigation decisions . These considerations motivate our focus on what can be learned from brief , localized samples . We do not address the issue of how to integrate odor samples with other sources of information . A key starting point for our analysis is the explicit recognition that the resources available for sampling and encoding an odor environment are finite , and that it is natural to quantify these resources in terms of bits . This leads to the framework of information theory , which has the advantage that it minimizes the assumptions about the odor distribution . As mentioned above , the sampling strategies we consider explore tradeoffs between the number of bits allocated to resolving concentration , and to sampling in space and time . The focus on these tradeoffs is motivated by the diversity of the sampling strategies that animals use . With regard to spatial aspects , most animals have two spatially separated antennae or nostrils which sample the olfactory environment , but the sensor spacing ranges from less than a mm to several cm . With regard to temporal aspects , insects’ olfactory receptors are continuously exposed to odorants , while rodents take periodic samples and adjust their sniff rate based on previous measurements [14–16] . In this article , we discuss sampling strategies based on local cues in light of how much information they provide about sampling location . To compare different sampling strategies , we computed the information that they conveyed about location , for three realistic olfactory environments . In each environment , odor concentration was empirically determined via physical measurements , planar laser-induced fluorescence [17] . We chose to use physical measurements of actual plumes not only to avoid the assumptions made by models of turbulence or the complexities of numerical simulations , but also because the non-idealities of physical measurements take into account the real-world issues that confront the olfactory navigator . Although the three environments differed with regard to flow rate , turbulence , and proximity to a boundary , a number of commonalities emerged . First , precise measurement of odor concentration is generally not useful . That is , after allocating one or two bits to a coarse representation of odor concentration , more information about location is gained by using additional bits for encoding concentrations at nearby locations in space or time , than by using these bits to refine the representation of concentration . We also demonstrate that using “histogram equalization” as a strategy to discretize odor concentration—which is optimal to convey information about intensity per se [18]—is not optimal when the goal is to determine location . That is , the optimal strategy for low-level sensory encoding depends on the ultimate use of the information . Finally , with regard to sampling in time , we find that the additional information gained from multiple samples is preserved even if the temporal order of the samples is ignored , and this provides a rationale for simple post-receptoral processing strategies .
Odor plume data were obtained experimentally using a surrogate odor ( acetone ) released in a turbulent flow within a benchtop low-speed wind tunnel . We imaged the odor structure using planar laser-induced fluorescence ( PLIF ) ; images were subsequently post-processed into calibrated matrices of normalized concentrations . We acquired three separate datasets varying in mean flow rates and proximity to a boundary . The wind tunnel has a test section measuring 1 m long , by 0 . 3 m tall , by 0 . 3 m wide . We collected odor plume data at flow speeds of 5 cm/s and 10 cm/s . Ambient air enters the tunnel through a contraction section and passes through a turbulence grid consisting of 6 . 4 mm diameter rods with a 25 . 5 mm mesh spacing . Air exits the test section through a 15 cm long honeycomb section used to isolate the test section from a fan located in the downstream contraction . The odor surrogate was released isokinetically through a 9 . 5 mm diameter tube on the tunnel centerline . The tube orifice was located 10 cm downstream of the turbulence grid . For one dataset , named boundary flow , a false floor spanning the length and width of the test section was placed directly below the release tube . Acetone vapor was used as a fluorescent odor surrogate . We generated the acetone vapor by bubbling a carrier gas through liquid acetone . Because acetone is denser than air , the carrier gas consisted of a mixture of air ( 59% v/v ) and helium ( 41% v/v ) such that the odor surrogate mixture was neutrally buoyant in the wind tunnel . We used a water bath to maintain the temperature of the odor mixture at ambient tunnel conditions . A 1 mm thick light sheet from a Nd:YAG 266 nm pulsed laser illuminated the odor plume in the test section , causing acetone vapor in the odorant mixture to fluoresce with an intensity proportional to its concentration . The laser sheet enters and exits the tunnel through longitudinal slits along the sides of the test section . Plume fluorescence was imaged through a glass window in the tunnel using a high quantum efficiency sCMOS camera , with a bit depth of 16 bit , at a framerate of 15 Hz synchronized with the laser pulses . To enhance signal-to-noise , images were binned to ( 512x512 ) pixels corresponding to a spatial resolution of 0 . 74 mm/pixel . Raw images were processed to correct for background according to the equation c ( t , x , y ) =1acI ( t , x , y ) F ( x , y ) , ( 1 ) where c is the normalized concentration , I is the image from the camera ( with background signal subtracted ) and F is the flatfield image ( also with the background signal subtracted ) . The calibration coefficient , ac , was used to normalize the concentrations based on the source concentration at the tube exit . Three datasets were collected , which had different combinations of wind tunnel flow rates and false floor configurations ( Table 1 ) . The first condition , named fast flow , had a mean free stream velocity of 10 cm/s , and the odor mixture was released into the center of the tunnel without a false floor . The second condition , named slow flow , had a free stream velocity of 5 cm/s , and acetone was also released into the center of the tunnel without a false floor . The third condition ( boundary flow ) had a free stream velocity of 10 cm/s , but in contrast to the first condition , acetone was released with the false floor in place . All datasets were collected in segments of 4 minutes . We had a total of 40 minutes ( 36000 frames ) for the first and third condition , and 36 minutes ( 32400 frames ) for the second dataset . The matrices of normalized concentrations provide a natural coordinate system . Time-averaged odor concentrations and two typical snapshots for the three conditions are shown in Fig 1 . To compare olfactory cues across different flow conditions , we chose two grids of 16 locations in each olfactory landscape ( G narrow and G wide ) . Coordinates of the locations for the grid choices ( inlet location at the origin ) are: G narrow = { ( x , y ) | x = ( 2 . 2 , 5 . 9 , 9 . 6 , 13 . 3 ) cm , y = ( - 4 . 4 , - 1 . 5 , 1 . 5 , 4 . 4 ) cm } , G wide = { ( x , y ) | x = ( 5 . 6 , 11 . 1 , 16 . 7 , 22 . 2 ) cm , y = ( - 2 . 6 , - 1 . 1 , 1 . 1 , 2 . 6 ) cm } . ( 2 ) The two grids were chosen to capture the environment close to the source and further away from it above and below the centerline . The locations are indicated as blue circles ( G narrow ) and green triangles ( G wide ) in Fig 1 . The distances between gridpoints and the odor source are directly relevant to walking flies and other small insects . Probability distributions of the odor concentrations of the upper half of all grid points are shown in S1 Fig . Our primary goal is to quantify the extent to which a small number of samples of odor concentration within a plume provide information about the location of the sample . A principled approach is to use Shannon’s mutual information ( MI ) [19] for this purpose . That is , using entropy as a measure of uncertainty , we will determine the extent to which a given encoding scheme reduces the uncertainty about the location of the sample . Thus , our two variables of interest are location ( L ) and discretized odor samples ( M ) ; these are related in a complex statistical fashion . Specifically , this analysis quantifies the ability to discriminate between the 16 locations of either G narrow or G wide when the only available information comes from odor intensity samples . The choice of 16 locations per grid is somewhat arbitrary , however , in order to get stable information estimates with a given amount of data one trades off the number of locations with the number of bits using for odor coding . We settled on 16 locations as they capture a good proportion of the environment while allowing for the analysis of coding of odor samples with up to 10 bits . As is well-known , the MI between two random variables L and M is [19 , 20]: I ( L , M ) = H ( L ) - ∑ m ∈ M p ( m ) H ( L | m ) , ( 3 ) where H ( L ) is the ( unconditional ) entropy of L , and H ( L | m ) is the entropy of the distribution of L conditional on m ∈ M . In our context , L is the set of sampling locations G narrow or G wide and m ∈ M is a measurement of the normalized odor concentration c ( t , x , y ) . The specific representation of c as a ( coarser ) measurement m is an integral part of the encoding schemes we consider . We assume that the a priori probability of the locations l ∈ L are equal . It follows that the unconditional entropy is H ( L ) = - ∑ l ∈ L p ( l ) log 2p ( l ) = log ( | L | ) , ( 4 ) where | L | is the number of sampling locations . Note that the MI ( Eq 3 ) is a property of the grid as a whole , not the individual points . Since all | L | grid points have the same a priori probability , the upper bound of the MI is log 2 ( | L | ) . If the navigator has log 2 ( | L | ) bits of information then it knows its location on the grid unambigously . Posterior ( conditional ) distributions p ( l|m ) were calculated by Bayes theorem . Specifically , we binned the odor concentrations c at each location p ( m|l ) and then normalized the likelihoods by p ( m ) . The entropy of these conditional distributions are given by H ( L | m ) = - ∑ l ∈ L p ( l | m ) log 2 p ( l | m ) . ( 5 ) This quantity , weighted by the probability that sample m occurs p ( m ) , is summed over all m ∈ M to determine the average conditional entropy in Eq ( 3 ) . We used two contrasting strategies for representing the odor concentration as discrete symbols ( bins ) . In the first strategy , we divided the data into equal quantiles , i . e . we chose boundaries such that the distribution p ( m ) is uniform . This histogram-equalization procedure maximizes the information conveyed about the odor concentration ( i . e . , M ) [20 , chap . 2] , but does not necessarily maximize the information conveyed about sampling location . In the second strategy , we adjusted these bin boundaries to increase the amount of information about location . Because finding the bin boundaries that yield an absolute maximum is a multidimensional discrete optimization problem , we used the following “greedy” iterative strategy to find an approximate maximum . The first bin boundary is chosen to maximize I ( L , M ) , and is identified by an exhaustive search of the range of concentrations . Then , iteratively , the k-th boundary is chosen to maximize I ( L , M ) while keeping the k − 1 bin boundaries fixed . This is also a one-dimensional search over the range of concentrations , and leads to a binary subdivision of one of the bins determined at the previous step . For analyses in which the odor at multiple temporal or spatial samples is encoded , we used the bin boundaries determined from these single-sample optimizations . The encoding strategies we considered are specified not only by the way that each sample is encoded ( i . e . , the bin boundaries ) , but also by the number of spatial samples rspat and the number of temporal samples rtemp . Specifically , S ( n bits ; r spat , r temp ) , ( 6 ) denotes an encoding strategy that uses nbits to discretize odor intensity , applies this discretization to rspat samples at nearby locations obtained at rtemp points in time . Note that the number of bins used to discretize odor concentration is given by 2 n bits . When investigating strategies with two sensors ( rspat = 2 ) , we take two samples at a distance of 0 . 3 cm ( four pixels ) centered around the locations specified in Eq ( 2 ) . For sampling strategies specified by the notation of Eq ( 6 ) , bin boundaries are obtained by histogram equalization . To indicate that the “greedy” strategy has been used for obtaining bin boundaries , we use the symbol n bits * . The total number of bits used for encoding a sample m is given by nbits ⋅ rspat ⋅ rtemp ( or n bits * · r spat · r temp ) . To ensure that our results do not reflect the idiosyncrasies of odor concentrations at specific locations , all calculations were repeated after jittering the grid location . Specifically , the grid was rigidly moved from its standard location ( as given in Eq ( 2 ) ) by 0 . 74–2 . 22 mm ( 1-3 pixels ) in x and y directions , yielding a total of 49 placements . In all figures of the results section , mutual information at these jittered locations are shown as shaded blue and green regions .
We considered encoding schemes that probed the three basic ways in which resources could be allocated to encoding the odor measurements: for resolving concentration , for sampling across space , and for sampling across time . Here and in the other analyses below , parallel calculations were carried out for three odor environments: fast flow ( A ) , slow flow ( B ) and boundary flow ( C ) , and for two sets of locations ( narrow grid ( blue ) and wide grid ( green ) ) within each environment . The fast flow and boundary flow conditions have the fastest inlet flow of 10 cm/s , but the boundary flow dataset was taken near a boundary where the odor surrogate’s dynamics are affected by boundary layer effects . Hence , boundary flow is the condition were diffusion has the biggest impact; see Methods for details . As a consequence of the more diffusive regime of the boundary flow condition the mutual information values we obtained for this condition are somewhat higher than in the other two conditions . The slow flow dataset has an inlet velocity of 5 cm/s . Except as noted , the analyses with different datasets and different grid choices yielded similar results . Fig 3A1–3C1 shows results for strategies that devote all bits to encoding concentration at one point in space and time ( S ( n bits ; 1 , 1 ) ) . As the resolution for odor concentration increases , so does MI , but only up to a point: once four bits are used to resolving odor concentration , additional resolution yields only minimal increases in MI . When measurements are made at two sensor locations ( transversely separated by 0 . 3 cm ) , using additional bits for coding allows MI to increase beyond the plateau encountered with a single sensor ( Fig 3A2–3C2 ) . The benefit of spatial sampling is not merely the result of having two independent samples . Specifically , MI computed after ignoring which sample corresponded to which sensor was smaller , by up to 0 . 1 to 0 . 2 bits ( dashed curves in Fig 3A2–3C2 ) , than the MI conveyed by a coding scheme that keeps track of which sample is which . This indicates that sampling with two sensors enables extraction of a spatial feature of the odor plume that varies along the vertical axis . This trend is also true for different spacing between two sensors , as shown for half intersensor distance and double intersensor distance in S5 Fig . Note that in the boundary flow condition , the curves continue to increase rapidly at the limits of measurement , suggesting that MI is not close to saturation . Encoding odor measurements at two consecutive times ( separated by 1 . 6 s ) also increases MI beyond the plateau of a single sample , but not by as much as for two spatial samples ( Fig 3A3–3C3 ) . While each additional bit used for resolving the concentration of two consecutive samples provides greater MI , the increases become progressively less , suggesting that MI has reached a plateau when five bits of resolution are devoted to two samples separated in time . Virtually identical results are obtained for longer intervals between samples; this is expected since MI reaches an asymptotic value as a function of sampling interval ( see section temporal encoding strategies below ) . In the above analysis , we discretized the odor concentration into sub-intervals of equal probability , as this histogram-equalization procedure provides the greatest amount of information about the odor concentration itself [18 , 20] . However , this does not yield the maximal MI about location , so we carried out a further analysis that explored the discretization strategy . For the simple case of discretization into two levels , we show how the MI depends on the binarization threshold in Fig 4 . For the boundary flow condition ( C ) the information curves are flat over a large range for the narrow grid , and has a maximum above the median for the wide grid . For the fast flow ( A ) and slow flow ( B ) condition the maximum of information is obtained when the threshold is above the median for both grids . This suggests the most informative samples about location occur at high concentration . A threshold above the median exploits this feature of the odor statistics and allows better discriminability between locations . A comparison between the bin boundaries obtained by histogram-equalization and the optimal bin boundary when binarizing odor can be seen in S2 Fig . It is evident that the optimal bin boundary occurs at a higher concentration than the median for all but the narrow grid of the most diffusive condition . To investigate how a different choice of bin boundaries affects the results of Fig 3 , we implemented a “greedy” partitioning scheme ( see Methods ) in which the first cutpoint was chosen to yield the maximal MI about location , and then successive cutpoints were chosen so that each maximized the MI about location , given the previous partitioning . Results ( see S3 Fig ) were very similar to the above analysis based on histogram-equalized bins ( Fig 3 ) . Although one- and two-bit encoding schemes ( two to four partitions ) yielded more MI than histogram equalization , the plateau seen in row 1 of Fig 3 was essentially unchanged . The advantage of encoding schemes based on two spatial or two temporal samples persisted . The above findings show that overall , there is surprisingly little benefit to allocating coding bits to resolving odor concentration , compared to allocating them to capture several samples across space or time . We hypothesized that resolution of odor concentration might become more important in regimes that were more diffusive , especially when coupled with sampling at two locations . To investigate this hypothesis , we compared coding schemes in which the same number of bits ( four bits at each of two spatial samples ) were allocated to one , two , or four samples in time , and in which the spatial sampling was across the flow axis ( as in Fig 3 ) , or along the flow axis . Fig 5 shows that this hypothesis is supported . Considering first bin boundaries based on histogram equalization , and sensor locations across the flow axis ( unshaded portions of plots in first row of Fig 5 ) , two or more bits were only beneficial for the most diffusive environment boundary flow ( Fig 5C ) . Likewise , for sensor locations along the flow axis ( shaded half of each subplot ) , more than one bit of resolution was only helpful in this environment ( boundary flow ( Fig 5C ) ) . Similar conclusions are reached when bin boundaries are determined via the “greedy” binning procedure: more than one bit of resolution for odor concentration is only useful in the most diffusive environment ( boundary flow ( Fig 5C ) ) , and has the greatest benefit when the two sensors are across to the axis of flow . In the fast flow condition , increasing resolution while decreasing the number of samples in time makes little difference ( Fig 5A ) , and for the slow flow condition ( Fig 5B ) , increasing resolution while decreasing the number of samples leads to a loss of information about location for either sensor orientation . In sum , the results of Figs 4C , 5C1 and 5C2 show that in a diffusive regime the exact choice of bin boundaries is not important , but devoting up to four bits to concentration resolution has a benefit over accumulating multiple temporal samples . When the flow conditions are more turbulent , a navigator benefits from classifying multiple odor samples at coarser resolution ( Fig 5A and 5B ) , but the choice of the discretization threshold becomes important ( Fig 4A and 4B ) . Consistent across conditions , sampling across the odor plume yielded more MI than sampling along the mean flow direction ( white vs . gray shaded regions in Fig 5 ) .
We now discuss the implications of our findings , first with regard to sensation and then with regard to navigation algorithms . As a starting point , we consider the simple scenario of a sensory system confronted with a continuous and widely varying input , but limited in the number of symbols that it can use for encoding . As is well-known , information is maximized when each of the symbols is used equally often , i . e . , histogram equalization . Histogram equalization can be implemented as a nonlinearity applied to the input prior to producing a neural output [18] . For a positively skewed distribution , such as light intensities or odor intensities , the nonlinearity is a highly compressive one , so that it takes into account the rarity of very large inputs . Here , however , we consider the task of maximizing information not about the sensory signal itself , but about location—which is related to odor concentration in a complex , stochastic manner . As we showed , most of the available information about location can be conveyed by a coarse discretization of the sensory range—in fact , by binarization . However , this only holds if the cutpoint is properly chosen . In the two more turbulent odor environments considered here , the optimal cutpoint is substantially higher than the median , which is the cutpoint associated with histogram equalization ( see Fig 4 ) . That is , discriminations in the upper range of odor concentrations play a disproportionately greater role in determining location , than in reconstructing the input per se . Correspondingly , implementation of this encoding requires a nonlinearity that is less compressive for higher intensities than histogram equalization . Optimal adaptation strategies , in the sense of being maximally informative , under naturalistic stimuli are ( to our knowledge ) unknown . The problem of optimally discretizing a signal is not just an olfactory problem but applies to other sensory modalities which face resource constraints as well ( e . g . vision [26–28] ) . While it is difficult to imagine a biologically-plausible mechanism that achieves the precisely optimal nonlinearity for conveying information about location , there is a simple and plausible mechanism that can achieve an approximation: ligand-receptor binding in olfactory receptor neurons [29] . In steady-state , this mechanism generates a nonlinear encoding described by the Hill equation [30] . This transformation compresses signals at high concentrations , because receptors become occupied , and more ligand is required to activate the remaining receptors [31] . Thus , the degree of compression depends on the apparent dissocation constant Kd , the odorant concentration at which half of the receptors are occupied . Setting Kd at the median odor concentration corresponds to histogram equalization: half of the time the ligand binding will be below the median , and half of the time it will be above . Interestingly , setting Kd at the mean concentration , rather than the median , leads to less compression than histogram equalization . This is because the measured odor concentrations are positively skewed . Since the mean odor concentration is larger than the median , this setting will produce a response that is less than half-maximal most of the time . Such a coding strategy results in more information about location than histogram equalization , as we have outlined above ( see Fig 4 ) . In order to implement this strategy , olfactory receptors or receptor neurons would have to have an apparent Kd close to the mean concentration in the environment . Adaptation of Kd to the mean has been observed in olfactory receptor neurons of the fruitfly [32–34] , and might serve to increase the amount of information that the fly olfactory system can encode about its location in a turbulent environment . With regard to odor navigation algorithms , we note that these fall into two categories: those that rely on local cues ( e . g . comparison of concentration differences in two sensors [35] , comparison of sample arrival times in two sensors [13] , the combination of local anemotactic and olfactory cues [36 , 37] ) , and those algorithms that construct a cognitive map ( like infotaxis [1] and mapless [2] ) . We do not intend to argue for one kind of strategy over the other , but rather to identify aspects of the odor navigation problem that apply to both , as both begin with the acquisition of sensory samples . Our work suggests that these algorithms can operate on a coarse representation of odor concentration since we find that a four-bit representation of the odor intensity reveals almost the same amount of information as finer odor concentration representation . We also found that sampling with two sensors adds substantially to the amount of information about location , and this improvement is not just due to obtaining two samples , but by comparing them in a labelled fashion ( as observed in the second row of Fig 3 ) . While this is directly exploited by comparison algorithms using two sensors , we suggest that , navigation algorithms that use an internal model of the odor distribution like infotaxis and mapless could also be improved by incorporating measurements from two sensors . Finally , an important caveat of our study is that animals have multi-sensory cues available; here we only consider the single modality of odor and do not integrate information of other modalities , e . g visual or mechanosensory flow information , that navigators have access to . In particular , it is crucial for moths and fruitflies to combine flow information via mechanosensory input when walking and visual input when flying for successful navigation [38–41] . For example , since the wind direction may meander substantially , a simple upwind movement can lead a navigator out of the odor plume [8 , 42] . Simultaneously recording flow and odor concentration , and analysis along the lines undertaken here , may shed light on useful sampling strategies for combining both sources of information . Determining the location of an odor source based on olfactory cues is a challenging problem . We focused on how to optimally sample from the odor distribution when the goal is to determine location with respect to the source . This study shows that the sampling strategy that maximizes information about location under finite resources utilizes two sensors , allowing for the comparison of spatially separated samples , while representing odor concentration in no more than three to four bits . Furthermore , temporal sequences of samples can be averaged to preserve resources while only minimally affecting the amount of information that the sequence conveys . | Navigating towards a food source or mating partner based on an animals’ sense of smell is a difficult task due to the complex spatiotemporal distribution of odor molecules . The most basic aspect of this task is the acquisition of samples from the environment . It is clear that odor concentration does not vary smoothly across space in many natural foraging environments . Using data from three different naturalistic environments , we compare different sampling strategies and assess their efficacy in determining the sources’ location . Our findings show that coarsely encoding the concentration of samples at separate sensors and/or multiple times provides more information than encoding fewer samples with higher resolution . Furthermore , coding resources should be focused on discriminating rare high-concentration odor samples , which are very informative about the sampling location . Such a nonlinear transformation can be implemented biologically by the receptor binding kinetics that bind odorants as a first stage of the sampling process . A further implication is that animals as well as computational models of algorithms can operate efficiently with a coarse representation of the odor concentration . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [
"classical",
"mechanics",
"chemical",
"compounds",
"fluid",
"mechanics",
"quantum",
"tunneling",
"neuroscience",
"organic",
"compounds",
"turbulence",
"probability",
"distribution",
"animal",
"behavior",
"mathematics",
"odorants",
"materials",
"science",
"zoology",
"quantum",
"mechanics",
"thermodynamics",
"entropy",
"animal",
"cells",
"behavior",
"materials",
"by",
"attribute",
"chemistry",
"fluid",
"dynamics",
"olfactory",
"receptor",
"neurons",
"probability",
"theory",
"continuum",
"mechanics",
"physics",
"animal",
"migration",
"cellular",
"neuroscience",
"acetones",
"organic",
"chemistry",
"cell",
"biology",
"animal",
"navigation",
"neurons",
"biology",
"and",
"life",
"sciences",
"cellular",
"types",
"physical",
"sciences",
"afferent",
"neurons"
] | 2018 | Information-theoretic analysis of realistic odor plumes: What cues are useful for determining location? |
Substrate dephosphorylation by the cyclin-dependent kinase ( Cdk ) -opposing phosphatase , Cdc14 , is vital for many events during budding yeast mitotic exit . Cdc14 is sequestered in the nucleolus through inhibitory binding to Net1 , from which it is released in anaphase following Net1 phosphorylation . Initial Net1 phosphorylation depends on Cdk itself , in conjunction with proteins of the Cdc14 Early Anaphase Release ( FEAR ) network . Later on , the Mitotic Exit Network ( MEN ) signaling cascade maintains Cdc14 release . An important unresolved question is how Cdc14 activity can increase in early anaphase , while Cdk activity , that is required for Net1 phosphorylation , decreases and the MEN is not yet active . Here we show that the nuclear rim protein Nur1 interacts with Net1 and , in its Cdk phosphorylated form , inhibits Cdc14 release . Nur1 is dephosphorylated by Cdc14 in early anaphase , relieving the inhibition and promoting further Cdc14 release . Nur1 dephosphorylation thus describes a positive feedback loop in Cdc14 phosphatase activation during mitotic exit , required for faithful chromosome segregation and completion of the cell division cycle .
Cellular reproduction is a highly regulated process that is controlled on a multiplicity of levels , ensuring orderly progression through the different phases of the cell cycle and accurate partitioning of the genome . At the heart of eukaryotic cell cycle control lie cyclin-dependent kinases ( Cdks ) and their opposing phosphatases [1] , [2] . In Saccharomyces cerevisiae , the Cdk subunit Cdc28 associates with a series of cell cycle stage-specific cyclins to bring about Cdk activity . It is opposed by the main Cdk-counteracting phosphatase Cdc14 , which reverses Cdk phosphorylation events during mitotic exit [3] . The changing balance between Cdk and Cdc14 phosphatase activities at this stage of the cell cycle serves to order mitotic exit events , such as spindle elongation and chromosome segregation followed by spindle disassembly and ultimately cytokinesis [2] , [4] . So far , only a few of the Cdk substrates , whose dephosphorylation brings about mitotic exit , have been characterized [5]–[9] . During mitotic exit , Cdc14 is also essential for the downregulation of Cdk activity , on the one hand by promoting mitotic cyclin degradation , via dephosphorylation of the Anaphase Promoting Complex ( APC ) activator Cdh1 , and on the other by promoting accumulation of the Cdk inhibitor Sic1 [3] , [10] , [11] . Cdc14 activity is stringently regulated . During most cell cycle phases , Cdc14 is sequestered in the nucleolus , and thus inactive , through inhibitory binding to Net1 [12]–[14] . It is thought that Cdc14 release from Net1 occurs following phosphorylation of the latter , which can be achieved by a series of kinases including Cdk , Polo and MEN kinases , thus reducing Net1 affinity for Cdc14 [15]–[17] . Net1 phosphorylation , and thus Cdc14 release , is prevented until early anaphase by the action of the phosphatase PP2ACdc55 , which keeps Net1 under-phosphorylated . At anaphase onset , the protease separase is activated after APC-mediated destruction of its inhibitor securin . Separase now cleaves the chromosomal cohesin complex to trigger sister chromatid segregation . At the same time , separase uses a non-proteolytic activity to downregulate PP2ACdc55 [18] , [19] . This swings the phosphorylation balance on Net1 towards phosphorylation by mitotic Cdk which , with additional help from components of the FEAR network [17] , [20] , initiates Cdc14 release . While Cdc14 release during the early stages of anaphase depends on mitotic Cdk activity [17] , [19] , mitotic cyclins are being degraded at this time and Cdk activity is in decline . It is thought that declining Cdk and increasing Cdc14 contribute to activation of the MEN , a G-protein coupled signaling cascade consisting of the GTPase Tem1 , its regulators Lte1 and Bub2/Bfa1 , and its downstream kinases Cdc15 and Dbf2/Mob1 [21]–[24] . Both Cdc15 and Mob1 are Cdk targets and their dephosphorylation in mid anaphase contributes to MEN activation [20] , [25] , [26] . Active MEN kinases in turn are thought to have the potential to maintain Net1 phosphorylation . However , how Cdc14 release is sustained while Cdk activity declines between anaphase onset and MEN activation has remained poorly understood . Timely Cdc14 activation in early anaphase is important for successful chromosome segregation . It is required to stabilize the anaphase spindle and is also required for completion of chromosome segregation , in particular telomeres and the late segregating rDNA locus [6] , [7] , [27] , [28] . Cdc14 promotes the condensation and segregation of the repetitive rDNA region during anaphase and cdc14 mutants display rDNA segregation failure despite unobstructed cohesin cleavage . How Cdc14 promotes rDNA segregation is still being debated . Condensin is recruited to the rDNA in anaphase in a Cdc14-dependent manner , where it appears to promote decatenation of the locus , making the condensin complex a prime candidate for Cdc14 regulation [28]–[30] . It has also been suggested that Cdc14 downregulates rDNA transcription by RNA polymerase I , which could facilitate condensin access to the locus [31] , [32] . On the other hand , rRNA synthesis continues unabated during mitotic exit , making this hypothesis appear less likely [33] . In any event , a Cdc14 target that is dephosphorylated to promote rDNA condensation and segregation in anaphase remains unknown . In this study , we take advantage of our recent phosphoproteome analysis of budding yeast mitotic exit [34] . In the search for Cdc14 targets that have a role in regulating rDNA segregation , we identified the nuclear rim protein Nur1 as a Cdc14 substrate . Failure to dephosphorylate Nur1 causes rDNA missegregation , however , this turns out to be the consequence of compromised Cdc14 activation rather than a specific rDNA segregation defect . This leads us to discover that Nur1 has a previously uncharacterized role in Cdc14 inhibition , and that its inhibitory activity is phosphorylation-dependent . Constitutive Nur1 phosphorylation delays Cdc14 activation , while non-phosphorylatable Nur1 causes premature Cdc14 activation . Thus , Cdc14-dependent Nur1 dephosphorylation in early anaphase forms a positive feedback loop to promote further Cdc14 release , with important implications for faithful chromosome segregation .
In the search for Cdc14 targets that promote rDNA condensation and segregation during anaphase , we reviewed the phosphoproteome of budding yeast mitotic exit . Cells were arrested in metaphase by depletion of the APC coactivator Cdc20 , then synchronous mitotic exit progression was induced by ectopic expression of the Cdc14 phosphatase . Mass spectrometry was used to survey the disappearance of phosphopeptides over the course of mitotic exit , with the original intention to identify proteins whose dephosphorylation controls cytokinesis [34] . Among the proteins that were dephosphorylated in response to Cdc14 expression , in addition to cytokinesis regulators , we identified the nuclear rim protein Nur1 ( Fig . 1A , B ) . Along with its binding partner Src1 , Nur1 is involved in tethering rDNA and telomeres to the nuclear envelope , its absence leading to decreased rDNA repeat stability , unequal rDNA segregation , as well as loss of telomere stability and silencing [35]–[37] . Nur1 was previously identified as a Cdk target , containing nine putative Cdk phosphorylation sites , of which four have been confirmed in mass spectrometry studies ( Fig . 1A ) [34] , [38] . Of these four , our phosphoproteome analysis covered three phosphorylation sites on two phosphopeptides . These disappeared with early to intermediate timing , relative to the phosphopeptides of all detected proteins , during Cdc14 induced mitotic exit ( Fig . 1B ) . To confirm cell cycle-dependent Nur1 phosphorylation , and the role of Cdc14 in its dephosphorylation , we monitored the electrophoretic mobility of Nur1 using Phos-tag gels during synchronous cell cycle progression . Cells were arrested in G1 by pheromone α-factor treatment , before release to progress through the cell cycle at 35 . 5°C and either rearrest in the next G1 phase by readdition of α-factor , or arrest in late mitosis following Cdc14 inactivation using the temperature sensitive cdc14-1 allele [39] . Protein extracts were prepared at the indicated times and cell cycle progression was monitored by FACS analysis of DNA content ( Fig 1C ) . Phos-tag gel analysis of the protein extracts revealed the appearance of slower migrating Nur1 isoforms 30 minutes after release from G1 arrest , coincident with the time of S-phase , presumably due to phosphorylation ( Fig . 1C ) . During undisturbed cell cycle progression , Nur1 reached its slowest migration at 45 minutes , in G2/M , before faster migrating forms appeared again at 60–75 minutes . This pattern is consistent with dephosphorylation during early anaphase , before cells completed cytokinesis at 90 minutes . When Cdc14 was inactivated , in the cdc14-1 strain , dephosphorylation was no longer observed and Nur1 accumulated in slow migrating forms in the late mitotic arrest . To confirm that the mobility shift observed on the Phos-tag gels is indeed due to Nur1 phosphorylation , Nur1 was immunoprecipitated from cells arrested in mitosis by nocodazole treatment and incubated in a control buffer , or in the presence of either λ-phosphatase or purified recombinant Cdc14 [4] . Incubation with either phosphatase , but not control incubation , led to conversion to faster migrating species on the Phos-tag gel , confirming that the slower migrating forms are the consequence of phosphorylation ( Fig . 1D ) . In order to examine the importance of Nur1 dephosphorylation , we took advantage of a strategy to create constitutively Cdk phosphorylated proteins by covalent fusion to a mitotic cyclin ( Fig . 1E ) [34] , [40] . In brief , gene targeting was used to fuse the NUR1 gene at its genomic locus with a mitotic cyclin Clb2 tagging cassette . Clb2 is modified to lack localization and destruction signals , so as not to interfere with Nur1 function . After gene targeting , NUR1 remains initially separated from the cyclin tag by a selectable marker , flanked by loxP recognition sites for the Cre recombinase . The marker is then removed through the action of β-estradiol-activatable Cre-ER recombinase , following hormone addition to the growth medium . As a control , a similar tag was created in which Clb2 harbors three additional point mutations that prevent it from interacting with and recruiting the Cdc28 kinase subunit ( denoted Clb2ΔCdk , see Materials and Methods for details ) . An HA epitope is also included in the tag to facilitate detection , both before and after excision of the marker . We found there was virtually complete conversion of Nur1 to Nur1-Clb2 and Nur1-Clb2ΔCdk , respectively , four hours following β-estradiol addition ( Fig . 1F ) . Furthermore , Western blotting revealed a haze of slower migrating forms in case of the Nur1-Clb2 fusion protein , but a sharp , faster migrating band in case of Nur1-Clb2ΔCdk , consistent with increased Nur1 phosphorylation due to the Clb2 fusion . We next examined the effect of the fusions on cell survival . Cells expressing Nur1-Clb2 are viable at 25°C but unable to grow at a higher temperature of 36°C ( Fig . 1G ) . In comparison , Clb2ΔCdk fusion did not affect cell growth at the high temperature , suggesting that temperature sensitivity is caused by the continuous presence of Cdk activity close to Nur1 . As the Nur1-Clb2 fusion appears to cause increased Nur1 phosphorylation , constitutive phosphorylation could be the cause of temperature sensitive growth . Alternatively , phosphorylation of proteins in the vicinity of Nur1 , due to the increased local Cdk concentration , could cause the temperature sensitivity . To differentiate between these possibilities , we created a version of Nur1 , in which its 9 Cdk consensus phosphorylation sites were replaced by alanines ( Nur1 ( 9A ) ) . We then repeated the process of generating Nur1 ( 9A ) -Clb2 and Nur1 ( 9A ) -Clb2ΔCdk fusions . The absence of Cdk phosphorylation sites on Nur1 restored temperature resistant growth following Clb2 fusion ( Fig . 1G ) . It also prevented the Nur1 mobility shift following Clb2 fusion . ( Fig . 1H ) . This suggests that indeed persistent Nur1 phosphorylation on its Cdk phosphorylation sites is the cause for a temperature sensitive growth defect . In order to study the effect of persistent Nur1 phosphorylation on rDNA segregation , we tagged the rDNA binding protein Net1 with YFP to visualize the behavior of the rDNA locus . We synchronized a cell population by α-factor block and release at 36°C and monitored rDNA segregation as cells progressed through mitosis . As an internal marker for segregation timing , we recorded the length of the elongating anaphase spindle in each cell , as well as whether or not the rDNA had separated and segregated into two opposite cell halves . In both a wild type control strain , as well as in the Nur1-Clb2ΔCdk or Nur1 ( 9A ) -Clb2 strain , rDNA segregation started at a spindle length of 5–6 µm and was complete by the time spindles reached 8 µm in length . In contrast , in Nur1-Clb2 cells , rDNA segregation only started at spindle lengths of 7–8 µm and never reached completion even when spindles were fully elongated ( Fig . 2A ) . Cytological observation of the rDNA locus showed that it reached its expected tightly condensed state in opposite cell halves in wild type and Nur1-Clb2ΔCdk cells ( Fig . 2B ) . In contrast , the rDNA often appeared stretched and uncondensed in Nur1-Clb2 cells . These observations are consistent with the possibility that Nur1 dephosphorylation promotes rDNA condensation , which in turn is required for its timely segregation . Notably , an rDNA segregation defect of similar extent to that in Nur1-Clb2 cells is seen after inactivation of the chromosomal condensin complex [27]–[30] . An rDNA segregation defect was also observed , albeit less pronounced , in Nur1-Clb2 cells at 25°C ( S1 Fig . ) . Taken together , we conclude that constitutive phosphorylation of Nur1 leads to delayed rDNA segregation , coincident with defective rDNA condensation . A consequence of rDNA segregation defects is unequal sister chromatid exchange and consequent chromosomal instability . As a measure for rDNA stability , we assessed the loss rate of an ADE2 marker within the rDNA repeats [36] , [41] . ADE2 loss causes accumulation of a red intermediate metabolite in the adenine biosynthesis pathway , thus causing the colony color to turn red . ADE2 loss during the first cell division after plating on agar medium will generate half red-sectored colonies . We therefore counted the fraction of half red-sectored colonies , among all colonies , as a measure for the ADE2 loss rate from the rDNA . The loss rate was approximately five-fold elevated in the Nur1-Clb2 strain at 25°C , compared to a wild type and Nur1-Clb2ΔCdk controls ( Fig . 2C ) . This is indicative of rDNA instability , probably as the consequence of rDNA segregation defects caused by the Nur1-Clb2 fusion , even at a permissive temperature . Nur1 is a nuclear rim protein that makes contact with the rDNA , so its phosphorylation status could directly affect rDNA condensation . Alternatively , Nur1 phosphorylation could indirectly affect rDNA condensation and segregation . In particular , delayed rDNA segregation is a hallmark of mitotic exit defects , as for instance observed in FEAR pathway mutants . In order to differentiate between these possibilities , we monitored cell cycle progression of Nur1-Clb2ΔCdk and Nur1-Clb2 cells . Cells were synchronized in G1 by α-factor treatment , released to pass through a synchronous cell cycle at 36°C , before being rearrested in the following G1 by α-factor readdition . FACS analysis of DNA content showed that Nur1-Clb2 cells spent at least 20 minutes longer with a 2C DNA content ( Fig . 3A ) , i . e . show delayed cell cycle progression between G2 and cytokinesis , compared to the control . To delineate where Nur1-Clb2 cells are delayed in cell cycle progression , we analyzed several cell cycle markers using Western blotting at frequent time intervals during the time course ( Fig . 3B ) . Appearance of securin and Clb2 at around the time of S-phase was indistinguishable between the control and Nur1-Clb2 cells , as was phosphorylation of Orc6 that takes place at that time . However , destruction of both securin and Clb2 were markedly delayed in Nur1-Clb2 cells , as was Orc6 dephosphorylation during mitotic exit and reaccumulation of the Cdk inhibitor Sic1 . This pattern suggests a delay of Nur1-Clb2 cells during mitosis and mitotic exit . To distinguish whether the mitotic delay takes place before or after the metaphase to anaphase transition , we monitored spindle morphology during cell cycle progression ( Fig . 3C ) . This revealed that Nur1-Clb2 cells persisted for somewhat longer with short metaphase spindles ( 1–3 µm in length ) , as well as a pronounced prolongation of the time that cell persisted with long anaphase spindles ( >3 µm in length ) . These observations are consistent with a delay of Nur1-Clb2 cells both at the metaphase to anaphase transition as well as during mitotic exit . A similar delay in mitotic progression , albeit less pronounced , was observed in Nur1-Clb2 cells at 25°C ( S1 Fig . ) . To confirm that the mitotic delay due to Nur1-Clb2 was caused by persistent Nur1 phosphorylation , we used cells harboring Nur1 ( 9A ) . Clb2 fusion to Nur1 ( 9A ) no longer delayed cell cycle progression ( S2 Fig . ) , indicating that indeed persistent Nur1 phosphorylation slows down mitotic progression . These findings further open the possibility that rDNA condensation and segregation defects seen in Nur1-Clb2 cells are a consequence of a mitotic exit delay . The mitotic delay observed in Nur1-Clb2 cells is reminiscent of that seen in cells with an inactive FEAR pathway [42] . In those cells , delayed Cdc14 phosphatase activation slows mitotic exit progression and rDNA segregation [27] , [28] . Given the phenotypic similarities , we examined the timing of Cdc14 release in Nur1-Clb2 , compared to wild type and Nur1-Clb2ΔCdk , cells . Again , we synchronized cells in G1 and then measured the timing of Cdc14 release as a function of spindle length as cells passed through mitosis . In several repeats of this experiment , we detected a small but statistically significant delay to Cdc14 release in the Nur1-Clb2 strain ( Fig . 4A ) . This suggests that phosphorylated Nur1 impedes Cdc14 activation , a likely cause for delayed mitotic exit and rDNA segregation defects . To investigate whether delayed Cdc14 activation is indeed the cause for mitotic defects in Nur1-Clb2 cells , we tested whether the dominant active CDC14TAB6-1 allele , which leads to Cdc14 being less tightly bound by Net1 [43] , restores cell survival at high temperature . Indeed , we found that CDC14TAB6-1 almost completely restored temperature-resistant growth of the Nur1-Clb2 strain ( Fig . 4B ) . As a control , we confirmed that CDC14TAB6-1 did not cause dephosphorylation of Nur1-Clb2 , e . g . due to the presence of higher than normal levels of Cdc14 activity . Western blotting revealed that Nur1-Clb2 mobility , in particular its slower migrating forms , remained unaffected by CDC14TAB6-1 ( Fig . 4C ) . This confirms that the temperature sensitive growth due to persistent Cdk phosphorylation of Nur1 is caused by defective Cdc14 activation . We also monitored the dynamics of cell cycle progression in the Nur1-Clb2 strain rescued by the CDC14TAB6-1 allele . This revealed that the mitotic delay caused by Nur1-Clb2 is largely reduced in cells carrying the CDC14TAB6-1 allele ( Fig . 4D ) . Both the delays at the metaphase to anaphase transition , as well as the delay during mitotic exit , were ameliorated . FACS analysis of DNA content as well as Western blotting analysis of Clb2 levels and of Orc6 dephosphorylation during a synchronous cell cycle confirmed that Nur1-Clb2 no longer affects mitotic progression in CDC14TAB6-1 cells . This indicates that the mitotic defects and associated loss of viability at high temperature in Nur1-Clb2 cells are caused by defective Cdc14 activation . If Nur1 , especially the Cdk phosphorylated form , prevents Cdc14 release in early anaphase , then eliminating Nur1 or removing its Cdk phosphorylation sites , should facilitate Cdc14 release . To investigate this possibility , we compared Cdc14 release kinetics in synchronized populations of wild type , nur1Δ and nur1 ( 9A ) cells . Strikingly , Cdc14 was released in over half of nur1Δ and nur1 ( 9A ) cells in early anaphase at spindle lengths below 3 µm , when Cdc14 release is almost never seen in wild type cells ( Fig . 5A ) . Even in metaphase cells with short ( 1-2 µm ) spindles , when Cdc14 is normally tightly sequestered , a substantial fraction of nur1Δ and nur1 ( 9A ) cells displayed released Cdc14 ( Fig . 5B ) . This suggests that phosphorylated Nur1 restricts Cdc14 release in early anaphase . Cdc14 activation is promoted by the FEAR network in early anaphase . To further study the impact of Nur1 on Cdc14 release at this time , we introduced the nur1Δ and nur1 ( 9A ) alleles into a spo12Δ strain , lacking a key component of the FEAR network [20] . In the spo12Δ background , Cdc14 release is delayed until spindles reach about 6 µm in length . Deletion of nur1 , or its replacement with nur1 ( 9A ) , restored Cdc14 early anaphase release ( Fig . 5C ) . The Cdc14 release profile in the nur1Δ spo12Δ and nur1 ( 9A ) spo12Δ double mutant strains approached that of wild type cells . However , at short spindle lengths , nur1Δ spo12Δ and nur1 ( 9A ) spo12Δ cells still showed premature Cdc14 release , as compared to wild type , while at longer spindle lengths the rescue did not fully match wild type release levels . These findings confirm that phosphorylated Nur1 is a potent inhibitor of Cdc14 release in early anaphase and that the FEAR network acts to overcome Cdc14 inhibition by Nur1 . However , the incomplete rescue of Cdc14 release in spo12Δ cells by Nur1 ablation suggests that the FEAR network acts at least in part by a mechanism different from inactivating Nur1 . We next compared the kinetics of mitotic progression between wild type , nur1Δ and nur1 ( 9A ) strains . Despite the advanced Cdc14 release in the latter strains , the timing of progression through mitosis , as measured by the fractions of cells displaying metaphase and anaphase spindles at each time point , was indistinguishable from wild type ( Fig . 5D ) . In a spo12Δ background , cells showed approximately a 20 minute delay in anaphase . In this case , nur1Δ or nur1 ( 9A ) restored the kinetics of mitotic progression close to wild type ( Fig . 5D ) . This confirms that Nur1 , specifically its phosphorylated form , can delay mitotic progression . Inactivation of Nur1 , by deletion or mutation of its Cdk phosphorylation sites , augments Cdc14 release in early anaphase and almost completely compensates for loss of FEAR network components . This could be because phospho-regulation of Nur1 has a role specific to early anaphase . Alternatively , Nur1 might be a general Cdc14 inhibitor at all stages of mitotic exit . To differentiate between these two possibilities , we tested whether Nur1 inactivation can also compensate for partial loss of MEN signaling . We used conditional thermosensitive alleles in two MEN kinases cdc15-2 and dbf2-2 and asked whether nur1 deletion would improve cell growth at a semi-permissive temperature when MEN signaling is partially compromised . However , cdc15-2 nur1Δ and dbf2-2 nur1Δ double mutants lost viability at increasing temperatures in a fashion indistinguishable from the parental cdc15-2 and dbf2-2 strains ( Fig . 6A ) . Thus , Nur1 does not appear to act later during mitotic exit , when the MEN takes control of Cdc14 release . Instead , Nur1 function as a Cdc14 inhibitor appears to be restricted to early anaphase . This conclusion is consistent with the Nur1 phosphorylation pattern during mitotic exit . Nur1 is Cdk phosphorylated during early mitosis , when it counteracts Cdc14 . In anaphase , Nur1 becomes dephosphorylated due to Cdc14 action and thus looses its inhibitory effect on Cdc14 . To investigate the mechanism of how Nur1 counteracts Cdc14 , we asked whether Nur1 is physically linked to components of mitotic exit control . Affinity purified fractions of Nur1 , analyzed by sensitive mixture mass spectrometry , contained Heh1 and mitotic monopolin , as well as a Net1 peptide [36] . However , while mitotic monopolin resides in the nucleolus , Nur1 could not be detected on the rDNA together with Net1 by chromatin immunoprecipitation [36] , [44] . To clarify whether Nur1 interacts with Net1 , we performed a co-immunoprecipitation experiment . A Net1-myc strain was synchronized in G1 and cell extracts prepared at regular time intervals following release into synchronous cell cycle progression . Net1 was immunoprecipitated from the extracts using an α-myc antibody . Nur1 coprecipitated with Net1 at all stages of the cell cycle , with little fluctuation to the efficiency of the interaction ( Fig . 6B ) . No Nur1 was recovered in a parallel control immunoprecipitation from extracts of a strain lacking the Net1 myc epitope . In addition , we detected Cdc14 in immunoprecipitates of Nur1 throughout the cell cycle . While we do not currently know with which of Net1 and/or Cdc14 Nur1 makes direct contact , this finding opens the possibility that Nur1 directly influences Cdc14 inhibition in conjunction with Net1 . A key event during Cdc14 activation in early anaphase is Cdk phosphorylation of Net1 on at least six Cdk consensus phosphorylation sites . It could therefore be that Nur1 antagonizes Cdc14 by counteracting Net1 phosphorylation . As a start to investigate this , we took advantage of the net1-6Cdk allele that lacks these six Cdk phosphorylation sites and , as a consequence , delays Cdc14 activation and causes a short mitotic exit delay [17] . If Nur1 impacts on Cdc14 by counteracting Net1 phosphorylation , then Nur1 inactivation will not be able to correct the mitotic exit delay of net1-6Cdk cells . In contrast , if Nur1 counteracts Cdc14 in a pathway different from Net1 phosphorylation , then its deletion should be able to advance Cdc14 activation in net1-6Cdk cells , just as it did in the spo12Δ background . However , nur1 deletion did not improve Cdc14 release nor reduce the mitotic exit delay of net1-6Cdk cells ( S3 Fig . ) . This suggests that Nur1 contributes to Cdc14 regulation most likely by influencing Cdk phosphorylation of Net1 . It will be an important task to directly study the influence of Nur1 on Net1 phosphorylation .
The paramount importance of the Cdc14 phosphatase during budding yeast mitotic exit is well established , with the list of its targets and functions increasing . For instance , the incompletely understood role of Cdc14 in cytokinesis has recently come into focus [34] , [45] , [46] . Concomitantly , our understanding of the regulation of nucleolar release and activation of this phosphatase is deepening [17] , [19] , [47]–[50] . In this study , we set out to further our molecular knowledge of the role of Cdc14 in rDNA condensation and segregation . A candidate Cdc14 target with potential to impact on rDNA segregation was identified as the nuclear rim protein Nur1 in our recent phosphoproteome screen of budding yeast mitotic exit . Nur1 has been previously linked to regulating rDNA stability . We confirmed Nur1 as Cdc14 substrate in vivo and in vitro , that is dephosphorylated in early anaphase . To study the role of Nur1 dephosphorylation , we created a constitutively phosphorylated form of Nur1 , by fusing it to a cyclin Clb2 moiety . The fusion led to an rDNA segregation delay and reduced viability , which was dependent on the Cdk phosphorylation sites , especially at high temperatures . This confirms Nur1 as a Cdc14 target whose dephosphorylation is critical for successful mitotic exit progression . Further investigation revealed that the overt rDNA segregation defect in Nur1-Clb2 cells was likely a secondary consequence of a primary defect in Cdc14 activation . While the Cdc14 substrate ( s ) that govern rDNA condensation , resolution and segregation in anaphase therefore remain elusive , we have uncovered a previously unknown Cdc14 substrate that acts to sustain Cdc14 activation , thus creating positive feedback of Cdc14 release in early anaphase ( Fig . 7 ) . While inhibition of Cdc14 release by Nur1 is normally restricted to early anaphase while Nur1 is phosphorylated , our Nur1-Clb2 fusion likely extends that period , potentially throughout mitotic exit . Nur1-Clb2 may thus elicit phenotypic consequences beyond the delayed Cdc14 release characteristic of FEAR network mutants . An example of a difference is the temperature sensitive growth of Nur1-Clb2 cells , which is not typically shared by FEAR network mutants . A higher temperature causes cells to progress faster through the cell cycle , presumably rendering them less tolerant to delays or alterations to cell cycle signaling pathways . Phenotypes of late mitotic exit mutants , in particular those involved in cytokinesis , are known to be exacerbated at high temperatures [34] , consistent with an impact of Nur1-Clb2 at later stages . Intriguingly , defective Cdc14 activation in Nur1-Clb2 cells caused not only a mitotic exit delay , but also a delay at the metaphase to anaphase transition . A small amount of Cdc14 is likely to be active by the time of metaphase , when it contributes to dephosphorylation of proteins like Dsn1 and possibly Spo12 [49] , [51] . Furthermore , securin dephosphorylation by Cdc14 has been suggested to sharpen the timing of anaphase [52] . Nur1-dependence of regulating these earliest Cdc14 targets might play a role at the metaphase to anaphase transition . We cannot exclude that Nur1 performs additional functions at this time , that might be affected by Nur1-Clb2 fusion and that might contribute to explaining the lethality of Nur1-Clb2 fusion strains at high temperatures . Our results demonstrate that Nur1 , through regulating Cdc14 , plays an important role in promoting the accurate timing of mitotic exit events . Nur1 was previously characterized as chromosome linkage protein , connecting chromatin and the inner nuclear membrane , with functions in the maintenance of genome stability and replicative life span [36] , [37] . Are these independent functions of this protein , or are they related ? Constitutive Nur1 phosphorylation , caused by the Nur1-Clb2 fusion , increased marker loss at the rDNA , likely due to unequal sister chromatid exchange . A similar level of marker loss at the rDNA was previously reported in nur1Δ cells [36] . It will be therefore interesting to examine whether Nur1's role in tethering the rDNA to the nuclear envelope is regulated by its phosphorylation status . It is as yet unknown whether rDNA tethering is modulated during the cell cycle , for instance during rDNA condensation and segregation , and how the rDNA moves relative to the nuclear envelope during anaphase . It has been speculated that cell cycle-dependent Nur1 phosphorylation within a putative nuclear localization signal might affect the nucleocytoplasmic shuttling of the protein , with possible implications both for Cdc14 regulation and rDNA tethering [37] . Our initial observations of Nur1 localization revealed faint Nur1 enrichment at the nuclear envelope , consistent with previous reports [36] , [53] . This localization did not noticeably change during cell cycle progression and was not altered as consequence of Clb2 fusion or Cdk phosphosite mutations . Nevertheless , the possibility of Nur1 regulation by localization merits further investigation . Nur1 interacts with Net1 and , in addition to Net1's role as a Cdc14 inhibitor , one of the phenotypes reported for net1-1 mutants is a high rate of chromosome loss [54] . This indicates that Net1 plays a role in accurate chromosome segregation that might go beyond its role as cell cycle regulator . The relationships between Net1 , Nur1 , the phosphoregulation of both proteins and their roles in Cdc14 phosphatase regulation and genome stability are an important topic for further studies . Removing Nur1 ( nur1Δ ) , or its Cdk phosphorylation sites ( nur1 ( 9A ) ) , resulted in premature Cdc14 release . Phosphorylated Nur1 therefore fulfills a role in attenuating Cdc14 activation in early anaphase . In a quantitative model in which Cdc14 substrate dephosphorylation timing is determined by the ratio of phosphatase activity versus Cdk kinase activity [2] , [4] , phospho-Nur1 ensures that Cdc14 activity is kept low and only its earliest targets are dephosphorylated in early anaphase . While preventing Nur1 phosphorylation causes premature Cdc14 activation , maintaining persistent Nur1 phosphorylation delays Cdc14 release and mitotic exit . Nur1 dephosphorylation by Cdc14 in early anaphase thus engages a positive feedback loop in which phospho-Nur1 as a Cdc14 inhibitor is removed by the very action of Cdc14 itself ( Fig . 7 ) . Whether Nur1 dephosphorylation actually stimulates release of Cdc14 , or just stops Nur1 from counteracting it , is still an open question . In either case , failure to turn off phospho-Nur1 results in a mitotic exit delay , compromised rDNA segregation and inability to survive at higher temperature . Thus , Nur1 dephosphorylation plays an important role in shaping the Cdc14 activation pattern in early anaphase , until the MEN takes over to sustain Cdc14 release . Additional players might contribute to Cdc14 feedback regulation , including Tof2 [50] or Spo12 [49] . Although we have found that Nur1 interacts with Net1 and Cdc14 , the molecular mechanism by which Nur1 counteracts Cdc14 release is not yet understood . The fact that Nur1 no longer influences Cdc14 activation in a net1-6Cdk strain suggest that Nur1 acts at the level of Cdk-dependent Net1 phosphorylation . A possible mechanism that is consistent with this observation takes into consideration that Net1 itself is a Cdc14 target in vivo and in vitro [12] , [25] . Cdc14 could prevent its own release if it were allowed access to Cdk phosphorylation sites on Net1 . A scenario can be envisaged in which phospho-Nur1 promotes dephosphorylation of Net1 by Cdc14 . The required change in accessibility of Net1 phosphorylation sites could be accomplished through conformational changes within a Net1/Nur1/Cdc14 protein complex , depending on the Nur1 phosphorylation state . In summary , we describe a positive feedback loop , in which Nur1 attenuates Cdc14 release until Nur1 itself is dephosphorylated by Cdc14 . This feedback imposes a requirement for the FEAR network to initiate Cdc14 release . Once Cdc14 becomes active , Nur1 dephosphorylation helps to sustain Cdc14 release while Cdk activity declines , until the MEN pathway takes over . In this model , Nur1 can be seen as a bridge between the FEAR network and the MEN signaling cascade .
All strains were of the W303 background and are listed in S1 Table . A strain containing ADE2 integrated within the rDNA repeats was a kind gift from M . Kaeberlein [41] . Epitope tagging of endogenous genes and gene deletions were performed by gene targeting using polymerase chain reaction ( PCR ) products [55] , [56] . The nur1 ( 9A ) mutant was engineered by endogenous gene replacement using an integrative plasmid , based on a synthetic DNA construct ( GeneArt , Life Technologies ) . The conditional Nur1-Clb2 and Nur1-Clb2ΔCdk fusions were created as described [34] . In brief , Clb2 lacking its destruction and KEN-boxes [57] , as well as its nuclear localization sequence [58] , was fused to Nur1 , separated by an unstructured 10-mer GGSGTGGSGT linker . In addition , the Clb2ΔCdk mutant contained further 3 point mutations that prevent it from interacting with the Cdc28 kinase subunit , as described [59] . Strains harboring the conditional Clb2-fusion cassettes were grown on medium lacking uracil to maintain the selectable marker and to prevent spontaneous recombination . Marker loop-out was then induced by β-estradiol-dependent activation of Cre recombinase fused to an estradiol binding domain ( Cre-EBD78; [60] ) , by addition of 1 µM β-estradiol to the growth medium . Yeast cultures were grown in rich YP medium supplemented with 2% glucose [61] . Cell synchronization using α-factor was as described [62] . For cell synchronization at higher temperatures , using the cdc14-1 or NUR1-CLB2 backgrounds , cultures were shifted to the higher temperatures at the time of release from α-factor arrest . Protein extracts for Western blotting were prepared following cell fixation using trichloroacetic acid , as described [63] , and analyzed by SDS-polyacrylamide gel electrophoresis ( SDS-PAGE ) . Antibodies used for Western detection were , α-Clb2 ( Santa Cruz , sc9071 ) , α-Orc6 ( clone SB49 ) ; α-Sic1 ( Santa Cruz , sc50441 ) , α-Tub1 ( clone YOL1/34 , AbD Serotec ) , α-HA ( clone 12CA5 ) , α-myc ( clone 9E10 ) , α-Pk ( clone SV5-Pk1 , AbD Serotec ) . Phos-tag was purchased from Wako Chemicals and added to SDS-polyacrylamide gels along with MnCl2 according to the manufacturer's instructions . For immunoprecipitation , cell extracts were prepared in EBXG buffer ( 50 mM HEPES pH 8 . 0 , 100 mM KCl , 2 . 5 mM MgCl2 , 10% glycerol , 0 . 25% Triton X-100 , 1 mM DTT , protease inhibitors ) using glass bead breakage in a Multi Bead Shocker ( Yasui Kikai ) . Extracts were precleared , incubated with antibody and finally adsorbed to Protein A Dynabeads . Beads were washed and elution was carried out in SDS-PAGE loading buffer . For the in vitro Nur1 dephosphorylation assay , immunoprecipitation was performed as above , then beads were resuspended in phosphatase buffer and 1 µg λ phosphatase ( New England Biolabs ) , or 8 µg purified recombinant Cdc14 [4] , were added , followed by incubation at 30°C for 30 minutes before the reaction was stopped and proteins eluted by addition of SDS-PAGE loading buffer . Indirect immunofluorescence was performed on formaldehyde-fixed cells using the following antibodies , α-GFP , ( clone TP401 , Torrey Pines Biolabs or ab6556 , Abcam ) , α-Tub1 ( clone YOL1/34 , AbD Serotec ) and FITC and Cy3-dye labeled secondary antibodies ( Sigma and Chemicon , respectively ) . Cells were counterstained with the DNA binding dye 4' , 6-diamidino-2-phenylindole ( DAPI ) . Fluorescent images were acquired using an Axioplan 2 imaging microscope ( Zeiss ) equipped with a 100x ( NA = 1 . 45 ) Plan-Neofluar objective and an ORCA-ER camera ( Hamamatsu ) . Spindle length measurements were carried out in ImageJ . | During the cell cycle , a specific sequence of events leads to the formation of two daughter cells from one mother cell . Progression through the cell cycle is tightly controlled , with events occurring in the right place at the right time . Exactly how this is achieved is still being elucidated . In budding yeast , the events occurring during the final cell cycle phase – “mitotic exit” – are controlled by the phosphatase Cdc14 . It is kept sequestered and inactive until it is needed for mitotic exit , at which time it is rapidly released . In this study , we have identified a new regulator of Cdc14 activity , the protein Nur1 . In a series of experiments , we saw that Nur1 acts both upstream and downstream of Cdc14 activation , thereby creating a positive feedback loop . On the one hand , Nur1 contributes to inhibiting Cdc14 until the start of mitotic exit . On the other hand , through the actions of Cdc14 itself , Nur1 is disabled as an opponent of the phosphatase . This creates a robust system , rapidly switching between two opposing states and thus driving forward the mitotic exit transition . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods",
"and",
"Materials"
] | [
"signal",
"transduction",
"fungi",
"feedback",
"regulation",
"cell",
"biology",
"mechanisms",
"of",
"signal",
"transduction",
"cell",
"cycle",
"and",
"cell",
"division",
"biology",
"and",
"life",
"sciences",
"cell",
"processes",
"saccharomyces",
"cerevisiae",
"saccharomyces",
"organisms",
"yeast"
] | 2015 | Nur1 Dephosphorylation Confers Positive Feedback to Mitotic Exit Phosphatase Activation in Budding Yeast |
The RNA exosome complex functions in both the accurate processing and rapid degradation of many classes of RNA . Functional and structural analyses indicate that RNA can either be threaded through the central channel of the exosome or more directly access the active sites of the ribonucleases Rrp44 and Rrp6 , but it was unclear how many substrates follow each pathway in vivo . We used CRAC ( UV crosslinking and analysis of cDNA ) in growing cells to identify transcriptome-wide interactions of RNAs with the major nuclear exosome-cofactor Mtr4 and with individual exosome subunits ( Rrp6 , Csl4 , Rrp41 and Rrp44 ) along the threaded RNA path . We compared exosome complexes lacking Rrp44 exonuclease activity , carrying a mutation in the Rrp44 S1 RNA-binding domain predicted to disfavor direct access , or with multiple mutations in Rrp41 reported to impede RNA access to the central channel in vitro . Preferential use of channel-threading was seen for mRNAs , 5S rRNA , scR1 ( SRP ) and aborted tRNAs transcripts . Conversely , pre-tRNAs preferentially accessed Rrp44 directly . Both routes participated in degradation and maturation of RNAPI transcripts , with hand-over during processing . Rrp41 mutations blocked substrate passage through the channel to Rrp44 only for cytoplasmic mRNAs , supporting the predicted widening of the lumen in the Rrp6-associated , nuclear complex . Many exosome substrates exhibited clear preferences for a specific path to Rrp44 . Other targets showed redundancy , possibly allowing the efficient handling of highly diverse RNA-protein complexes and RNA structures . Both threading and direct access routes involve the RNA helicase Mtr4 . mRNAs that are predominately nuclear or cytoplasmic exosome substrates can be distinguished in vivo .
In Eukaryotes , the exosome is the major RNA degradation complex responsible for quality control of most transcripts in both the nucleus and cytoplasm , processing of stable RNA precursors and turnover of pre-mRNAs , mRNAs and large numbers of non-coding RNAs ( ncRNAs ) . A puzzling aspect of exosome substrate targeting is the basis of the distinction between precise 3’ processing of stable RNA species and the rapid , complete degradation of “constitutive” degradation substrates or aberrant RNAs and RNA-protein complexes ( reviewed in [1] ) . Processing targets include precursors to the 5 . 8S rRNA , small nucleolar RNAs ( snoRNAs ) and small nuclear RNAs ( snRNAs ) . Constitutive nuclear exosome targets include pre-rRNA spacer regions and several hundred different non-protein coding RNAs ( ncRNAs ) . Constitutive cytoplasmic targets are the ~6 , 000 mRNA species . Aberrant RNAs apparently arise from all classes of transcription unit , including pre-rRNAs , pre-tRNAs , pre-mRNAs and the precursors to many other stable RNA species . The exosome core has a barrel-like overall structure [2] and in the archaeal complex the nuclease active sites are positioned within a central channel ( Fig 1A ) [3] . In yeast and human cells the central channel is well conserved compared to Archaea , but point mutations have apparently eliminated the ancestral phosphorolytic activity of the complex [4–6] . Instead , hydrolytic exonuclease activity of the core exosome is provided by an associated protein termed Rrp44/Dis3 in yeast or Dis3 in humans . In the yeast nucleus and human nucleolus , a second exonuclease associates with the exosome , termed Rrp6 in yeast or EXOSC10 in humans . Rrp44 is composed of an N-terminal PIN domain responsible for endonuclease activity , two continuous RNA-binding cold-shock domains ( CSD domains ) , an RNB domain carrying the exonuclease active site , and an RNA-binding S1 domain ( Fig 1B ) . RNA substrates can reach the exonuclease site by threading through the central channel or via direct access [5 , 7–9] . In the major pathway to the exonuclease active site of Rrp44 , single-stranded substrates are threaded through the central channel , which protects around 33 nt of RNA [10] . Functional analyses of the PIN domain endonuclease activity of Rrp44 identified only the 7S pre-rRNA and excised 5’ ETS pre-rRNA fragments as targets for cleavage [11–13] , whereas UV crosslinking identified apparent interactions between the PIN domain and many exosome substrates [14] . Exosome crystal structures indicate that the PIN domain is not accessed by substrates via the central channel [7] . However , partial occlusion of the central channel by temperature sensitive ( ts ) mutations in Rrp41 was reported to inhibit both the exonuclease and endonuclease activity of Rrp44 [15] . The exonuclease site in Rrp44 can also be accessed by a more direct route . This involves a structural rearrangement that disrupts the route for RNA through the central channel to the Rrp44 exonuclease site [8 , 9] . Mutations in Rrp44 that are predicted to disfavor adoption of this direct access structure were reported to impair the degradation of two characterized exosome substrates , hypomodified tRNAiMet and truncated 5S ribosomal RNA ( rRNA ) [9] . The Rrp44 G916E mutation disturbs the OB-fold ( oligonucleotide/oligosaccharide-binding fold ) of the S1 RNA binding domain [16 , 17] and abolished RNA binding to hypomodified tRNAiMet in vitro [6] . We anticipated that inactivation of the S1 domain would reduce RNA recruitment by direct access ( Fig 1A , right panel ) , but not via threading through the central channel ( Fig 1A , left panel ) . Conversely , charge-reversal mutations in Rrp41 that impair entry of RNA to the central channel should reduce utilization of the threaded pathway with little impact on direct access substrates . In addition , we anticipated that substrates following a pathway of direct access to Rrp44 might show limited crosslinking to exosome components located in the barrel of the exosome . On RNAs threaded through the channel , we anticipated that the distribution of exosome proteins might be resolved , at least on highly abundant substrates with a well-defined site of stalling . The extent to which substrates for degradation by Rrp6 pass through the central channel in vivo , and in which orientation , remains unclear [18 , 19] . However , substrate passage through the channel or , indeed , any interaction with the exosome , is apparently not obligatory for Rrp6 activity . Similar RNA phenotypes are seen following depletion of any of the ten “core” exosome components , including Rrp44 , whereas loss of Rrp6 results in distinctly different effects showing that its functions are at least partially independent of the exosome barrel [20 , 21] . Many highly structured RNA substrates are preferentially targeted by Rrp6 , including the 5 . 8S+30 pre-rRNA , mature tRNAs , small nuclear ( snRNAs ) and small nucleolar RNAs ( snoRNAs ) [14 , 22] . The core exosome complex appears to have little activity , but functions in vivo with several different activating cofactors in the nucleus and cytoplasm . Key nuclear cofactors include the RNA helicase Mtr4 [23 , 24] [25] , which can function either alone or in the context of the Trf4/5-Air1/2-Mtr4 polyadenylation ( TRAMP ) complexes [26–31] . In vitro analyses have given major insights into the structure of the exosome and its interactions with RNA and cofactors . However , despite this outstanding work it remains much less clear , in vivo , which RNAs and RNA-protein complexes are partitioned between the multiple different routes to the exosome active sites , or whether specific routes are favored for individual substrates that are destined for degradation versus accurate 3’-end processing . The aim of this work was to resolve these questions using a combination of UV-crosslinking and analysis of cDNAs ( CRAC ) on exosome subunits or cofactors ( Mtr4 ) with mutations in the exosome to identify different classes of directly interacting RNAs .
Pre-ribosomal RNAs are characterized exosome substrates and were enriched in our datasets . In particular , the excised 5’ external transcribed spacer ( 5’ ETS ) that is released by cleavage at site A0 ( Figs 2 and S2 ) and the internal transcribed spacer 2 ( ITS2 ) region that is present on the 7S pre-rRNA ( 5 . 8S rRNA with ~140 nt 3’ extension ) and 5 . 8S+30 pre-rRNA ( S2 Fig ) . Notably , both the 5’ ETS and the ITS2 region of the 7S pre-rRNA are also characterized substrate for the endonuclease activity of Rrp44 [11 , 13] . It was , however , unclear whether they access Rrp44 directly or via threading . To address this point , we performed affinity-purification experiments without crosslinking . Rrp44-exo and Rrp44-exo-S1 both coprecipitated the full length 5’ ETS—A0 fragment and multiple degradation intermediates , notably a cluster of bands ranging from ~110–130 nt , indicating threading through the channel to Rrp44 ( Fig 2A ) . However , the shortest fragments detected ( ~80 nt ) using a transcription start site ( TSS ) proximal probe ( +49–67 ) notably failed to coprecipitate with the exosome lacking S1 RNA binding activity . Analysis of the read distribution across the 5’ ETS ( Fig 2B ) showed two major peaks for Rrp44-exo . The peak over +90 to +140 is retained in the Rrp44-exo-S1 double mutant and is also present in the datasets for the Csl4 , Rrp41 and Rrp6 exosome components and the TRAMP component Mtr4 . Point mutations in the cDNA sequence data indicate the direct sites of protein-RNA contact . Mutation analysis ( S2A Fig ) indicates that this region contains multiple Rrp44-associated sites , consistent with the multiple bands observed in the northern blot data ( Fig 2A ) . This strongly indicates that the +90 to +140 peak represents a set of major intermediates in Mtr4-mediated degradation by threading to Rrp44 , in cooperation with Rrp6 . In contrast , the TSS proximal peak ( ~+20 - +50 nt ) in Rrp44-exo was greatly reduced in the exo-S1 double mutant and largely absent from the Rrp41 and Csl4 crosslinking data , and recovered at only a low level withRrp6 . It was , however , retained in the Mtr4 crosslinking data . Downstream of the +90 to +140 region , distributed binding is seen in all datasets including Rrp44-exo-S1 and Mtr4 . This would be consistent with the northern blot data and indicates largely processive degradation through this region . The data would support a model in which the region from the 5’ end of the ETS to site A0 ( 5’ ETS-A0 ) would be threaded through the channel , contacting Csl4 and Rrp41 . We predict that this species is initially submitted to repeated cycles of oligo-adenylation by TRAMP , as previously proposed [26] , each facilitating unwinding by Mtr4 and threading ( Fig 2C ) . A number of fragments are detected from this region , which do not appear to be lost in the exo–S1 double mutant . From the location of the +130 nt fragment further degradation may be impeded by strong secondary structure that is predicted in the 5’ domain of the ETS . Short 5’ fragments with 3’ ends matching the major peaks and crosslinks in the ETS around +50 would not be observed in the northern blot analysis . However , the lack of crosslinking to Rrp41 , Csl4 or Rrp6 strongly indicates that these RNAs are not threaded though the channel . Despite this , the crosslinking data indicate that this region is a target for Mtr4 , strongly suggesting that Mtr4 can also facilitate the direct access pathway to Rrp44 . The 7S pre-rRNA was coprecipitated with similar efficiency by Rrp44-exo and Rrp44-exo-S1 suggesting that is processed by threading through the channel ( S2B Fig ) . However , the shorter 6S pre-rRNA ( 5 . 8S rRNA + 5–8 nt 3’ overhang ) was less efficiently recovered by Rrp44-exo-S1 than Rrp44-exo . The distribution of reads across the 7S pre-rRNA ( S2C Fig ) showed a strong peak upstream of the 3’ end of the 5 . 8S+30 pre-rRNA ( indicated by a dashed black line ) with all core exosome subunits and , to a lesser extent , with Rrp6 . The processing of 7S to 5 . 8S+30 is normally dependent on the core exosome rather than Rrp6 , whereas processing from 5 . 8S+30 to 6S is strongly Rrp6-dependent [35] . Binding of Rrp44 3’ to 5 . 8S+30 is not sensitive to inactivation of the S1 RNA binding domain , and this region is also bound by Rrp41 , Csl4 and Mtr4 , strongly indicating that the 3’ end of the 7S pre-rRNA is threaded through the channel . At 5 . 8S+30 we speculate that further processing via the channel is blocked by the RNA-protein structure of the mature 5 . 8S region ( see also [36] ) . The remaining ITS2 region must presumably then be extracted from the channel and re-targeted to Rrp6 . The peak of exosome association with the 5 . 8S+30 region may reflect the time required for this reorganization . The overall distribution of target RNAs ( Fig 1C ) suggested that binding of RNAPIII transcripts , in particular snR6 and tRNAs , is sensitive to the Rrp44-S1 mutation . To better characterize the interactions of individual RNAPIII transcripts with the channel and direct access pathways , we applied k-means clustering algorithms to the Rrp44-exo and Rrp44-exo-S1 data ( Fig 3A ) . Clustering is based on reads per million total mapped reads ( RPM ) for each RNAPIII transcript , averaged between two datasets . Clustering identified four groups of RNAPIII transcripts ( Fig 3A ) . For the large majority of RNAPIII transcripts ( Clusters 1–3 ) interactions with Rrp44 were strongly reduced by the S1 mutation , since substantially fewer reads were recovered with Rrp44-exo-S1 than Rrp44-exo . This indicates that these RNA species predominately access Rrp44 in a channel-independent manner . For Cluster 4 transcripts , relative recovery was higher with Rrp44-exo-S1 compared to Rrp44-exo . In the double mutant many RNAs show reduced binding , so species for which binding is unaltered will show an apparent increase in relative recovery . We conclude that Cluster 4 transcripts are insensitive to loss of the S1 RNA binding activity , indicating that they are predominately threaded . These species included the 5S rRNA , scR1 ( RNA component of the signal recognition particle , SRP ) , RNA170 and a subset of pre-tRNAs . Most tRNA isoacceptors are transcribed from multigene families in which the mature tRNAs are the same but the flanking , transcribed pre-tRNA regions are unique . Pre-tRNAs were differentiated by aligning sequences to a database composed only of pre-tRNA regions ( with 15 nt overhang on each end ) and using only uniquely mapped reads . For the mature tRNA regions , the genomic source cannot be differentiated within gene families , and they appear as a single entry in Fig 3A . Reads that match internal tRNA sequences may therefore be generated from mature tRNAs or precursors . However , given the large excess of mature tRNAs , these are likely to predominate and the reads are designated as “tRNA” . In the cluster analysis , mature tRNAs predominately fall in cluster 2 , which is characterized by particularly low binding to Rrp44-exo-S1 relative to Rrp44-exo . In contrast , pre-tRNAs and other RNPIII transcripts were distributed between different clusters . A subset of yeast pre-tRNAs contain introns ( shown in green in “intron” column of Fig 3A ) , but these were not clearly segregated from non-intronic pre-tRNAs . Notably , pre-tRNAs for the same isoacceptor did not systematically fall into the same clusters , suggesting they are not processed in the same way . The RNAPIII cluster data derived from Rrp44 crosslinking was compared to the association of the same transcripts with the exosome channel components , Csl4 and Rrp41 , exonuclease Rrp6 and cofactor Mtr4 ( Fig 3B ) . Here , binding to different proteins was averaged between two independent experiments and is presented relative to Rrp44-exo binding ( set to 1 ) . For Rrp41 CRAC , which reproducibly recovered relatively low numbers of reads , only the experiment with the highest number of reads was used , to reduce noise in the calculation . Association with Csl4 and Rrp41 was significantly higher for cluster 4 , consistent with threading of these substrates through the channel . Conversely , Csl4 and Rrp41 binding was the lowest for cluster 1 , in which transcripts are the most affected by the S1 mutation , with intermediate levels for clusters 2 and 3 . Association with Rrp6 was strong in all clusters suggesting that a major turnover pathway for tRNAs involves targeting to Rrp6 . This is consistent with a previous report that tRNAs show higher binding with Rrp6 than Rrp44 [14] . RNAs from cluster 4 showed more than two-fold higher relative binding by Rrp6 or Csl4 than Rrp44 . We speculate that these substrates can be degraded either by Rrp6 , which directly contacts Csl4 [18] , or by threading through the channel to Rrp44 . Association with Mtr4 was seen across all clusters , indicating that it functions as a cofactor in both channel-mediated and direct access to Rrp44 . The highest association with Mtr4 was on cluster 3 transcripts , with relative binding similar to Rrp6 and slightly greater than Rrp44-exo . This suggests that these substrates could use either Rrp6 or direct-access to Rrp44 for their degradation , facilitated by Mtr4 in both cases . The cluster analyses indicated that while pre-tRNAs show some differences in their interactions with Rrp44 , mature tRNAs were grouped together in cluster 2 and all were sensitive to the S1 mutation . This was further analyzed by comparing each tRNA and pre-tRNA on a 2D plot ( Fig 3C ) . Mature tRNAs showed markedly decreased binding to Rrp44-exo-S1 relative to Rrp44-exo , suggesting they directly access Rrp44 in vivo . Pre-tRNAs showed a much greater spread in their relative interactions with Rrp44-exo-S1 and Rrp44-exo , consistent with the cluster analysis . These data suggest that a subset of pre-tRNAs are threaded , whereas the majority use the direct access pathway . To independently assess dependence of RNAPIII transcripts belonging to different clusters on the Rrp44 S1 binding domain for exosome association , coprecipitation without crosslinking was performed with Rrp44-HTP , Rrp44-exo-HTP and Rrp44-exo-S1-HTP ( Fig 3D and 3E ) , followed by Northern blotting . The cluster analyses indicated that exosome association of snR6 ( U6 snRNA ) ( cluster 2 ) , RPR1 ( RNA component of RNase P ) ( cluster 3 ) and pre-tRNAProUGG ( pre-tP ( UGG ) ) ( cluster 2 , 3 , 4 ) should be dependent on the S1 RBD , whereas binding of 5S rRNA and scR1 ( cluster 4 ) was expected to be less S1-dependent . Northern hybridization confirmed that this is the case , with U6 , RPR1 and pre-tP ( UGG ) showing reduced coprecipitation with the Rrp44-exo-S1 mutant . The 3’ truncated 5S species ( 5S* ) is a well characterized exosome substrate and its binding was not clearly affected by the S1 mutation ( see also S3A–S3C Fig ) . We investigated the distribution of protein-binding sites in more detail across individual RNAPIII transcripts . As examples , ( S3A–S3C Fig ) shows these data for RPR1 , snR6 and 5S rRNA . On RPR1 , binding of Rrp44 , Mtr4 and Rrp6 showed strong accumulation close to the TSS ( S3A Fig ) . Their distribution was similar to the RNAPIII subunit Rpo31 [37] and was not clearly altered in the S1 mutant . This suggests a degree of pausing or stalling leading to release of nascent RNAs that are degraded via the exosome channel . In contrast , a strong peak at the 3’ end of the transcript does not correlate with high RNAPIII occupancy or binding to Rrp41 , and the association of this site with Rrp44 is lost in Rrp44-exo-S1 . This indicates that post-transcriptional 3’ processing or degradation of RPR1 involves direct access to Rrp44 , in cooperation with Mtr4 and Rrp6 . The pattern was broadly similar at the 3’-end of the U6 snRNA ( S3B Fig ) but distinct on the 5S rRNA ( S3C Fig; and see above ) . To assess the distribution of protein-binding sites across pre-tRNA and tRNA species , each individual tRNA gene was displayed in 2-dimensional plots ( Fig 4A–4F ) . Each line on the Y axis corresponds to a transcription unit . The X axis shows the absolute position on the gene aligned to the 3’ end of the mature tRNA . Metagene plots below the heat maps show the sum of binding across all individual genes . Rrp44-exo was strongly bound to both the 5’ and 3’ ends of pre-tRNAs ( Fig 4A ) , indicating that both 5’ and 3’ extended pre-tRNAs are Rrp44 substrates . In comparison , binding of Rrp44-exo-S1 was strongly reduced over the 3’ extended pre-tRNAs and the 3’ region of the mature tRNAs ( Fig 4B ) . The distribution of reads mapped to Rrp41 , located within the central channel , closely resembled Rrp44-exo-S1 , strongly indicating that ( pre- ) tRNA 5’ regions are threaded substrates . Distributions of reads recovered with the exosome cap component Csl4 , Rrp6 or Mtr4 were similar to each other but distinct from Rrp44 , with a sharp drop at the 3’ end of the mature tRNA region ( Fig 4D–4F ) . In summary , this indicates that 3’ ends of pre-tRNAs directly access the Rrp44 active site , while 5’ regions of ( pre- ) tRNAs are threaded through the central channel . In contrast , mature tRNAs are degraded via a different pathway involving Rrp6 , Mtr4 and perhaps Csl4 in the exosome cap . Notably , in published structures , Mtr4 directly contacts Rrp6 , which binds to Csl4 , and the route taken by substrates to the active site of Rrp6 is likely to involve interactions with the Rrp4/Rrp40/Csl4 ring [19 , 30] . Mtr4 is a component of the Trf4/5-Air1/2-Mtr4 polyadenylation complex ( TRAMP ) , which adds oligo ( A ) tails to RNAs prior to targeting them to the exosome for degradation . Mapping of reads that carry oligo ( A ) tails that are not encoded in the genomic sequence can therefore identify TRAMP targets . ( Pre- ) tRNA reads containing non-encoded oligo ( A ) tails represent ~23% of total reads for both Rrp44-exo and Rrp44-exo-S1 , but only ~2% for reads recovered with the RNAPIII subunit Rpo31 . Fig 4G presents the alignment of oligo-adenylated reads to tRNAs in comparison with all reads . Notably , the pre-tRNA 3’ regions that are bound by Rrp44-exo but not Rrp44-exo-S1 or Mtr4 also showed low oligoadenylation ( Fig 4A , 4B , 4E and 4G ) , indicating that they are not predominately TRAMP substrates . Strong binding of Rrp44-exo across pre-tRNA 5’ regions was very similar to Rpo31 ( Fig 4G ) [37] . RNAPIII pausing or slowing was previously observed over the box A internal promoter region [37] and we speculate that this can result in the release of truncated pre-tRNAs . These are apparently targeted by the TRAMP complex ( Fig 4E ) , oligoadenylated , and reach Rrp44 through the channel , since they are not clearly affected by the S1 mutation ( Fig 4G and 4H ) . Notably , short , truncated RNAPII transcripts are also targeted by the TRAMP and exosome complexes [38–40] ( and see below ) . For each tRNA , relative protein-binding to the 5’ region versus the 3’ region was calculated . The correlations between the binding profiles obtained for Rpo31 ( RNAPIII ) , Rrp44-exo and Rrp44-exo-S1 were determined by calculating Pearson coefficients ( S3D Fig ) . The total Rrp44-exo-S1 and oligo-A Rrp44-exo-S1 datasets are highly correlated with each other and with Rpo31 . Rrp44-exo oligo ( A ) + reads are substantially better correlated with RNAPIII than were total Rrp44-exo reads . This is consistent with the model that Rrp44 is involved in 2 events: A cotranscriptional ( pre- ) tRNA degradation pathway , in which TRAMP and the exosome channel play major roles ( Fig 4H ) , and a post-transcriptional processing pathway , in which ( pre- ) tRNAs directly access Rrp44 ( Fig 4I ) . Clustering analysis was also performed for RNAPII transcripts , allowing us to identify four clusters ( Fig 5A ) . Recovery of Cluster 1 transcripts was strongly reduced by the S1 mutation , indicating predominately channel-independent access to Rrp44 . Transcripts in Cluster 2 show a lower degree of dependence on the S1 RNA binding activity . Cluster 3 is enriched for transcripts on which relative binding was unaffected by the S1 mutation indicating that they can use either pathway . For Cluster 4 , relative recovery was higher with Rrp44-exo-S1 , indicating they are predominately threaded . The distribution of different classes of RNAPII transcripts was broadly similar across the different clusters , indicating that they do not systematically differ in their pathway dependence , in contrast to RNAPIII transcripts . The relative association of exosome channel components Csl4 and Rrp41 with the different RNAPII clusters correlated with higher relative binding to Rrp44-exo-S1 , consistent with increased threading ( Fig 5B ) . Rrp6 and Mtr4 exhibited high binding to cluster 3 and 4 transcripts , suggesting they would play a major role in their degradation , also consistent with the lower sensitivity of these groups to the Rrp44 S1 binding domain mutation . However , we anticipate that mRNAs , in particular , will interact with the exosome and its cofactors in different ways during nuclear pre-mRNA surveillance and cytoplasmic mRNA turnover . The box C/D and box H/ACA classes of snoRNA were present in all four clusters ( Fig 5A ) . Alignment of the 3’ ends of all box H/ACA snoRNAs ( Fig 6A ) shows strong Rrp44 binding downstream of the mature 3’ end , but no significant difference between Rrp44-exo and Rrp44-exo-S1 . In contrast , Rrp44 binding downstream of the 3’ end of box C/D snoRNAs ( Fig 6B ) is elevated in Rrp44-exo-S1 , and extends into the mature snoRNA . Examples of individual box C/D snoRNAs , U14 and U3 , are shown in Fig 6C and 6D . Similar read distributions over the 3’ flanking region of the U14 snoRNA were observed for the Rrp44-exo and Rrp44-exo-S1 CRAC datasets ( Fig 6C ) . Single nucleotide deletions ( indicating RT errors at the actual site of crosslinking ) were mapped to the same nucleotides , showing that both Rrp44 mutants contact U14 at sites 24 nt and 29 nt downstream of the mature 3’ end ( indicated by a solid line ) . Interestingly , even though the contact points are the same for Rrp44-exo and Rrp44-exo-S1 , reads are extended further upstream of the 3’ mature end in Rrp44-exo-S1 . This suggests that normal processing of pre-U14 involves both direct access and channel threading to Rrp44 . Supporting threading , the peak binding sites of Csl4 , Mtr4 and Rrp6 lie progressively further upstream of the Rrp44-exo and Rrp44-exo-S1 peaks . Their locations would be consistent with crosslinking to 3’ extended pre-U14 that is threaded through the exosome with a 3’ end located in the Rrp44 active site . RNA coprecipitation ( Fig 6E ) confirmed that binding of the major 3’ extended form of U14 is not lost when the S1 domain is disrupted , whereas a shorter extended form of U14 was not recovered in association with Rrp44-exo-S1 . In the case of the U3 snoRNA , previous analyses had shown that rapid cotranscriptional cleavage by Rnt1 ( RNase III ) is followed by binding of the Lsm2-8 complex and Lhp1 ( La ) to 3’ oligo ( U ) tracts [41–43] . Rrp44 , Mtr4 and Rrp6 all showed binding predominately over the 3’ region of the mature U3 sequence ( Fig 6D ) , whereas little association was observed for the channel proteins Csl4 or Rrp41 . Crosslinking of Rrp44-exo also extended into the 3’ flanking region and encompassed the oligo ( U ) tracts , whereas crosslinking of Rrp44-exo-S1 was limited to the region 5’ to the oligo ( U ) tracts ( Fig 6D; see zoom in ) . RNA coprecipitation of 3’ extended U3 was lost in the Rrp44-exo-S1 mutant ( Fig 6F ) , indicating direct access to Rrp44 . We conclude that this is the major pathway for initial 3’ maturation of U3 snoRNA following Rnt1 cleavage , which is likely to be cotranscriptional . All exosome components , as well as Mtr4 , showed additional strong association over the 5’ region of exon 2 of U3 . The lack of effect of the S1 mutation plus association with the core exosome components Rrp41 and Csl4 indicates that these U3 regions are threaded through the channel for degradation , possibly of the mature snoRNA . Overall snoRNA recovery was not strongly affected by the S1 mutation ( S4A Fig ) . The 3’ ends of all snoRNAs were aligned for each of the proteins along the threaded path to Rrp44 ( S4B Fig ) . This revealed the displacement in binding from the top to bottom of TRAMP-exosome complex: Mtr4 occupied the most upstream position , followed by Csl4 and Rrp41 , with Rrp44 most downstream . These results indicate that snoRNA turnover mainly proceeds via the threaded pathway . In principle , a similar distribution of factors might be expected on other threaded substrates , however , this is only clearly resolved on strongly expressed substrates with a well-defined site of exosome stalling , such as that predicted to be induced by the snoRNA-associated proteins . Binding of the top 1000 expressed mRNAs to Rrp44-exo and Rrp44-exo-S1 was compared ( S5A Fig ) . Their distribution close to the diagonal showed that they are predominately threaded through the channel . On protein-coding genes , Rrp44-exo showed a pronounced peak of TSS proximal binding and this profile was not clearly affected by loss of S1 RNA binding ( S5B Fig ) . Among the few mRNAs with strongly reduced recovery in Rrp44-exo-S1 was INO1 , which encodes Inositol-3-phosphate synthase [44 , 45] . Rrp44-exo binding was strikingly high immediately upstream of the start codon of the gene , whereas these reads were lost completely in Rrp44-exo-S1 ( S5C Fig ) . Other exosome subunits showed few if any hits across INO1 . The INO1 gene has a 5’ UTR region of 437 nt , which is exceptionally long for yeast . This highly unusual structure suggests that the region is involved in regulation , and we speculate that direct access to Rrp44 might contribute to the regulation of INO1 mRNA levels . The Rrp44-exo exosome complex protects some 33 nt of RNA in vitro [10] . However , this protection is lost in complexes that include a mutant form of Rrp41 with four reverse-charge point mutations at the RNA entry site ( K62E , S63D ) and exit site ( R95E , R96E ) of the channel [10] . In an attempt to further define substrates reaching Rrp44 through the channel , we performed CRAC on wild type Rrp44-HTP in a strain where Rrp41 carries these four point mutations ( Rrp41-channel ) . Unexpectedly , comparison of wild type Rrp44-HTP with the Rrp44-HTP , Rrp41-channel strain showed clear crosslinking differences only for mRNAs ( Fig 7A ) , with overall mRNA binding substantially reduced by channel mutation . In the crystal structure of an Rrp6-associated nuclear exosome complex , the channel was seen to be widened relative to the complex lacking Rrp6 [18] . We therefore postulate that the Rrp41-channel mutation may inhibit RNA passage through the cytoplasmic exosome , but not through the widened channel in the nuclear complex . This model is in agreement with our conclusion that only cytoplasmic exosome substrates , which are predominately mRNAs , are affected by the Rrp41-channel mutation . Analysis of the distribution of Rrp44 binding over the top 1000 expressed mRNAs showed that the TSS proximal peak of association was reduced , but not abolished by the Rrp41-channel mutation ( Figs 7B , S5D and S5E ) . This 5’ peak was previously attributed to the presence of truncated transcripts generated through premature transcription termination [38–40] . These transcripts are also strongly bound by the nuclear TRAMP polyadenylation complex , and are expected to be degraded in the nucleus . TRAMP substrates are characterized by non-templated oligo ( A ) tails and we therefore filtered Rrp44 hits for the presence of 3’ oligo ( A ) tracts . Strikingly , crosslinking of the oligo ( A ) + transcripts showed no reduction in the Rrp41-channel mutant ( Fig 7C ) . This supports the model that reduced Rrp44 association in the Rrp41-channel mutant is a specific feature of cytoplasmic , but not nuclear exosome substrates . For each mRNA species as well as the CUT ( Cryptic Unstable Transcripts ) and SUT ( Stable Unannotated Transcripts ) classes of ncRNAs , we determined the ratio of Rrp44 binding in Rrp41 and Rrp41-channel strains as a measure of sensitivity to channel narrowing ( Fig 7D–7F ) . Comparison of binding of the nuclear-specific exosome cofactor Mtr4 with mRNA sensitivity to channel narrowing revealed a striking correlation ( Fig 7G–7I ) . Those mRNAs that were least affected by Rrp41-channel ( ratio <0 . 8 ) were highly bound by Mtr4 , whereas mRNAs with Rrp44 association that were strongly reduced in the channel mutant showed progressively decreased Mtr4 association . As examples , hit distributions along two mRNAs are presented ( Fig 7H and 7I ) . TDH3 , encoding the glyceraldehyde-3-phosphate dehydrogenase , showed almost no binding for Mtr4 or Rrp44 ( Rrp41-channel ) , whereas Rrp44 alone was robustly bound , strongly indicating cytoplasmic turnover ( Fig 7H ) . In contrast , RPS14B was targeted by Mtr4 and by Rrp44 in both Rrp41 wild-type and Rrp41-channel strains . Coupled with the recovery of hits across the intron , this indicates that both spliced and unspliced forms of RPS14B pre-mRNA can be degraded in the nucleus ( Fig 7I ) . Based on this insight , mRNAs were analyzed to identify species for which Rrp44 binding was insensitive to the Rrp41-channel mutation , indicating predominant nuclear degradation ( Figs 7D , S5D and S5E ) . This identified a subset of nuclear-degraded mRNAs ( above line in Fig 7D ) . Notably , these include mRNAs encoding four factors implicated in nuclear pre-mRNA surveillance , Nrd1 , Nab3 , Hrp1 and , most strikingly Dbp2 ( S5 Table ) [39 , 46–49] . Moreover , Dbp2 and Nrd1 are auto-regulated by nuclear RNA processing [50 , 51] , and this may also be the case for Nab3 and Hrp1 . NRD1 expression is auto-regulated through transcription termination and degradation by the exosome , induced by Nrd1 binding . The peak of Rrp44 binding on NRD1 corresponded well with the locations of Nrd1-binding sites [52] and consensus Nrd1-binding motifs ( UGAUG ) ( S5F Fig ) . The mRNA encoding the transcription factor Tye7 is regulated by nuclear RNA surveillance [53] and was also strongly affected ( S5 Table ) . The CUT class of ncRNAs are well characterized as nuclear exosome substrates and showed little sensitivity to the Rrp41-channel mutation ( Fig 7E ) . In contrast , members of the more stable SUT class of ncRNAs showed an intermediate behavior between mRNAs and CUTs , with different species showing either increased or reduced Rrp44 association in the Rrp41-channel mutant ( Fig 7F and S6 Table ) . We conclude that the Rrp41-channel mutation offers a tool to distinguish nuclear and cytoplasmic sites of degradation for major exosome substrates .
The central channel of the nuclear exosome is widened in vivo relative to the cytoplasmic complex , presumably reflecting allosteric changes induced by Rrp6 binding [18] . In consequence , Rrp41 charge-reversal mutations inside the channel [10] inhibited the degradation only of cytoplasmic ( non-Rrp6 associated ) substrates . This allowed us to distinguish mRNAs and ncRNAs that are preferentially degraded in the nucleus and cytoplasm . In the nucleus , our data implicate Mtr4 in targeting substrates to both the threaded and direct-access pathway , strongly indicating interactions with the exosome that have yet to be observed in vitro . Many RNA species showed clear preferences for either the threaded or direct-access pathway . However , these generally appeared not to be absolute requirements . We speculate that this redundancy reflects the outcome of selective pressure . The exosome degrades and/or processes thousands of different substrates , including large numbers of ncRNAs that tend to change rapidly during evolution . The system may therefore have been selected for versatility and redundancy to allow the efficient handling of highly diverse RNA-protein complexes and RNA structures .
All sequence data are available from GEO under accession number GSE90647 . | In all organisms , diverse classes of RNA perform a wide range of functions , including programing of protein synthesis , transcriptional and post-transcriptional regulation of gene expression and forming key structural and functional components for pre-mRNA splicing , ribosome synthesis and translation . In consequence , control of the quantity and quality of RNA synthesis and maturation is of key importance . In Eukaryotes , the RNA exosome complex plays a major role in both degradation of aberrant RNA molecules and maturation of functional RNAs . The exosome has a barrel-like structure , with a central channel through which RNAs are threaded to the exonuclease active site of Rrp44 . However , in a different exosome conformation , RNAs can also directly access the Rrp44 active site . Here we report the interaction of RNAs with different exosome components and its major cofactor Mtr4 in wild type and mutant strains of yeast . We identified numerous RNAs with clear channel specificities , whereas channel-threaded and direct access routes cooperated for many other substrates . The pathway used by RNA to access Rrp44 reflects the nature and structure of the RNA and whether it is addressed to the exosome for accurate processing or complete degradation , helping us to understand these crucial but highly complex pathways in vivo . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Availability",
"of",
"data",
"and",
"materials"
] | [
"transfer",
"rna",
"small",
"nucleolar",
"rnas",
"chemical",
"bonding",
"chemical",
"characterization",
"vesicles",
"gene",
"regulation",
"messenger",
"rna",
"cellular",
"structures",
"and",
"organelles",
"physical",
"chemistry",
"research",
"and",
"analysis",
"methods",
"rna",
"structure",
"gene",
"expression",
"chemistry",
"cross-linking",
"binding",
"analysis",
"molecular",
"biology",
"ribosomes",
"biochemistry",
"rna",
"ribosomal",
"rna",
"cell",
"biology",
"nucleic",
"acids",
"exosomes",
"genetics",
"biology",
"and",
"life",
"sciences",
"physical",
"sciences",
"non-coding",
"rna",
"macromolecular",
"structure",
"analysis"
] | 2017 | Transcriptome-wide analysis of alternative routes for RNA substrates into the exosome complex |
Ischemic stroke is the second leading cause of death worldwide . Only one moderately effective therapy exists , albeit with contraindications that exclude 90% of the patients . This medical need contrasts with a high failure rate of more than 1 , 000 pre-clinical drug candidates for stroke therapies . Thus , there is a need for translatable mechanisms of neuroprotection and more rigid thresholds of relevance in pre-clinical stroke models . One such candidate mechanism is oxidative stress . However , antioxidant approaches have failed in clinical trials , and the significant sources of oxidative stress in stroke are unknown . We here identify NADPH oxidase type 4 ( NOX4 ) as a major source of oxidative stress and an effective therapeutic target in acute stroke . Upon ischemia , NOX4 was induced in human and mouse brain . Mice deficient in NOX4 ( Nox4−/− ) of either sex , but not those deficient for NOX1 or NOX2 , were largely protected from oxidative stress , blood-brain-barrier leakage , and neuronal apoptosis , after both transient and permanent cerebral ischemia . This effect was independent of age , as elderly mice were equally protected . Restoration of oxidative stress reversed the stroke-protective phenotype in Nox4−/− mice . Application of the only validated low-molecular-weight pharmacological NADPH oxidase inhibitor , VAS2870 , several hours after ischemia was as protective as deleting NOX4 . The extent of neuroprotection was exceptional , resulting in significantly improved long-term neurological functions and reduced mortality . NOX4 therefore represents a major source of oxidative stress and novel class of drug target for stroke therapy .
Ischemic stroke has outstanding medical relevance as it is the second leading cause of death in industrialized countries [1] . Due to the aging of the population , the incidence of stroke is projected to rise even further in the future [2] . Despite tremendous research activity , with more than 100 clinical trials in human stroke patients [3] , only one therapy approved by the United States Food and Drug Administration is available , i . e . , thrombolysis using recombinant tissue plasminogen activator ( rt-PA ) . However , the efficacy of rt-PA on functional outcomes is moderate at best , and more than 90% of all stroke patients must be excluded from rt-PA treatment because of over 25 labeled contraindications . Therefore , there is an unmet need for more effective therapies in acute stroke . Although a plethora of drugs for the treatment of acute stroke are effective in animal models , their translation into clinical practice has completely failed [3] , [4] . As a result , many pharmaceutical companies have withdrawn from drug discovery in this area . To overcome this lack of clinically effective neuroprotective drugs , innovative strategies are urgently needed to identify pathways that can be targeted with innovative therapies [5] . Higher quality study designs are also required [6] , [7] . One such high-potential pathway in ischemic stroke may be the occurrence of oxidative stress , i . e . , the increased occurrence of reactive oxygen species ( ROS ) above physiological levels . Oxidative stress has been suggested for many years to cause tissue damage and neuronal death . The toxicity of ROS can be further increased by nitric oxide to produce reactive nitrogen species such as peroxynitrite ( ONOO− ) , a molecule that causes oxidation and nitration of tyrosine residues on proteins [8] . Disappointingly , there is no conclusive evidence of a causal link between oxidative stress and the development of disease , and there is no successful therapeutic application targeting oxidative stress . To date , clinical attempts to scavenge ROS by applying antioxidants did not result in clinical benefit [9] or even caused harm [10] , [11] . However , the characterization of the relevant enzymatic sources of oxidative stress may allow therapeutic targeting of oxidative stress by preventing the formation of ROS in the first place , instead of scavenging ROS after they have been formed . A potential source of ROS are NADPH oxidases , the only known enzyme family that is only dedicated to ROS production [12] . Four rodent genes of the catalytic subunit NOX , Nox1 , Nox2 , Nox3 , and Nox4 , have been identified , of which Nox1 , Nox2 , and Nox4 are expressed in the vasculature . NOX4 is the most abundant vascular isoform; its expression is even higher in cerebral than in peripheral blood vessels [13] and , further , induced in stroke [14] . Therefore , we hypothesized that NOX4 is the most relevant source of ROS in stroke . To test this hypothesis , we generated constitutively NOX4-deficient ( Nox4−/− ) mice and directly compared them to NOX1-deficient ( Nox1y/− ) and NOX2-deficient ( Nox2y/− ) mice . NOX4 has been implicated in the regulation of systemic and hypoxic vascular responses . Therefore , we had to exclude systemic vascular effects of NOX4 deletion on blood pressure , which may affect stroke outcome independent of a specific neuronal or neurovascular mechanism . Finally , to examine the therapeutic potential of NOX4 as a drug target , we infused the specific NADPH oxidase inhibitor VAS2870 [15] after ischemia , thus mirroring the clinical scenario .
Because NOX4 mRNA is expressed at higher levels in cerebral than in peripheral blood vessels [13] and is induced in stroke [14] , we first sought to validate these data not only at the mRNA but also at the protein level . In all experiments , we followed current guidelines defining methodological standards for experimental stroke studies [4] , [6] , [7] , [16] , [17] . Here we chose a model of acute ischemic stroke in which mice are subjected to transient middle cerebral artery occlusion ( tMCAO ) . This disease model is thought to involve oxidative stress and an induction of Nox4 expression [18] . Indeed , expression of NOX4 mRNA was significantly higher 12 h and 24 h after tMCAO in the basal ganglia and neocortex of wild-type mice than in sham-operated controls , in which basal NOX4 expression was low ( Figure 1A ) . This result was validated by immunohistochemistry using a specific NOX4 antibody . We detected a stronger staining in neurons and cerebral blood vessels in wild-type mice subjected to tMCAO than in sham-operated controls . Although immunohistochemistry is not quantitative , this finding suggests higher levels of NOX4 protein ( Figure 1B ) . Importantly , NOX4 staining was also stronger in brain samples from stroke patients . Although NOX4 was barely detectable in healthy brain regions , clear positive labeling of NOX4 was seen in neurons and vascular endothelial cells from the forebrain cortex of stroke patients . This finding was confirmed by double labeling for NeuN ( a neuronal marker ) or von Willebrand factor ( an endothelial marker ) and NOX4 in brain tissue ( Figure 1B ) . These data indicate that NOX4 protein is induced during brain ischemia in mice , and this observation would be in agreement with a major functional role for NOX4 in ischemic stroke . Our limited observations in a small number of human cases provide some support to the hypothesis that these processes are also important in human stroke . We first subjected 6- to 8-wk-old male Nox4−/− mice to tMCAO and , after 24 h , assessed infarct volumes by staining brain sections with 2 , 3 , 5-triphenyltetrazolium chloride ( TTC ) ( Figure 2A ) . Infarct volumes were significantly smaller , by approximately 75% , in male Nox4−/− mice than in sex-matched wild-type controls ( 25 . 5±14 . 8 mm3 versus 78 . 7±19 . 5 mm3 , respectively ) . The smaller infarct volume was functionally relevant: compared with wild-type mice , Nox4−/− mice had significantly better overall neurological function ( Bederson score 1 . 2±0 . 7 in Nox4−/− mice versus 3 . 7±1 . 1 in wild-type mice ) as well as better basal motor function and coordination ( grip test score 4 . 3±0 . 7 in Nox4−/− mice versus 1 . 7±1 . 3 in wild-type mice ) 24 h after tMCAO ( Figure 2B ) . Gender can significantly influence stroke outcome in rodents [4] , [16] , [17] . Therefore , we also subjected female Nox4−/− mice to 60 min of tMCAO . In line with the results in male mice , Nox4-deficient female mice also developed significantly smaller infarctions ( 30 . 1±6 . 7 mm3 versus 89 . 5±22 . 2 mm3 , respectively ) and less severe neurological deficits ( p<0 . 001 ) than female controls ( Figure 2A and 2B ) . Histological analysis revealed that all infarcts in Nox4−/− mice were restricted to the basal ganglia ( arrow in Figure 2A and 2C ) , whereas in wild-type mice , the neocortex was also consistently affected . Serial magnetic resonance imaging of living mice up to 6 d after stroke showed that in Nox4−/− mice the infarct volume did not increase over time , thus indicating that deletion of the Nox4 gene provides sustained protection against stroke ( Figure 2C ) . Moreover , infarcts always appeared hyperintense on blood-sensitive constructive interference in steady state ( CISS ) sequences . Hypointense areas , which typically indicate intracerebral hemorrhage , were absent from Nox4−/− mice and wild-type controls . This finding excludes the possibility of an increased rate of bleeding complications caused by Nox4 deficiency . To establish any potential specificity of this function for NOX4 compared to NOX1 and NOX2 in stroke , we carried out identical experiments in 6- to 8-wk-old Nox1y/− and Nox2y/− mice . However , in contrast to Nox4−/− mice , we observed no protection in these animals , neither in terms of infarct volumes nor on functional outcomes on day 1 after tMCAO , even with large subject sample sizes ( n = 19 for Nox2y/− mice , p>0 . 05; Figure 2A ) . Ischemic stroke is usually a disease of the elderly and , consequently , one should verify any stroke-protective effects observed in young adult laboratory animals also in an older cohort [4] , [16] , [17] . Indeed , 18- to 20-wk-old Nox4−/− mice also developed significantly smaller brain infarctions ( 27 . 8±15 . 1 mm3 versus 81 . 8±19 . 0 mm3 , respectively ) and less severe neurological deficits than age-matched controls , thereby confirming our results in young animals ( Figure 2B ) . We also determined the functional outcome and mortality of 6- to 8-wk-old male Nox4−/− mice and matched wild-type controls over a longer time period after ischemic stroke ( Figure 2D ) . Five days after 60 min of tMCAO , 15 of 15 wild-type mice ( 100% ) had died , which is in line with previous reports [19] . In contrast , seven of ten Nox4−/− mice ( 70% ) survived until day 5 , and five of these were still alive after 1 wk ( p = 0 . 0039 ) ( Figure 2D ) . In line with these findings , Nox4-deficient mice showed significantly better Bederson scores than controls over the whole observation period , and neurological deficits remained low until day 7 ( Figures 2D and S4 ) . According to the current experimental stroke guidelines [4] , [16] , [17] , any protective effect also requires evaluation in models of both transient and permanent ischemia . We therefore subjected Nox4−/− mice to filament-induced permanent middle cerebral artery occlusion ( pMCAO ) , a procedure in which no tissue reperfusion occurs ( Figure 2E ) . In the absence of Nox4 , infarct volumes ( 66 . 7±28 . 6 mm3 versus 120 . 1±15 . 6 mm3 , p<0 . 05 ) and neurological deficits ( Bederson score 2 . 3±1 . 7 versus 3 . 4±0 . 8 , p<0 . 05 ) at day 1 after pMCAO were significantly reduced compared with those in wild-type controls , although to a lesser extent than they were in the tMCAO model ( Figures 2E and S5 ) . Brain infarctions following filament-induced pMCAO are large , and the infarct borders are often not very well defined , which limits the accuracy of any estimation on infarct volumes . We therefore used another model of permanent stroke , cortical photothrombosis , to further verify our findings . Here , the lesions are restricted to the cortex and highly reproducible in size and location . Moreover , photothrombosis has been shown to induce early and profound ROS formation and blood-brain-barrier leakage [20] , [21] , two key readout parameters of the present investigation . Importantly , photothrombosis-induced infarct volumes were as reduced in Nox4−/− mice relative to wild-type mice ( 3 . 3±4 . 6 mm3 versus 25 . 0±12 . 8 mm3 , respectively , a difference of 86 . 8%; Figure 2F ) as they were in the tMCAO model . Based on the physiological distribution of NOX4 in kidney [22] , lung [23] , and aorta [24] , as well as cell biology data obtained using small interfering RNA approaches [23] , one would predict basal phenotypes in a Nox4−/− mouse , such as arterial hypotension , reduced hypoxic pulmonary hypertension , and altered renal function . Importantly , these effects could potentially modulate or interfere with stroke outcome even in the absence of a specific neuronal or neurovascular mechanism . Surprisingly , systemic elimination of Nox4 did not result in any apparent abnormal vascular phenotype ( Text S1; Figures S1 and S2; Table S1 ) . In particular , blood pressure was normal , and hypoxic pulmonary hypertension still occurred despite a 20-fold induction of NOX4 in wild-type animals [23] . In contrast , Nox1- and p47phox-deficient mice ( a Nox2 subunit ) have a lower basal blood pressure , and their blood-pressure response to angiotensin II is reduced [25]–[27] . Our data suggest that any phenotype caused by deleting Nox4 , unlike those caused by deleting Nox1 and Nox2 , would indeed be brain-specific . Next we sought to elucidate the underlying mechanisms of this NOX4-specific neurotoxicity in stroke . NOX4 can form superoxide or H2O2 , which can interact with nitric oxide to form reactive nitrogen species . Therefore , we stained brain sections with broad-spectrum indicators of oxidative/nitrative stress , i . e . , dihydroethidium [28] and nitrotyrosine [8] . At 12 h and 24 h after tMCAO , brains from wild-type mice exhibited a significantly larger amount ( by a factor of 2 . 5–3 . 5 ) of ROS in neurons than brains from sham-operated animals , as quantified by dihydroethidium staining ( Figure 3A ) . Neurons from Nox4−/− mice , in contrast , showed only very small ischemia-induced increases in ROS relative to those in sham-operated controls ( p>0 . 05 ) . ROS formation from neurons after 24 h was also significantly reduced in Nox4−/− mice subjected to pMCAO ( Figure S6 ) . Because the dihydroethidium stain may also indicate oxidative chemistry events , including formation of ONOO− and nitration of protein tyrosine residues [8] , we analyzed the extent of protein nitration in Nox4−/− and wild-type mice subjected to tMCAO . In agreement with our findings on the generation of ROS , tissue nitration occurred to a lesser extent in ischemic brains from Nox4−/− mice than in those from wild-type controls ( Figure 3B ) . Oxidative chemistry events such as the formation of ROS and peroxynitrite , as detected by dihydroethidium staining and nitrotyrosine immunolabeling , can induce neuronal apoptosis , which is a well-established mechanism of tissue damage in ischemic stroke [29] , [30] . Indeed , superimposed TUNEL and NeuN immunolabeling revealed widespread apoptosis of neurons in wild-type mice 24 h after stroke onset ( Figure 3C ) . In contrast , the number of apoptotic neurons in Nox4−/− mice subjected to tMCAO was significantly lower , and the basal apoptotic turnover rate in Nox4−/− mice fell within the range found in sham-operated mice ( p>0 . 05 ) ( Figure 3C ) . We also detected NOX4 in cerebral blood vessels ( Figure 1B , white arrow indicates endothelial cells ) . Therefore , we hypothesized that Nox4 deficiency also influences the disruption of the blood–brain barrier and edema formation mediated by ROS [31] . Integrity of the blood–brain barrier was preserved in Nox4−/− mice on day 1 after tMCAO . This finding correlated with significantly less brain edema in Nox4−/− mice than in wild-type controls , as assessed by the extent of extravasation of Evans blue stain ( 8 . 0±5 . 9 mm3 in Nox4−/− mice versus 96 . 2±5 . 9 mm3 in wild-type mice ) . Importantly , almost no brain edema was seen in the brain regions where infarcts were regularly present in Nox4−/− mice ( basal ganglia; Figure 3D , area delineated by the broken white line ) . This result indicates that the lesser edema seen in the Nox4−/− mice was a specific phenomenon and mechanistically relevant but was not due to smaller infarct volumes . Finally , we wanted to examine whether these genetic insights into the biology of oxidative stress in stroke and the role of NOX4 in general can be translated into a therapeutic intervention . Importantly , this intervention would have to be effective post-stroke and ideally it would be pharmacological . Therefore , we examined the efficacy of a validated , low-molecular-weight NADPH oxidase inhibitor , VAS2870 [15] , [32]–[34] , in vital brain slices and in vivo . VAS2870 equally inhibits the ROS-generating activity of all NOX subunits , i . e . , NOX1 , NOX2 , and NOX4 . Vital brain slices [35] taken from wild-type mice 12 h after tMCAO produced significantly less ROS after pretreatment with 10 µM VAS2870 , as did brain slices from untreated Nox4−/− mice ( Figure 4A ) . Importantly , incubating ischemic slices from Nox4−/− mice with VAS2870 had no additional inhibitory effect on superoxide formation ( Figure 4A ) . This finding further underlines the extraordinary role of NOX4 in generating oxidative stress during the course of ischemic stroke , while other NOX isoforms such as NOX1 or NOX2 are obviously less relevant . To determine whether VAS2870 is also active when applied in vivo , we administered 2 mg of VAS2870 intrathecally to wild-type mice 2 h and 12 h after tMCAO . This experimental therapeutic approach significantly reduced brain infarct volumes ( 20 . 7±4 . 0 mm3 in VAS2870-treated mice versus 82 . 4±6 . 4 mm3 in vehicle-treated controls ) and significantly improved neurological function , to the same extent as observed for the deletion of Nox4 in mice ( Figure 4B and 4C ) . Moreover , less oxidative stress was detected in ischemic brains from VAS2870-treated animals than in those from vehicle-treated controls ( Figure 4D ) . Again , post-stroke application of VAS2870 to Nox4−/− mice had no additive neuroprotective or superoxide-lowering effect compared to the outcomes in wild-type animals treated with VAS2870 or untreated Nox4−/− mice ( Figure 4B–4D ) . This observation is consistent with our ex vivo findings in ischemic brain slices and reaffirms that NOX4 rather than NOX1 or NOX2 is critically involved in the pathophysiology of ischemic stroke . Another , less specific inhibitor that also targets molecules other than NADPH oxidases [36] , [37] , apocynin , had no effect on infarct size or functional outcome when given post-stroke and did not reduce the formation of ROS in vivo ( Figure 4B and 4C ) . To further examine whether the neuroprotective effect observed in Nox4−/− mice is specifically related to reduced ROS formation and not due to other nonspecific or developmental defects , we performed a rescue experiment by restoring cerebral ROS levels in Nox4−/− mice during the course of ischemic stroke by applying exogenous H2O2 ( Figure 4B–4D ) . Indeed , intrathecal administration of H2O2 rescued the phenotype in Nox4−/− mice , and infarct volumes , functional deficits , and stroke-induced ROS formation returned to the levels observed in wild-type mice ( Figure 4B–4D ) .
Here we identify NOX4 as a relevant molecular source of oxidative stress in cerebral ischemia , including some cases of human stroke . Our data suggest that NOX4-mediated oxidative stress leads to neuronal damage via leakage of the blood–brain barrier and neuronal apoptosis—two pathophysiological hallmarks of ischemic stroke . The extent of neuroprotection conferred by the absence of NOX4 in male and female Nox4−/− mice was exceptional and preserved in old animals . Importantly , the outcomes of these genetic experiments were mimicked when we pharmacologically inhibited NADPH oxidases within a clinically relevant time after induction of stroke . We consider this a key finding for the wider concept of oxidative stress , which might also be of relevance for other disease states , such as neurotrauma and neuroinflammation , where oxidative stress , blood–brain barrier damage , and neurotoxicity are involved . Rather than focusing on antioxidants and the disappointing outcomes of their application , the identification of the relevant source of oxidative stress and preventing its formation may represent an approach with clinical potential . The hypothesis that free radicals are involved in acute ischemic stroke and account for secondary infarct growth dates back to the 1970s [38] but has remained unproven [38] , [39] . The extent of neuroprotection that we observed is exceptional compared with that seen in many other pre-clinical stroke studies , in which the reduction of infarct size usually does not exceed 30%–40% [40] . Such moderate reductions of infarct volume have not translated into improvement of neurological status [3] . Most notably , continuous assessment of functional deficits until 7 d after stroke revealed that Nox4-null mice indeed showed a better amplitude rather than simply altered kinetics of recovery . This protection in Nox4−/− mice was further underlined by a significantly reduced post-stroke long-term mortality . Secondary infarct growth mediated for example by edema formation or hemorrhagic transformation is common during the course of brain ischemia and can lead to worsening of neurological symptoms [39] . Serial magnetic resonance imaging revealed that infarcts in Nox4−/− mice remain small , even at later stages of infarct development , and signs of intracerebral hemorrhage were consistently absent , thus indicating that NOX4 inhibition is likely to be safe and persistently effective . A plethora of compounds have provided neuroprotection in animal models of brain ischemia , but they all failed in human clinical trials [4] . This translational roadblock has been attributed mainly to inadequate pre-clinical study design and severe methodological shortcomings . Important confounding factors are a lack of randomization or rater-blinded evaluation of study results , and use of only one stroke model [16] . Strictly adhering to current expert recommendations for basic stroke trials , we here demonstrate that in the absence of NOX4 , brain tissue can be salvaged after ischemia or reperfusion injury ( as occurs in the tMCAO model ) . Most importantly , neuroprotection was preserved in old male and female Nox4−/− mice as well as in Nox4−/− mice subjected to permanent ischemia ( i . e . , cortical photothrombosis or pMCAO ) . Compared to in the tMCAO model , however , the reduction of infarct size in the pMCAO model was less pronounced though still significant . Distinct pathomechanisms that can be positively influenced only in the presence of tissue reperfusion , i . e . , after tMCAO but not pMCAO , such as progressive thrombus formation in the cerebral microvasculature [41] , might account for this quantitative difference . Indeed , preliminary results suggest that clotting is attenuated in the cerebral vessels of Nox4−/− mice subjected to tMCAO but not pMCAO ( unpublished data ) . Clearly , elimination of NOX4 remains beneficial in the absence of arterial recanalization , a condition frequently observed in human stroke . In our experiments , deficiency of NOX1 or NOX2 had no impact on infarct size or functional outcome after tMCAO . Although others have described protective effects of NOX2 deficiency after experimental stroke [42]–[44] , we could not reproduce those findings . The exact reasons for this discrepancy are unclear at present . Differences in the experimental protocols and middle cerebral artery occlusion times , which varied between 30 min and 120 min in previous investigations , might play a role here [42]–[44] . In contrast to these previous studies , however , we used especially high numbers ( n = 19 ) of Nox2y/− mice to verify our findings . Moreover , type-II ( beta ) error of the differences between infarct volumes in Nox2y/− mice and wild-type controls was only 7% in our study ( 93% power , respectively ) ( Tables S3–S5 ) , which is a very powerful result compared to the positive reports on Nox2 deficiency in cerebral ischemia [42]–[44] as well as to many other experimental stroke studies in general [4] , [45] . Moreover , the fact that VAS2870 , which specifically inhibits NADPH oxidases , could not further decrease infarct size and ROS formation in Nox4−/− mice ex vivo and in vivo ( Figure 4 ) clearly argues against a major role of NOX1 or NOX2 in the pathophysiology of acute ischemic stroke . Finally , protein expression levels of NOX1 and NOX2 were almost unchanged in the brains of Nox4−/− mice ( Figure S3C ) , underlining that the profound neuroprotection we observed is mediated by deficiency or blockade of NOX4 itself and not by secondary effects . Nevertheless , we cannot completely rule out contributions of other sources of ROS . Referring to this , Block et al . recently reported that a functional NOX4 is present and regulated in mitochondria , indicating the existence of a hitherto undescribed source of mitochondrial ROS [46] . An unprecedented need exists for more effective therapies for acute stroke , the second leading cause of death worldwide [1] . We have demonstrated that pharmacological inhibition of NADPH oxidases using the specific NADPH oxidase inhibitor VAS2870 [15] , [32]–[34] protects mice from brain ischemia within a clinically meaningful 2-h time window . In contrast , the commonly used organic compound apocynin may not be a NOX inhibitor in vascular cells but rather acts as a nonspecific antioxidant [36] . It also inhibits Rho kinase inhibitor [37] , an activity that increases its nonspecific actions . If apocynin inhibits NADPH oxidases at all , it supposedly blocks the migration of the cellular NADPH oxidase complex subunit p47phox to the membrane , thus interfering with assembly of the functional NOX complex [47] . Therefore , it is unlikely to inhibit the NOX4-containing NADPH oxidase , which acts independently of any cytosolic subunits [12] . Indeed , in our experiments , application of apocynin had no effect on the formation of ROS or of functional outcome after experimental stroke in vivo . In summary , we have demonstrated that NOX4-derived oxidative stress is a crucial player in the pathophysiology of acute ischemic stroke , while Nox4 deletion does not affect basal vascular or renal function . Nox4 gene reconstitution experiments in Nox4−/− mice and studies of the effects of different , structurally unrelated NOX inhibitors—should they become available—would be desirable to further substantiate the causality between NOX4 deficiency and protection from cerebral ischemia . Pharmacological inhibition of NADPH oxidases using specific compounds may also pave new avenues for the treatment of ischemic brain injury in humans . Because NADPH oxidase–mediated production of ROS may represent a general mechanism of neurotoxicity , our findings may extend to other ischemic disorders and neurodegenerative or inflammatory diseases . Further studies in relevant disease models are warranted .
Specimens from patients who had experienced a stroke were collected during routine autopsy at the Department of Neuropathology , University of Würzburg , Germany . Detailed study characteristics are provided in Table S2 . We strictly followed the recent international expert recommendations for conducting research in mechanism-driven basic stroke studies [4] , [6] , [7] , [16] , [17] , [40] . If not otherwise mentioned , we performed 60 min of tMCAO in 6- to 8-wk-old male mice weighing 20–25 g , as described previously [48] , [49] . To exclude age- and gender-specific effects , 18- to 20-wk-old male and 6- to 8-wk-old female mice were used in some subgroups . For pMCAO the occluding filament was left in situ until sacrificing the animals [41] . At 2 h and12 h after the induction of tMCAO , subgroups of wild-type mice or Nox4−/− mice were randomly selected to receive either 2 mg of the NOX-specific inhibitor VAS2870 ( Vasopharm GmbH [32] , [33] ) or carrier solution ( 10% dimethyl sulfoxide , Sigma ) intrathecally , as described previously [50] . In another group , wild-type mice were injected intravenously with 100 µg of apocynin 1 h after the occlusion of the middle cerebral artery . In order to restore ROS levels in Nox4−/− mice , animals received repetitive intrathecal injections of H2O2 ( 15 mg/kg ) immediately after the occlusion of the middle cerebral artery and then every hour until 6 h after stroke induction . Cortical photothrombosis was induced in 6- to 8-wk-old wild-type or Nox4−/− mice as described previously [51] , [52] . Stroke analysis was performed as described previously [53] , [54] . To determine infarct size , mice were killed 24 h after tMCAO , pMCAO , or cortical photothrombosis . Brains were cut in 2-mm-thick coronal sections using a mouse brain slice matrix ( Harvard Apparatus ) . The slices were stained with 2% TTC ( Sigma-Aldrich ) to visualize the infarcts . Planimetric measurements ( ImageJ software , United States National Institutes of Health ) , calculating lesion volumes , were corrected for brain edema as described previously [55] . Determination of brain edema using Evans blue dye was performed as described previously [19] . Magnetic resonance imaging was performed repeatedly at 24 h and 6 d after stroke on a 1 . 5-T magnetic resonance unit ( Vision Siemens ) as described previously [56] . We used a custom-made dual channel surface coil designed for examining mice ( A063HACG; Rapid Biomedical ) . The imaging protocol comprised a coronal T2-weighted sequence ( slice thickness 2 mm ) and a blood-sensitive coronal three-dimensional T2-weighted gradient echo CISS ( slice thickness 1 mm ) sequence . Magnetic resonance images were assessed with respect to infarct morphology and the occurrence of intracerebral bleeding . Vital brain slices from infarcted mouse brains ( between –2 mm and –4 mm from bregma ) were prepared as described previously [57] . After RNA isolation , we quantified NOX4 mRNA expression using real-time PCR and the TaqMan system ( TaqMan Gene Expression Arrays for murine NOX4 , assay ID Mm00479246_m1 , Applied Biosystems ) , using 18s rRNA ( TaqMan Predeveloped Assay Reagents , part number 4319413E , Applied Biosystems ) to normalize the amount of sample RNA . Histology was performed by using formalin-fixed mouse brains on day 1 after tMCAO . Samples were embedded in paraffin and cut into 4-µm-thick sections ( 0 . 5 mm anterior from bregma ) . After deparaffinization and rehydration , tissues were stained with hematoxylin and eosin or Nissl staining solution ( Sigma-Aldrich ) . Immunohistochemical detection of NOX4 was performed on formalin-fixed human brain slices or cryopreserved mouse brain slices . A NOX4-specific primary antibody [58] was applied at a dilution of 1∶200 overnight at 4°C . To identify the cellular origin we performed double staining of NOX4 with the neuronal marker NeuN ( 1∶1 , 000 ) and the endothelial marker von Willebrand Factor ( 1∶25 ) . The presence of ROS and other oxidants such as ONOO− was visualized on frozen mouse brain sections 12 h and 24 h after tMCAO or 24 h after pMCAO using dihydroethidium ( Sigma; 2 µM stock ) staining , as described previously [59] , in coronal brain sections taken from identical regions ( –0 . 5 mm from bregma ) of sham-operated controls , wild-type and Nox4−/− mice that had undergone stroke , and wild-type mice and Nox4−/− mice treated with VAS2870 or H2O2 . Immunohistochemical staining for nitrotyrosine to visualize additional reactive nitrogen species was conducted on cryopreserved brain sections taken from identical regions of the mouse brain ( –0 . 5 mm from bregma ) 12 h and 24 h after tMCAO , using a polyclonal nitrotyrosine antibody . Apoptotic neurons in the ischemic hemisphere 24 h after tMCAO were visualized by TUNEL on paraffin-wax-embedded slices , using the TUNEL in situ cell death detection kit , TMR red ( Roche ) . NeuN/TUNEL double staining was performed on cryopreserved brain slices . We quantified amounts of NOX1 , NOX2 , and NOX4 protein in the cortex and basal ganglia by Western blot analysis . Data are expressed as mean ± standard deviation and were analyzed statistically using the PrismGraph 4 . 0 software package ( GraphPad Software ) . In the case of multiple group comparisons , data were tested for Gaussian distribution with the D'Agostino and Pearson omnibus normality test and then analyzed by Bonferroni-corrected one-way ANOVA or two-way ANOVA . Otherwise , the two-tailed Student's t-test was applied . For comparison of survival curves the log-rank test was used . P-values less than 0 . 05 were considered significant . Detailed power and type-II ( beta ) error calculations on infarct volumes are provided in Tables S3–S5 . The GenBank ( http://www . ncbi . nlm . nih . gov/Genbank ) accession numbers for the genes discussed in this paper are NOX1 , NM_172203; NOX2 , NM_007807; and NOX4 , NM_015760 . | Stroke is the second leading cause of death worldwide . Today , only one approved therapy exists—a drug that breaks down blood clots—the effectiveness of which is moderate , and it can only be used in about 10% of patients because of contraindications . New therapeutic strategies that are translatable to humans and more rigid thresholds of relevance in pre-clinical stroke models are needed . One candidate mechanism is oxidative stress , which is the damage caused by reactive oxygen species ( ROS ) . Antioxidant approaches that specifically target ROS have thus far failed in clinical trials . For a more effective approach , we focus here on targeting ROS at its source by investigating an enzyme involved in generating ROS , known as NADPH oxidase type 4 , or NOX4 . We found that NOX4 causes oxidative stress and death of nerve cells after a stroke . Deletion of the NOX4-coding gene in mice , as well as inhibiting the ROS-generating activity of NOX with a pharmacological inhibitor , reduces brain damage and improves neurological function , even when given hours after a stroke . Importantly , neuroprotection was preserved in old male and female Nox4−/− mice as well as in Nox4−/− mice subjected to permanent ischemia . NOX4 thus represents a most promising new therapeutic target for reducing oxidative stress in general , and in brain injury due to stroke in particular . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"physiology/cardiovascular",
"physiology",
"and",
"circulation",
"neurological",
"disorders/cerebrovascular",
"disease",
"non-clinical",
"medicine/research",
"methods",
"pharmacology/drug",
"development",
"cardiovascular",
"disorders/cardiovascular",
"pharmacology",
"genetics",
"and",
"genomics/gene",
"function",
"neurological",
"disorders/neuropharmacology"
] | 2010 | Post-Stroke Inhibition of Induced NADPH Oxidase Type 4 Prevents Oxidative Stress and Neurodegeneration |
During the last 20 years , the epidemiology of Japanese encephalitis virus ( JEV ) has changed significantly in its endemic regions due to the gradual displacement of the previously dominant genotype III ( GIII ) with clade b of GI ( GI-b ) . Whilst there is only limited genetic difference distinguishing the two GI clades ( GI-a and GI-b ) , GI-b has shown a significantly wider and more rapid dispersal pattern in several regions in Asia than the GI-a clade , which remains restricted in its geographic distribution since its emergence . Although previously published molecular epidemiological evidence has shown distinct phylodynamic patterns , characterization of the two GI clades has only been limited to in vitro studies . In this study , Culex quinquefasciatus , a known competent JEV mosquito vector species , was orally challenged with three JEV strains each representing GI-a , GI-b , and GIII , respectively . Infection and dissemination were determined based on the detection of infectious viruses in homogenized mosquitoes . Detection of JEV RNA in mosquito saliva at 14 days post infection indicated that Cx . quinquefasciatus can be a competent vector species for both GI and GIII strains . Significantly higher infection rates in mosquitoes exposed to the GI-b and GIII strains than the GI-a strain suggest infectivity in arthropod vectors may lead to the selective advantage of previously and currently dominant genotypes . It could thus play a role in enzootic transmission cycles for the maintenance of JEV if this virus were ever to be introduced into North America .
Japanese encephalitis virus ( JEV ) is a zoonotic flavivirus endemic in the Asian Pacific region . Transmission of JEV in endemic regions predominantly alternates between amplification hosts and Culex tritaeniorhynchus mosquitoes . Infected ardeid birds and pigs often develop high-titer viremia , which sustain subsequent transmission [1] . Similar to West Nile virus , another member in the JEV serocomplex , JEV has been known to utilize multiple species of mosquitoes for transmission and maintenance in nature [2] . In addition to the endemic vector Cx . tritaeniorhynchus , other species of mosquitoes have been demonstrated to be competent for transmission of JEV including: Cx . fuscocephala , Cx . pipiens , and Cx . quinquefasciatus [3–9] . In the process of its introduction to Australia , mainly in the northern islands , Cx . annulirostris was demonstrated to be a competent vector species for the first time [10] . Additional species belonging to the genus of Aedes , Anopheles , and Armigeres are also susceptible to JEV infection [2] . Although the majority of human infections are asymptomatic , childhood encephalitic diseases caused by JEV can often be lethal with a 20–30% mortality rate . Neurological disorders are reported in 30–50% of survivors [11] . Several vaccines have been developed and used for childhood immunization programs , which have substantially reduced the disease burden in the late 20th century [12] . The evolution of JEV has led to at least five genotypes in epidemic and endemic regions in Asia . Strains in genotype III ( GIII ) have been reported to be the dominantly circulating genotype and associated with multiple outbreaks between 1935 and the 1990s in various countries; whereas , the first isolate of genotype I ( GI ) JEV strain did not occur until 1967 in Cambodia [13] . In the last two decades , the GI strains have been more frequently isolated and have resulted in the displacement of GIII strains in several countries in East Asia [14 , 15] such that GI and GIII strains have been continuously circulating in Asia since the 1990s . Based on the isolates collected between the 1960s and 1990s , GI strains can be divided into two distinct clades , GI-a and GI-b [13 , 16] . Both GI-a and GI-b originated from the endemic Southeast Asia region but showed distinct dispersal patterns and epidemiological characteristics . GI-a strains were first found in Thailand and Cambodia with its subsequent spread to Australia . In contrast to the restricted geographic distribution of GI-a strains , the GI-b strains have rapidly dispersed throughout the majority of East Asia and are ultimately responsible for the displacement of the GIII strains . Although previously published phylogenetic data established our fundamental knowledge of viral genetics and epidemiology , our understanding of the mechanisms responsible for the emergence of GI strains is very rudimentary . To date , only phenotypic studies have been undertaken using in vitro cell culture models [13] . A significant knowledge gap exists in the lack of experimental evidence derived from in vivo models to characterize the mechanism for genotype replacement . Vector competence of various Cx . quinquefasciatus populations for JEV has been previously demonstrated by oral infection of GIII strains . Recently , our group has demonstrated that North American Cx . quinquefasciatus mosquitoes are competent vectors for the transmission of the JEV GIII Taira strain [17] . Cx . quinquefasciatus mosquitoes from Asia , Australia , and Brazil have also been shown to be competent vectors of JEV [3–5] . Collectively , these studies have demonstrated that JEV can be transmitted by Cx . quinquefasciatus . Despite the rapid emergence of GI-b strains , it has not been established whether or not Cx . quinquefasciatus can transmit GI strains . To determine vector competence of Cx . quinquefasciatus for GI strains and gain further insight into the emergence of GI strains , three JEV strains , each representing GI-a , GI-b and GIII , were orally administered to Cx . quinquefasciatus . Our results demonstrate distinct phenotypic differences between GI-a and GI-b strains in mosquitoes and suggest the difference in infectivity in competent mosquito species may be a critical determinant contributing to the emergence of GI-b strains .
African green monkey kidney Vero76 cells ( source: Arthropod-Borne Animal Disease Research Unit , Agriculture Research Service , United States Department of Agriculture ) and Aedes albopictus C6/36 cells ( source: Arthropod-Borne Animal Disease Research Unit , Agriculture Research Service , United States Department of Agriculture ) were maintained in L-15 media as previously described [18] . C6/36 cells were used to propagate virus stocks subsequently used in oral infection experiments . Vero76 cells were used for the detection of infectious viruses in the homogenized mosquito samples using the tissue culture 50% infectious dose ( TCID50 ) method . Based on the published phylogenetic analysis , three strains of JEV were chosen as representatives for GI-a , GI-b and GIII [16 , 19–21] . Strain KE-93-83 ( source: existing virus culture collection in the laboratory of Dr . Alan D . T . Barrett , University of Texas Medical Branch ) was used as a representative for GI-a . Prior to the study , it was passaged twice in Vero cells and once in C6/36 cells . Strain JE-91 ( source: existing virus culture collection in the laboratory of Dr . Alan D . T . Barrett , University of Texas Medical Branch ) , originally isolated from mosquitoes collected in Korea in 1991 followed by one passage in Vero cells and one passage in C6/36 cells , was chosen as a representative for GI-b . The Taira strain ( source: existing virus culture collection in the laboratory of Dr . Alan D . T . Barrett , University of Texas Medical Branch ) originally derived from an infected human in an epidemic in Japan in 1948 represents GIII and was passaged twice in Vero cells . All three strains used in viremic blood meals were generated by a single passage in C6/36 cells at 28°C and harvested when greater than 80% of cytopathic effect was present . Culex quinquefasciatus mosquitoes collected from Valdosta , GA ( source: the laboratory of Dr . Mark Blackmore , Valdosta State University ) were used in this study as previously described [17] . Mosquitoes were maintained in 12" cages with 10% sucrose ad libitum at 28°C . A 16hr:8hr light:dark photoregime was used for all experiments . For per os infection , eight-to-10-day-old female mosquitoes of generations below F6 were used . Prior to the infection , mosquitoes were deprived of sugar and water for 48 and 24 hours , respectively . Viremic blood meals were prepared by mixing virus stocks with defibrinated sheep blood ( Colorado Serum , CO ) and delivered through a Hemotek membrane feeding apparatus ( Discovery Workshop ) and cotton pledget for one hour . Mosquitoes were cold anesthetized on ice prior to sorting engorged mosquitoes . Three-to-5 engorged mosquitoes were immediately collected to assess the quantities of viruses ingested through the blood meals . Titers of viremic blood meals and three engorged mosquitoes are summarized in Table 1 . Mosquitoes were collected at 7 and 14 days post-infection ( d . p . i . ) by mechanical aspiration . Dissections were performed to separate the body ( abdomen ) section and secondary tissues ( head , wings , and legs ) of individual mosquitoes . Whole mosquitoes were also collected to assess the growth kinetics . All tissue samples were collected in 1ml L-15 medium supplemented with amphotericin B and sodium deoxycholate at 1 and 0 . 8 μg/ml , respectively . Homogenization and titration of samples were performed using previously published methods [18] . At 14 d . p . i . , saliva of each mosquito was collected prior to dissection [18] . Viral RNA was extracted with the QIAamp viral RNA extraction kit ( Qiagen ) and reverse transcribed by Superscript III reverse transcriptase ( Life Technologies ) with the reverse primer ( Integrated DNA Technologies ) prMR3 ( 5'-CATGAGGTATCGCGTGGC-3' ) . cDNA was amplified by Taq DNA polymerase ( New England BioLabs ) using the semi-nested PCR cycles described by Johansen et al . [22] . Primer sets target the conserved region between the capsid and pre-membrane genes . The outer set of primers were forward primer FV128 ( 5'-CCGGGCTGTCAATATGCT-3' ) and reverse primer prMR3 . The inner set of primers were forward primer FV128 and reverse primer JE659 ( 5'-CACCAGCAATCCACGTCCTC-3' ) . Infection status of orally challenged mosquitoes was determined by the detection of infectious viruses from homogenized samples . Infection rate was calculated by dividing the number of infected mosquitoes over the number of mosquitoes tested . Disseminated infection was defined by the detection of infectious viruses from homogenized secondary tissues among dissected mosquitoes . Dissemination rates were determined by the numbers of dissected mosquitoes containing positive secondary tissues divided by the number of infected mosquitoes dissected . Viral transmission was demonstrated by the presence of viral RNA in the saliva detected by semi-nested RT-PCR . Transmission rate was calculated by dividing the numbers of mosquitoes containing positive saliva by the numbers of infected mosquitoes . Infection , dissemination and transmission rates were compared among three JEV strains tested by Chi-square test or Fisher’s exact test coupled with a Tukey-type multiple comparison test . Analysis of growth kinetics was performed by comparing the titers of whole mosquitoes collected at 7 and 14 d . p . i . with Friedman’s two-way nonparametric analysis of variance . All statistical analyses were conducted using the SAS ( Statistical Analysis System ) software ( version 9 . 4 , Cary , NC ) .
Infection data following oral challenge with different JEV strains are summarized in Table 1 . At 7 d . p . i . , there was a significant difference in the infection rates among the three strains tested ( Chi-square = 25 . 49 , df = 2 , p<0 . 001 ) with the highest infection rate observed in mosquitoes exposed to the GIII Taira strain ( 95 . 1% , 39/41 ) and the lowest infection rate observed among mosquitoes exposed to the GI-a KE-93-83 strain ( 43 . 9% , 18/41 ) . The infection rate of the GIII Taira strain group was also significantly higher than the infection rate of the GI-b JE-91 strain ( 57 . 6% , 19/33 ) . No statistical difference was found in the infection rates between the KE-93-83 strain and the JE-91 strain . Infection rates for the three JEV strains were significantly different at 14 d . p . i . ( Chi-square = 11 . 95 , df = 2 , p = 0 . 003 ) . The two representatives of the endemic genotypes GI-b and GIII showed no statistical difference; the JE-91 strain ( 55 . 6% , 25/45 ) and the Taira strain ( 66 . 7% , 44/66 ) . Interestingly , the representative of the non-endemic GI-a , the KE-93-83 strain , maintained the lowest infection rate among the three strains . Tukey type multiple comparison showed a significantly lower infection rate ( 35 . 2% , 19/54 ) than the Taira strain ( p<0 . 05 ) . Overall , the KE-93-83 strain , the representative strain of non-endemic GI-a , showed the lower infectivity than the other two strains , which are the strains representing the endemic GI-b and GIII . Based on the observation that JEV strains can successfully establish infections , we investigated whether or not the infection will ultimately lead to viral dissemination . Titration of secondary tissues demonstrated that infection of all three strains led to dissemination ( Table 1 ) . There were no statistical differences in the dissemination rates among the three strains tested at 7 d . p . i . ( Fisher’s Exact test , p = 0 . 231 ) or 14 d . p . i ( Fisher’s Exact test , p = 0 . 664 ) . At 7 d . p . i . , the dissemination rate of the KE-93-83 strain was 23 . 1% ( 3/13 ) . Comparable dissemination rates were observed among the mosquitoes infected by the JE-91 strain ( 30 . 0% , 3/10 ) and the Taira strain ( 8 . 3% , 2/24 ) . Dissemination rates were also similar at 14 d . p . i . among the KE-93-83 strain ( 16 . 7% , 2/12 ) , the JE-91 strain ( 28 . 6% , 4/14 ) , and the Taira strain ( 32 . 1% , 9/28 ) . The results suggest that all three JEV strains in this study were able to disseminate into secondary tissues after the establishment of infection . Detection of JEV viral RNA in the saliva of virus-infected mosquitoes was achieved by semi-nested RT-PCR indicating the capacity of viral transmission among Cx . quinquefasciatus tested in this study as shown in Table 1 . Among the mosquitoes infected by the KE-93-83 strain , 5 . 3% ( 1/19 ) of saliva samples tested were positive for the presence of JEV viral RNA . Similar results were also observed in mosquitoes infected by the JE-91 strain ( 8 . 0% , 2/25 ) and the Taira strain ( 6 . 8% , 3/44 ) . There was no demonstrable difference in the transmission rates among the three strains tested in this study ( Fisher’s Exact test , p = 0 . 999 ) indicating North American Cx . quinquefasciatus can serve as competent vectors for GI-a , GI-b and GIII JEV strains . As dissemination and transmission require viral replication in various tissues in infected mosquitoes , the replication kinetics of three JEV strains was analyzed by the titration of infected whole mosquitoes . The results are summarized in Fig 1 . At 7 d . p . i . , the average virus infectivity titer in mosquitoes infected by the KE-93-83 strain was 3 . 5 logTCID50/ml ( n = 5 ) whereas for the JE-91 and Taira strains it was 2 . 8 ( n = 9 ) and 3 . 5 ( n = 15 ) logTCID50/ml , respectively . The average infectivity titer in mosquitoes infected by the KE-93-83 strain at 14 d . p . i . was 3 . 1 logTCID50/ml ( n = 7 ) while the JE-91 and Taira strains maintained average titers of 2 . 8 ( n = 9 ) and 3 . 5 ( n = 37 ) logTCID50/ml , respectively . Friedman’s two-way nonparametric analysis of variance showed there was no significant difference in the average titers of infected mosquitoes among the three different strains ( p = 0 . 174 ) at 7 or 14 d . p . i .
Our observations improve our understanding of the potential for the establishment of enzootic transmission cycles by the GI-b strains of JEV in North America . Although the GI-b strains have gradually become the dominant genotype in Asia , since 1990s , there is no information available to assess the risk of it establishing enzootic transmission cycles in North America or other regions , where JEV has not been reported . Our results support the potential role of North American Cx . quinquefasciatus as a competent vector for the newly emerging GI-b strains . With the additional evidence showing that North American avian species can develop viremia after being challenged with a GI strain [23] , it is reasonable to conclude that JEV remains as an important human and veterinary public health threat after the shift of the dominant genotype from GIII to GI-b . To the best of our knowledge , this is also the first report identifying phenotypic difference among different JEV genotypes in vivo . Significantly higher infection rates were observed at 7 and 14 d . p . i . in Cx . quinquefasciatus challenged by endemic GI-b and GIII strains of JEV; whereas , the non-endemic GI-a strain consistently showed significantly lower infection rates . Because Cx . quinquefasciatus does not belong to the Cx . vishnui complex , which is more directly related to the endemic transmission of JEV in nature , the results should be interpreted with caution . However , the importance in characterizing JEV in Cx . quinquefasciatus should not be overlooked because of its role as a source of JEV isolates in nature and the published evidence demonstrating its vector competence for JEV in multiple laboratory studies [3 , 8 , 17 , 24–26] . Given the anticipated high mortality from transporting eggs of Cx . tritaeniorhynchus from Asia [13] , using Cx . quinquefasciatus is an acceptable substitute for characterization of JEV in vivo . The results suggest that endemic GI-b and GIII strains of JEV are more infectious to Cx . quinquefasciatus through oral infection than GI-a strains , which have not become endemic since its emergence . This finding agrees with the epidemiological observations and directly contributes to much needed knowledge as to why GI-b strains of JEV can emerge after continuous evolution . Previously , superior multiplication kinetics were reported for the GI-b JE-91 strain in C6/36 Ae . albopictus cells compared to GIII Tiara with GI-a KE-93-83 having the lowest multiplication kinetics of the three virus strains . It was suggested this might result in a selective advantage explaining the emergence of GI-b strains and the subsequent displacement of GIII strains [13] . An alternative hypothesis that could explain the emergence of the GI-b strains is the subtle difference in the amino acid sequences in the NS5 RNA-dependent RNA polymerase between the GI and GIII strains may result in an increase in replication efficiency [27] . These two hypotheses are not mutually exclusive , but previous studies showed the three virus strains had indistinguishable multiplication kinetics in Duck embryo fibroblast cells [13] . A comparative analysis of the multiplication kinetics of the GI-b and GIII strains was previously performed in North American avian species and demonstrated that multiplication kinetics of the GI-b strain was at least as high as the GIII strain and in most cases multiplied to higher titer [23] . Detection of viremia in North American avian species suggests the possibility of establishment of enzootic transmission by providing sources of infectious blood meals for competent mosquitoes . Higher viremic titer in infected avian species caused by the GI-b strain further suggests the newly emerging GI-b strains can have higher epidemic potential than the previously dominant GIII strains . Our 14 d . p . i oral infection data for Cx . quinquefasciatus in this study showed the GI-b strain was as infectious and disseminated similarly to the GIII strain whereas the GI-a strain did not . Multiplication kinetics between the GI-b strain and GIII strain also did not differ significantly and subsequently resulted in different titers of infected mosquitoes . The lack of a distinguishable difference between the GI-b and GIII strains in our study may be due to the difference between the in vitro and in vivo experimental conditions utilizing C6/36 cells and orally infected Cx . quinquefasciatus , respectively . Therefore , with respect to mosquito data , we did not observe a selective advantage of the GI-b strains over the GIII strains and so cannot conclude the displacement of the GIII strains by the GI-b strains in various endemic countries is due to differences in infectivity for vectors , rather enhanced infection of mosquitoes by GI-b compared to GI-a viruses together with higher viremias than GIII viruses in birds may have led to the selective advantage of the GI-b viruses . Through per os infection , our and others’ studies demonstrated Cx . quinquefasciatus , which is commonly present in the sub-tropical region of the American continent , can be competent enzootic vectors for members of the JEV serocomplex such as Saint Louis encephalitis virus ( SLEV ) and West Nile virus ( WNV ) [28 , 29] . In contrast to our observation on the infection of endemic JEV strains in Cx . quinquefasciatus , oral ingestion of WNV resulted in comparable or significantly higher infection and dissemination rates regardless of genotypes used in the studies [28 , 30–35] . Similar to our observations on endemic strains of JEV in this study , previously published studies also demonstrated strains of SLEV with higher epidemic potential are more infectious than non-epidemic strains when orally delivered to Cx . quinquefasciatus collected from Argentina [29] . A more recently published study also demonstrated the difference in infectivity between genotype III and V of SLEV in Cx . quinquefasciatus [36] . Based on the infection study performed with Cx . quinquefasciatus collected from Gainesville , FL , the higher infection rate of SLEV ( 93–100% ) was achieved with relatively low viremic titers compared to JEV challenged at higher viral titers in this study . However , there was no distinguishable difference between SLEV and endemic strains of JEV in the capacity for dissemination into secondary tissues from the infected midguts [28] . In addition to the evidence supporting the potential role of Cx . quinquefasciatus as a competent vector , it is important to keep in mind that there are multiple species of mosquitoes found in North America , which have been previously demonstrated to be competent for the transmission of JEV and other related flaviviruses [37] . As observed with the process of WNV becoming endemic in North American since 1999 , it is certainly favorable for arboviruses to utilize multiple species of mosquitoes as enzootic vectors in order to establish its transmission cycles and achieve viral maintenance in adverse climatic conditions , especially winter [38 , 39] . Further evaluation on other medically important Culex species mosquitoes will be critical for understanding the relative risk of the introduction of JEV and the establishment of its enzootic transmission cycles . | Japanese encephalitis virus ( JEV ) is a zoonotic flavivirus , which is primarily transmitted by Culex species mosquitoes and a leading cause of pediatric encephalitis in Asia . JEV is also an important public health threat to countries outside the endemic region because collections of Cx . quinquefasciatus from around the world have demonstrated competence for the transmission of JEV and are capable of establishing enzootic transmission cycles between viremic avian and swine species . In the last two decades , the dominantly circulating genotype of JEV in endemic regions has experienced a significant shift ( genotype III to Genotype I ) . It is unclear if the newly dominant circulating G1-b genotype can still be vectored by Cx . quinquefasciatus . In this study , Cx . quinquefasciatus collected from North America was demonstrated to be competent for the transmission of the newly dominant genotype . Different infectivities observed between the endemic strains and non-endemic strain provides the mechanistic knowledge of the selection and emergence of endemic genotypes after continuous viral evolution . | [
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] | [
"japanese",
"encephalitis",
"virus",
"invertebrates",
"medicine",
"and",
"health",
"sciences",
"body",
"fluids",
"pathology",
"and",
"laboratory",
"medicine",
"pathogens",
"geographical",
"locations",
"microbiology",
"vector-borne",
"diseases",
"vertebrates",
"saliva",
"animals",
"viruses",
"rna",
"viruses",
"insect",
"vectors",
"evolutionary",
"emergence",
"infectious",
"diseases",
"birds",
"medical",
"microbiology",
"epidemiology",
"microbial",
"pathogens",
"disease",
"vectors",
"insects",
"arthropoda",
"people",
"and",
"places",
"mosquitoes",
"west",
"nile",
"virus",
"asia",
"flaviviruses",
"anatomy",
"viral",
"pathogens",
"physiology",
"biology",
"and",
"life",
"sciences",
"evolutionary",
"biology",
"amniotes",
"evolutionary",
"processes",
"organisms"
] | 2016 | Differential Infectivities among Different Japanese Encephalitis Virus Genotypes in Culex quinquefasciatus Mosquitoes |
Zambia is endemic for Taenia solium taeniosis and cysticercosis . In this single-centered , cross-sectional , community-based study , the role of neurocysticercosis ( NCC ) as a cause of epilepsy was examined . People with epilepsy ( PWE , n = 56 ) were identified in an endemic area using a screening questionnaire followed by in-depth interviews and neurological examination . Computed tomography ( CT ) was performed on 49 people with active epilepsy ( PWAE ) and their sera ( specific antibody and antigen detection , n = 56 ) and stools ( copro-antigen detection , n = 54 ) were analyzed . The CT scan findings were compared to a group of 40 CT scan controls . Of the PWE , 39 . 3% and 23 . 2% were positive for cysticercal antibodies and antigens , respectively , and 14 . 8% for coproantigens ( taeniosis ) . Lesions highly suggestive of NCC were detected in 24 . 5% and definite NCC lesions in 4 . 1% of CT scans of PWAE . This compares to 2 . 5% and 0% , respectively , in the control CT scans . Using the Del Brutto diagnostic criteria , 51 . 8% of the PWAE were diagnosed with probable or definitive NCC and this rose to 57 . 1% when the adapted criteria , as proposed by Gabriël et al . ( adding the sero-antigen ELISA test as a major criterion ) , were used . There was no statistically significant relationship between NCC , current age , age at first seizure and gender . This study suggests that NCC is the single most important cause of epilepsy in the study area . Additional large-scale studies , combining a community based prevalence study for epilepsy with neuroimaging and serological analysis in different areas are needed to estimate the true impact of neurocysticercosis in endemic regions and efforts should be instituted to the control of T . solium .
Neurocysticercosis ( NCC ) is a parasitic infection of the brain caused by the metacestode larval stage of the pork tapeworm Taenia solium and occurs primarily in low-income countries . Pigs are the natural intermediate host for the tapeworm , while humans are the only definitive host but may accidentally also serve as the intermediate host [1] . Humans become tapeworm carriers ( taeniosis ) after ingestion of undercooked pork that is infected with cysticerci ( cysticercosis ) . Humans and pigs acquire cysticercosis after ingestion of infective eggs that are passed in stool by a tapeworm carrier . The cysts have the propensity to localize in the brain leading to NCC [1] . NCC can cause a wide range of neurological disorders including epileptic seizures , headaches and focal neurological symptoms/signs [2–4] . It is reported as one of the major causes of acquired epilepsy in endemic areas [5 , 6] , which has profound social , physical and psychological consequences . Population based studies have reported that approximately 30% of cases of epileptic seizures are attributable to NCC [7 , 8] . The prevalence of epilepsy as determined in community-based studies throughout Africa shows varying results depending on the study population and the methodologies used , ranging from 5 . 2–74 . 4/1000 with a median of 15/1000 inhabitants [9 , 10] . This is compared to high-income countries where the prevalence of active epilepsy is estimated to be about 4–8/1000 inhabitants [11] . Epilepsy represents the most common chronic neurological disorder in sub-Saharan Africa , and involves unprovoked recurrent ( two or more ) afebrile seizures that may result in injury , disability and social stigmatization [12 , 13] . For individuals with epilepsy in many low-income countries , including Zambia , the etiology of the disease often remains unclear . The endemicity of T . solium infections in Zambia has been mentioned in many reports . Porcine cysticercosis prevalence estimates have been reported to range from 8 . 2–64 . 2% [14] while human cysticercosis is reported to range from about 6–13% ( based on circulating antigen detection ) in studies carried out in the Eastern province [15 , 16] . The Eastern province harbors almost 50% of the total number of pigs in the country , which are mostly reared under the small-scale management system [17] . Due to the lack of sanitary facilities such as latrines , free-ranging pigs have access to human feces . Such conditions increase the risk of both porcine cysticercosis and human taeniosis/cysticercosis infection . The high percentage of active cysticercosis infections ( 12 . 5% ) , determined in a study in Katete district ( Eastern province ) [15] , indicates the urgent need to identify the importance of NCC as a cause of epilepsy in this area . Therefore , the main objective of this cross-sectional study was to analyze the role of NCC as a cause of epilepsy in a study area of Katete district of the Eastern province of Zambia .
The University of Zambia Biomedical Research Ethics Committee ( IRB0001131 ) and the Ethical Committee of the University of Antwerp , Belgium , both granted ethical clearance for the study ( ITG:12084813 ) . Further approval was sought from the Ministry of Health of Zambia , from the local district health authorities and the area chief . Meetings were held with the people in the villages through their leaders ( headmen ) to explain the purpose of the study , request their permission to conduct the study and also to invite them to participate . Further permission was sought from the individual subjects to enter the study after oral consent for the screening questionnaires and written informed consent for further examinations . The use of oral consent in the screening phase was explicitly approved by both ethics committees and no special documentation of the oral consent was required . For individuals below the consenting age , their parents were asked for informed consent on their behalf . All collected samples were assigned an identification number that was linked to the questionnaire , neurological examination , laboratory test results and neuroimaging results . For neuroimaging , participants that were diagnosed with active epilepsy were transported to Lusaka accompanied by two medical officials from the local rural health center . For children below the age of 16 , a parent or guardian accompanied them to Lusaka . In collaboration with the regional health center Mtandaza all patients received free antiepileptic drugs according to national guidelines and niclosamide in case they were diagnosed with taeniosis . Patients with active NCC were offered a combination of dexamethasone and anthelmintic treatment in an inpatient setting and , according to the resolution of their symptoms a re-scan free of charge . The study was carried out in Mtandaza community in Katete district of the Eastern province of Zambia ( Fig 1 ) . The Mtandaza Rural Health Center ( RHC ) provides health care in this community with a catchment population of approximately 20000 . The people in this area practice subsistence agriculture raising animals like cattle , goats , pigs and chickens and growing crops like maize , groundnuts , bananas and cotton . Their homes are of adobe , have few sanitary facilities with hand pumps being the source of water for most villages . Pigs are commonly bought and sold among villagers and also to intermediary traders from nearby towns . Since sanitation is poor , pigs have access to the feces in nearby bushes that are used as latrines by the villagers . The district of Katete was chosen because endemicity of human and porcine T . solium cysticercosis has been demonstrated during previous studies [16 , 18] . The rural community was selected because of free-range pig keeping , absence of active ongoing sanitation programs , reports of people with epilepsy ( PWE ) , and the common observation of cysticerci in slaughtered pigs . Furthermore , the RHC had shown willingness to collaborate with this study . This was a single-center , cross-sectional and community-based study . In August 2012 , the study began with the screening of all inhabitants of villages within a 7 km radius from the Mtandaza RHC . Local health workers , after undergoing basic questionnaire administration training , used a standardized screening questionnaire modified and adapted by Birbeck and Kalichi [19] to identify people with possible epilepsy ( see S1 File ) . The screening questionnaire was administered door-to-door to any individual that lived in the study area and was willing to participate . Altogether , 4443 individuals were screened ( Fig 2 ) . Due to financial restriction the number of computed tomography ( CT ) scans had to be limited to 50 . Therefore , the recruitment started with those people with the highest number of positive answers to the screening questions 1–8 and a negative answer in question nine . Finally , all people with a minimum of four positive answers and one with three positive answers , giving a total of 101 individuals , were invited for further examinations at the RHC . Twenty-five individuals were either not available for examination or refused to participate in the study . After written informed consent , we examined 76 individuals with suspected epilepsy . Twenty individuals dropped out because they had no epilepsy or they withdrew their consent ( see also Table 1 ) . The 56 identified PWE provided a detailed medical history , underwent neurological examination and provided blood and stool samples for the immunological diagnosis of cysticercosis and taeniosis , respectively ( Fig 2 ) . Epilepsy was diagnosed according to a classification for resource poor settings , which is in agreement with the definitions of epilepsy by the International League against Epilepsy , but was adapted to local circumstances . The four major diagnostic groups were “Generalized seizures within specific age range” , “Generalized seizures outside specific age range” , “Generalized seizures with diffuse brain damage” and “Generalized seizures with focal signs” . The first two groups were distinguished by the age at onset of epilepsy being either within or outside the age span of six to 25 years [20] . Fifty-one patients were diagnosed with active epilepsy ( PWAE ) , which was defined that the patient had had seizures within the last two years or had been on antiepileptic medication , while five had inactive epilepsy ( PWIE ) and did not fulfill the criteria mentioned before . The PWAE were transported 480 km to Lusaka and examined for lesions compatible with NCC by CT scanning of the brain at the public Levy Mwanawasa General Hospital . About 5 ml of blood were collected into sterile plain blood collecting tubes and allowed to clot . Serum was extracted from all the blood samples by first allowing them to stand at 4°C overnight and then centrifuging at 3000g for 15 minutes . The supernatant ( serum ) was then aliquoted into 1 . 8 ml vials and stored at -20°C until use . Submitted stool samples were placed in 10% formalin and stored until use . All the samples were analyzed in the Cysticercosis Working Group for Eastern and Southern Africa Reference Centre for Cysticercosis at the School of Veterinary Medicine at the University of Zambia in Lusaka . The serum samples were tested for circulating cysticercal antigens using the monoclonal antibody based B158/B60 antigen enzyme linked immunosorbent assay ( sero-Ag-ELISA ) as described by Dorny et al . [21] . To determine the cutoff , the optical density ( OD ) of each serum sample was compared with a series of eight reference negative human serum samples at a probability level of p < 0 . 001 [21] . The serum samples were also tested for specific antibodies against cysticercosis using a commercial kit , Immunetics ( Immunetics Inc . , Boston , Massachusetts , USA ) . The assay was performed according to the manufacturer’s instructions . In brief , the assay is an enzyme linked immunoelectrotransfer blot ( EITB ) , which uses seven purified T . solium antigens ( diagnostic bands being Gp50 , Gp42-39 , Gp24 , Gp21 , Gp18 , Gp14 and Gp13 , whereby Gp stands for glycoprotein and the number is the molecular weight of each antigen expressed in kilodaltons ) . Reactions to any one or more of the bands are considered positive . The stool samples were examined for copro-antigens using the polyclonal antibody based ELISA ( copro-Ag-ELISA ) as described by Allan et al . [22] with slight modifications [15] . To determine the test result , the OD of each stool sample was compared with the mean of a series of eight reference Taenia negative stool samples plus three standard deviations ( cut-off ) . Of the 51 PWAE , 49 ( two refused ) underwent CT examination at the Levy Mwanawasa General Hospital in Lusaka . The CT machine used was a Neusoft helical multi-slice scanner ( Neusoft Medical Systems , Shenyang , China ) . The thickness of slices for the whole scan was 5 mm . Intravenous contrast medium was not applied . A set of 101 consecutive cerebral CT scans were also obtained from the hospital data base . These were from patients presenting for CT scanning for various indications; however no further clinical information was available . Therefore , we selected those scans ( n = 40 ) that had a clear pathology assuming that this was the reason for the performance of the CT scan and searched for NCC related pathologies on the CT scan images . The scans were checked by two people independently , a neuroradiologist ( KS ) and a neurologist ( JB ) . In case of disagreement , the scans were discussed until a consensus was achieved . CT based diagnosis of NCC was divided into three groups: definite NCC lesions , lesions highly suggestive of , and those compatible with NCC . Definite NCC lesions were cystic lesions showing the scolex while those without a visible scolex , multiple and parenchymal brain calcifications were categorized as lesions highly suggestive of NCC [23] . Compatible cases were those with any other pathology that could be caused by NCC , such as hydrocephalus [24 , 25] , and those with single calcifications in brain parenchyma [26] . NCC was diagnosed as being either active for any cystic lesions or inactive for only parenchymal calcifications [27 , 28] . Diagnosis of NCC was based on the Del Brutto diagnostic criteria [23] . Different diagnostic findings are stratified into absolute , major and minor criteria . The combination of these criteria results in a diagnosis of “Definitive NCC” , “Probable NCC” and “No NCC” [23] . All our participants had an epidemiological criterion , because they lived in an area endemic for T . solium taeniosis/cysticercosis and all PWE had a minor criterion , because they had a clinical manifestation suggestive of NCC . Additional to the Del Brutto criteria , we also used sero-Ag-ELISA as a major criterion as proposed by Gabriël et al . [29] . Both diagnostic classifications are presented in the result section . Data from the screening questionnaire and all further examinations was entered into a Microsoft Excel spreadsheet ( double entry and review , according to Good Data Practice ( GDP ) and analyzed with R ( 3 . 1 . 0 ) ( http://www . r-project . org/ ) . The statistical test used is mentioned in the text . The significance level was set to 0 . 05 . In the analysis of age differences , cases with unknown age are excluded from the analysis . For the diagnosis of NCC missing values for minor , major and absolute criteria are set to negative .
Of the 51 PWAE , 24 ( 47 . 1% ) were females . For the five PWIE , one ( 20 . 0% ) was female . In the group of 40 CT control scans , 20 ( 50 . 0% ) were women . There was no significant difference in gender distribution among the three groups ( Fisher’s exact test , p = 0 . 552 ) . The mean age with standard deviation of the 56 people with active epilepsy and inactive epilepsy was 32 . 3 ± 15 . 3 and 28 . 8 ± 20 . 4 years , respectively . The mean age of the patients from the CT control scans was 41 . 4 ± 20 . 9 years . There was a significant difference in age among the three groups ( ANOVA , p = 0 . 048 ) . A Scheffe-Post-Hoc-Test revealed a borderline difference between CT controls and PWAE ( p = 0 . 070 ) . After history taking and neurological examination of the 56 patients , epilepsy was diagnosed and classified as shown in Table 2 . The majority of PWE had clinically generalized seizures with about one third ( 17/56 ) having “Generalized seizures within specific age range” and a quarter ( 14/56 ) with “Generalized seizures outside specific age range” . Of the 56 PWE , 39 . 3% ( 22/56 ) had T . solium cysticercosis antibodies while 23 . 2% ( 13/56 ) had circulating cysticercus antigens . Fifty-four PWE provided a stool sample and 14 . 8% ( 8/54 ) were positive for taeniosis on copro-Ag ELISA ( Table 3 ) . The diagnostic tests results according to activity of epilepsy and the CT scan result are presented in Table 3 . A list with age , gender , the clinical diagnosis of epilepsy , the descriptive diagnosis of the CT scan , number of active and inactive NCC lesions , serological results and the diagnosis of NCC for every participant is shown in the supplement material S1 Table . There were no significant differences comparing results of the laboratory tests with the result of the classification of NCC lesions in CT scanning ( Fisher test; EITB: p = 0 . 452 , Ag-ELISA: p = 0 . 100 , copro-Ag ELISA: p = 0 . 542 ) . CT scanning was performed on 49 ( 96 . 1% ) of the 51 PWAE . Lesions compatible with NCC were found in 12 . 2% ( 6/49 ) , lesions highly suggestive of NCC in 24 . 5% ( 12/49 ) and definite NCC lesions in 4 . 1% ( 2/49 ) of CT scans of PWAE . This compares to 15 . 0% ( 6/40 ) , 2 . 5% ( 1/40 ) and 0 . 0% ( 0/40 ) , respectively , in the control CT scans . The difference between the two groups was statistically significant ( Fisher’s exact test , p < 0 . 001 ) . A detailed description of all CT scans results can be found in supplement material ( S1 Table ) . Two people with active epilepsy had a calcification in the masseter muscle visible on CT scan , which could be interpreted as cysticercosis outside the central nervous system ( CNS ) . However , Del Brutto defines “cysticercosis outside CNS” , which is a minor criterion , as multiple calcification in thigh or calf muscles [23] . Therefore , we did not consider these calcifications as “cysticercosis outside CNS” . Combining CT scan and serology EITB results , 15 . 7% ( 8/51 ) of PWAE were diagnosed with definitive and additionally 35 . 3% ( 18/51 ) with probable NCC as described by Del Brutto [23] ( Table 4 ) . Of the five PWIE , three ( 60% ) were diagnosed with probable NCC . Combining definitive and probable NCC , this resulted in an overall NCC prevalence of 51 . 8% ( 29/56 ) among PWE . 31 . 0% ( 9/29 ) of people with probable or definitive NCC had no NCC lesion on CT scan . Adding the sero-Ag-ELISA results as a major diagnostic criteria , as proposed by Gabriël et al . [29] , 25 . 5% ( 13/51 ) and 31 . 4% ( 16/51 ) of PWAE were diagnosed with definitive and probable NCC , respectively ( Table 4 ) . Of the five PWIE , two ( 40 . 0% ) were diagnosed with definitive NCC and another one ( 20 . 0% ) with probable NCC . The overall NCC prevalence in PWE was therefore 57 . 1% ( 32/56 ) . 34 . 4% ( 11/32 ) of people with probable or definitive NCC had no NCC lesions on CT scan . Combining probable and definitive NCC into a dichotomous variable , there were no significant differences in terms of gender and age at examination between people with and those without NCC ( gender: Fisher test , Gabriël et al . : p = 0 . 280 , Del Brutto: p = 0 . 788 , age at examination: T-test , Gabriël et al . : p = 0 . 817 , Del Brutto: p = 0 . 666 ) . Also , the age at first seizure was not significantly different between the two groups ( T-test , Gabriël et al . : p = 0 . 894 , Del Brutto: p = 0 . 705 ) . There was no significant association between the clinical diagnosis of epilepsy according to Winkler et al . [20] and NCC diagnosis ( Fisher test , Del Brutto: p = 0 . 265 , Gabriël et al . : p = 0 . 298 ) . The age , gender and clinical diagnosis of epilepsy are shown in Table 5 .
The results of this study , indicating NCC as the single most important cause of epilepsy in the study area , add to the body of knowledge on the endemicity of T . solium in Zambia , besides the reports on the parasite in pigs [17 , 18 , 30] and more recently in humans [15 , 16] . This will therefore contribute to the burden assessment of this important , yet neglected parasite . The prevalence of NCC among PWE reported in this study is one of the highest reported so far in the sub-Saharan African region . At a prevalence of 57 . 1% , based on the inclusion of the sero-Ag-ELISA test result , as proposed by Gabriël et al . [29] , it ranks higher than what has been reported in the pooled world estimate of 29% [7] . It also ranks higher than those reported elsewhere in the world . In Latin America , a recent meta-analysis of epilepsy and NCC revealed an NCC proportion of 32 . 3% among PWE [31] with the highest being 47% reported in a CT based study in Guatemala [32] . The estimates in Africa range from 11% in Burkina Faso [7] to 37% in the Eastern Cape Province of South Africa [33] with Tanzania reporting a prevalence of 16 . 5% [26] . The mean age of PWAE with NCC in this study ( 32 . 9 years ) is comparable to that reported elsewhere . Blocher et al . [26] reported a mean age of 32 . 5 years in a study in Tanzania . Though we had a limited sample size , the lack of association between NCC and age may point towards NCC being a cause of epilepsy in both the elderly and the young . Similarly , the age at first seizure was not significantly different between PWE with and without NCC , which is in line with results from Tanzania [26] . There was also no significant difference in gender distribution in our study cohort . This was similar to findings by Blocher et al . [26] in Tanzania who also did not find a significant difference regarding gender . A prevalence of copro-Ag-ELISA positives of 14 . 8% among the 54 PWE may indicate a possible autoinfection with T . solium eggs that could lead to NCC . Adult tapeworm carriers are not only a risk for cysticercosis to themselves , in light of poor personal hygiene , but also to other members of their households and the community . In Mexico , Garcia-Garcia et al . demonstrated that the presence of tapeworm carriers in households is the main risk factor attributed to human cysticercosis [34] . These carriers intermittently shed proglottids and/or substantial numbers of infective eggs in their feces , thereby exposing the people in their environment to cysticercosis for example due to the unhygienic preparation and serving of food [35] . With the prevailing low standards of sanitation and personal hygiene , infection and re-infection is therefore a high possibility , hence the increased levels of cysticercosis in people living in the area . At almost 60% , we have highlighted the importance of NCC as the most important single cause of epilepsy in our study area . This may also apply to other areas of Eastern and Southern provinces of Zambia , where factors for the maintenance of the parasite have been reported and high levels of environmental contamination were highlighted by the elevated porcine cysticercosis [18 , 30] and human taeniosis/cysticercosis prevalence estimates [15 , 16] . A study by Mwape et al . [16] reported a high taeniosis prevalence of 11 . 9% in an area in Katete district , indicating a substantial number of adult tapeworm carriers with the potential to result in marked environmental contamination with the taeniid eggs . The study also reported an average cysticercosis sero-prevalence of 13 . 1% ( sero-Ag ) and 35 . 4% ( sero-antibody ( Ab ) ) during the study period . Incidence rates of 6300 ( sero-Ag , per 100000 persons-year ) and 23600 ( sero-Ab , per 100000 persons-year ) were also determined . The sero-reversion rates recorded were 44% and 38 . 7% for sero-Ag and sero-Ab , respectively , over the whole period [16] . The study by Mwape et al . showed the dynamic nature of T . solium infections and that many of the people at risk become ( re ) infected due to the substantial environmental contamination , with a high number turning sero-negative within a year after infection . With increased levels of ( re ) infection rates , the possibility of NCC becomes very likely . These results , though from a different part of the district , may explain the high prevalence of NCC in PWAE in addition to the prevailing risk factors for T . solium transmission in the study area of the current study . The sero-prevalence of cysticercosis antibodies of 39 . 3% in our population of PWE is similar compared to the 35 . 4% ( mean ) reported in a population not suffering from epilepsy in another part of the district [15] . The prevalence of anticysticercal antibodies in PWE in our study was similar to the one reported in a study in South Africa ( 41 . 7% ) [33] , but is higher than that reported from Tanzania ( 15 . 9% ) [4] . Concerning cysticercal antigens , 23 . 2% in our population with epilepsy compared to 13 . 1% in a population without epilepsy [16] . Currently , the only published criteria for the diagnosis of neurocysticercosis , which are used in epidemiological studies , are those of Del Brutto [23] . These criteria basically combine results of neuroimaging ( CT and/or Magnetic Resonance Imaging ) and serological diagnostic tools ( circulating cysticercal antigen in cerebrospinal fluid and specific antibody detection in serum/cerebrospinal fluid ) . Gabriël et al . suggested an adaptation of these criteria by using sero-Ag-ELISA as a major criterion [29] . Diagnostic criteria are important in order to compare prevalence data from different countries; however they have limitations for the diagnosis of individual patients . The aim of this paper was not to discuss the different diagnostic criteria of neurocysticercosis . However , it is important to note that 34 . 4% ( 11/32 ) of those with a diagnosis of “Probable NCC” or “Definitive NCC” according to Gabriël et al . [29] and 31 . 0% ( 9/29 ) according to Del Brutto [23] have no NCC lesion on CT . Whether these patients have cysticercosis outside the central nervous system , the serological test was false positive or the NCC lesions were missed because of lack of sensitivity of a CT scan cannot be resolved . The latter has been observed in six of seven sero-Ag-ELISA positive patients with calcified NCC lesions but without viable cysts on CT scan [36] . The majority of NCC lesions were calcifications , which may support the hypothesis that mainly calcified NCC lesions are epileptogenic [27] and seizures usually develop at a later stage of the disease . When reporting low numbers of viable cysts the probably lower sensitivity of CT scan for cystic lesions has to be considered . Two people had calcifications outside the CNS , more specifically in the masseter muscles . However , the Del Brutto criterion “Cysticercosis outside the CNS” does not include this in its description , but rather mentions multiple ‘cigar-shaped’ calcifications in the thigh and calf muscles [23] . He states that , in endemic regions , a patient may have systemic cysticercosis and neurological manifestations due to an unrelated cause . Though this is true , we nevertheless propose that evidence on either CT scan or plain X-ray films of extra CNS cysticercal lesions in the muscles of the head/thorax be included in the minor criterion “Cysticercosis outside the CNS” . Although the data presented in this study is indeed valuable in analyzing the importance of NCC in PWE , there are some limitations . The selection of people with the highest number of positive answers in the screening questionnaire could have resulted in a bias towards more severe cases of epilepsy , i . e . with more generalized tonic-clonic seizures compared to a typical epilepsy population . We cannot rule out that there are different etiologies in different types of severity of epilepsy . Additionally , focal seizures without secondary generalization are rarely diagnosed as epileptic seizures in resource poor countries . This might have led to an underestimation of people with focal seizures in our study population . Because focal seizures have usually an underlying cause , we assume that in people with only focal seizures NCC lesions might be even more common , but we cannot prove this hypothesis with our data . Unfortunately , there was no clinical information about the consecutive hospital cerebral CT scans available . Therefore , some patients in the consecutive scans could have had epilepsy as well and can therefore not be considered a true control group . By selecting controls with pathologic lesions other than NCC , the related symptoms of which were probably the reason for the scan , we tried to lower the probability of including PWE in the CT controls . However , we still think , it is valuable to get more information about the prevalence of NCC in a set of CT scans , because it might help to compare the prevalence of NCC among countries at a larger scale and therefore we have included the information of all 101 consecutive CT scans in the supplement material ( S1 Table ) . In sub-Saharan Africa many countries have very few neuroimaging facilities . Consecutive images of these centers might be used as a sentinel to evaluate the NCC prevalence in a country . In a similar study in Tanzania , Winkler et al . reported a CT based prevalence of 1% of definite NCC in controls without epilepsy [4] . In the current study , the CT scans were performed only once and without application of contrast medium . As a consequence ring enhancing lesions and resolution of lesions after treatment , which are part of the Del Brutto criteria , could not be gathered . Similarly , a fundoscopic examination to visualize subretinal cysts , an X-ray of calf or thigh muscles or histology was not available , which might have led to a slight underestimation of the prevalence of NCC among PWE . In addition , in the people with the consecutive CT scans no serological tests were available , which could have also resulted in an underestimation of NCC in this group . In conclusion , our study results of almost 60% of PWE suffering from additional NCC should be taken seriously , despite the study limitations . This is one of the highest prevalence rates based on neuroimaging reported in sub-Saharan Africa and entails the need for the commencement of concerted efforts towards the control of T . solium in endemic areas of Zambia . The findings also highlight the importance of NCC as a differential diagnosis for patients presenting with epileptic seizures not only in rural areas but also in other areas of . The present study suggests that NCC seems the most important secondary cause of epilepsy , a cause that is preventable and even eradicable [37] and therefore much more attention should be directed to its control in T . solium taeniosis/ ( neuro ) cysticercosis endemic countries like Zambia . | Neurocysticercosis , an infection caused by the larval stage of Taenia solium ( the pork tapeworm ) , is a preventable and treatable disease and one of the main causes of epilepsy in low-income countries . Studies in Zambia have indicated high endemicity of T . solium in rural communities where a high prevalence of epilepsy has been reported . It was therefore important to analyze the role of neurocysticercosis as a cause of epilepsy . Diagnosis of the condition , however , relies on neuroimaging techniques , which are not routinely available in our study area . In this imaging-based study , people with epilepsy were identified , examined by a neurologist , their brains scanned using computed tomography , and their stools and sera analyzed for specific antigens and antibodies . The study suggests that 57 . 1% of these people suffered from neurocysticercosis , making it the single most important cause of epilepsy in the study area . The results show the need to institute control measures to alleviate the suffering of the affected communities . In summary , the recognition of neurocysticercosis as an important cause of epilepsy will aid in the treatment and prevention of epilepsy in affected communities and alleviate the burden of neurocysticercosis on the rural people . | [
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] | [] | 2015 | Prevalence of Neurocysticercosis in People with Epilepsy in the Eastern Province of Zambia |
A bivalent killed whole cell oral cholera vaccine has been found to be safe and efficacious for five years in the cholera endemic setting of Kolkata , India , when given in a two dose schedule , two weeks apart . A randomized controlled trial revealed that the immune response was not significantly increased following the second dose compared to that after the first dose . We aimed to evaluate the impact of an extended four week dosing schedule on vibriocidal response . In this double blind randomized controlled non-inferiority trial , 356 Indian , non-pregnant residents aged 1 year or older were randomized to receive two doses of oral cholera vaccine at 14 and 28 day intervals . We compared vibriocidal immune responses between these schedules . Among adults , no significant differences were noted when comparing the rates of seroconversion for V . cholerae O1 Inaba following two dose regimens administered at a 14 day interval ( 55% ) vs the 28 day interval ( 58% ) . Similarly , no differences in seroconversion were demonstrated in children comparing the 14 ( 80% ) and 28 day intervals ( 77% ) . Following 14 and 28 day dosing intervals , vibriocidal response rates against V . cholerae O1 Ogawa were 45% and 49% in adults and 73% and 72% in children respectively . Responses were lower for V . cholerae O139 , but similar between dosing schedules for adults ( 20% , 20% ) and children ( 28% , 20% ) . Comparable immune responses and safety profiles between the two dosing schedules support the option for increased flexibility of current OCV dosing . Further operational research using a longer dosing regimen will provide answers to improve implementation and delivery of cholera vaccination in endemic and epidemic outbreak scenarios .
As a disease of poverty and inequity , cholera is often prevalent in areas of compromised sanitation , overcrowded conditions , and poor quality of water supply . An increasing number of longer lasting outbreaks have dramatically impacted the least developed countries ( LDCs ) , including those in Africa , South Asia , and the Hispaniola island region [1] . Living conditions in LDC populations often favor disease transmission and improvements can take a long time to achieve . In these settings , V . cholerae O1 can cause large , rapidly spreading severe outbreaks that cripple public health systems with already limited medical and financial resources . Many recent epidemics have occurred in highly susceptible and vulnerable populations ( Haiti , Zimbabwe , Central and West Africa ) , where behavioral , social , and environmental factors , as well as lower background exposure to cholera have contributed to increased duration and severity of the outbreaks [2] . Effective interventions combining surveillance , treatment , and improving water , sanitation , and hygiene ( WASH ) measures are paramount . Vaccination can complement these preventive and control strategies in areas of endemic disease or areas at risk for outbreak [3] . Recently , a killed , bivalent oral cholera vaccine ( OCV ) has been prequalified and recommended for use by the WHO . Still , this OCV has not been widely implemented in endemic areas and its use is limited to areas with established or imminent outbreaks . Safety and immunogenicity of this OCV has been demonstrated in Vietnam , India , and Bangladesh [4–6] . Seroconversion with serum vibriocidal antibodies following vaccination was found to be lower in hyper-endemic areas ( India ) compared to less endemic areas ( Vietnam ) . When participants with only low baseline serum vibriocidal titers were analyzed in the Kolkata trial , seroconversion and geometric fold rise were similar in both populations [7] . A large phase three randomized clinical trial ( RCT ) of the two-dose , killed bivalent OCV demonstrated a cumulative 65% efficacy in endemic populations over five years [8] . Earlier studies with the cholera toxin whole cell O1 vaccine revealed protection for three years in adults and for 6–12 months in young children [9–11] . In Kolkata , a RCT evaluating immune responses of the bivalent killed whole cell OCV without cholera toxin B ( Shanchol , Shantha Biotechnics Limited ) in adults and children found robust responses to a first dose but no further rise following the second dose [12] . This observed blunted immune response following the second dose may be due to the increased LPS content , as compared with older versions of killed OCV . Proposed mechanisms of the blunted immune response include blocking of subsequent antibody production by the increased LPS or a booster like effect occurring after the first dose due to recurrent natural exposure in an endemic setting . Some questions still remain unanswered with regards to the most optimal dosing regimen to assist the effective deployment of OCV in field conditions . An alternate 28 day interval could facilitate inclusion of OCV into a routine immunization schedule in cholera endemic regions . No significant difference in immune response following a four week schedule may further support the hypothesis of whether adequate protection can be offered by a single dose in endemic areas—this is currently being assessed in a large , placebo controlled RCT in Bangladesh . We aimed to assess if immune responses in a prolonged 28 day dosing interval is non-inferior to the standard 14 day schedule .
This was a double-blind , RCT conducted at the Clinical Trials Unit of the National Institute of Cholera and Enteric Diseases ( NICED ) , Kolkata , India . Recruitment , dosing and follow up were completed between January-December 2011 . The study was performed in the cholera endemic urban slums of Kolkata with similar access to water , sanitation , and health care throughout the study area . Healthy males and non-pregnant females aged ≥1 year were recruited . Exclusion criteria consisted of serious chronic disease , pregnancy , immune-compromised conditions , gastrointestinal disease , antibiotic usage in the past 14 days , or previous receipt of cholera vaccine . Potential participants with acute illness or fever had dosing deferred pending recovery . The objectives of this trial were to compare safety and serum vibriocidal antibody responses in participants receiving two OCV doses either 14 days or 28 days apart . The primary endpoint was the proportion of participants exhibiting four-fold or greater rises in serum vibriocidal antibody titers , 14 days following the second dose relative to baseline . Secondary endpoints included measurement of geometric mean titers of serum vibriocidal antibody at the above time points . Safety of the vaccine was also evaluated throughout the follow up period . Written informed consent was obtained by study physicians for all adults and parents/guardians of participating children , as well as written assent for 11–17 year old participants . The trial protocol was approved by the Scientific and Ethics Committee of NICED and the International Vaccine Institute ( IVI ) . Independent safety monitoring was conducted , with external monitoring & GCP audits performed by Shantha Biotechnics Limited . This trial was registered in India ( CTRI/2010/091/002807 ) and clinicaltrials . gov ( NCT 01233362 ) . All data from trial volunteers used for analysis was anonymized . The study vaccine ( Shanchol ) consisted of 600 ELISA units of LPS of V . cholerae O1 El Tor Inaba; 300 ELISA units of LPS each of V . cholerae O1 classical Ogawa , 300 ELISA units of LPS of V . cholera O1 Inaba and 600 ELISA units of LPS of V . cholerae O139 . Placebo vials contained E . coli K12 cells , whose appearance was identical to the study vaccine . Dosing of the study agent was administered as in Table 1 . Both placebo and vaccine were packaged as liquid formulations in identical vials containing 1 . 5 mL doses and were stored at 2–8°C . The study agent was given in two doses separated by a two week or a four week interval and administered by oral syringe , after which each participant was offered a cup of water . Participants were observed in the trials unit for 30 minutes following dosing , as well as for 3 days after each dosing . During each follow up day , study physicians conducted a structured interview regarding the participant’s overall health and any occurrence of adverse events . Diarrhea was defined as three or more loose or liquid stools in a 24 hour period . Blood samples were obtained prior to the first dose and 14 days after each study agent dose . Sera were separated and stored at -70°C until paired testing was performed . The microtiter technique was used to detect serum vibriocidal antibodies to V . cholerae O1 El Tor Inaba , O1 Ogawa , and O139 [13] . Participants were stratified by age group ( 1–5y , 6–10y , 11–17y , and ≥18y ) . Randomization numbers were generated in blocks of at least four , which included equal numbers of each arm , to ensure that balance between treatments was maintained . These lists were prepared by a statistician not involved in the study . Study agents were pre-labeled by Shantha personnel , who were not involved in the conduct or monitoring of the trial . All study staff and participants were blinded to treatment assignment for the duration of the study . Sample size calculation was driven by seroconversion after two doses under 14 day and 28 day dosing intervals . Among participants , we assumed 45% seroconversion in adults and 80% in children after 2 doses . If the seroconversion rate in the 28 day dosing interval is no less than 20% than that in the 14 day dosing interval , it will be considered to be non-inferior . This threshold was selected based on seroconversion rates and their corresponding lower bounds of the one tailed 95% confidence interval from previous studies using the same vaccine in the same setting[5 , 12] . Assuming a one tailed α = 0 . 05 , 80% power , a 15% drop-out rate , and using the score method of non-inferiority test [14] , a total of 89 participants per study group had been considered . Thus , a total of 356 participants were targeted , 178 in each dosing regimen . Data were entered in a web-based data capture system and analyses were performed in SAS 9·3 ( SAS Institute , Cary NC ) . Analyses for comparisons of dichotomous outcomes such as adverse events and seroconversion were performed with the chi-square test or Fisher’s exact test if cell counts were sparse . For comparisons of vibriocidal titers , Student’s t-test was performed using the pooled or Satterthwaite method depending on whether the variances were equal or not . Nonparametric Wilcoxon rank-sum test and Kolmogorov-Simirnov test were performed when data were not normally distributed . Comparisons of the primary outcomes , vibriocidal seroconversion were evaluated with one-tailed 95% confidence intervals using the Wilson Score method[15] . Statistical evaluations of all other comparisons were two tailed .
Recruitment of participant flow is illustrated in Fig . 1A total of 356 participants ( 178 children , 178 adults ) were recruited from January 2010 to October 2011 . Among eligible participants , 86/89 adults ( 96 . 6% ) and 84/89 children ( 94 . 4% ) in the 14 day interval arm and 84/89 adults ( 94 . 4% ) and 82/89 children ( 92 . 1% ) of the 28 day interval arm took all three doses of the assigned study agent and provided all four blood samples . A total of 20 participants ( 5 . 6% ) were lost to follow up or were found to be ineligible to continue following study visit screening . There were no significant differences in demographic characteristics between intervention arms among each age group ( Table 2 ) . All participants randomized in the study were included in safety outcome analysis . No statistically significant differences in the rates of adverse events between each intervention group were noted ( Table 3 ) . A total of 10 adverse events ( AE ) were reported within 3 days of either dose . The most commonly reported AEs were fever ( n = 3 ) , general ill feeling ( n = 2 ) , vomiting , diarrhea , and headache ( with n = 1 each ) , with no statistically significant differences between children or adults . No serious adverse events were reported during the trial . A per protocol analysis was conducted for immunogenicity data , including 336 participants who completed all planned study visits . Immune responses to V . cholerae O1 Inaba , O1 Ogawa , and O139 following administration of two doses of vaccine in a 28-day schedule were non-inferior to those of a 14 day schedule , as the difference measured was greater than the pre-defined cut-off of-20% ( Tables 4 , 5 ) . No significant difference between dosing schedules was observed in percentage of seroconversion after the first or second dose . Baseline vibriocidal geometric mean titers ( GMT ) to O1 Inaba ranged from 94 to 275 in adults and from 29 to 140 in children . The geometric mean fold ( GMF ) rise was higher in children ( ranging from 7 . 5–26 . 9 ) than in adults ( 3 . 4–6 . 4 ) . No statistically significant difference was noted between intervention arms in seroconversion or geometric fold rise . The GMF rise from baseline was higher for O1 Inaba in adults , after receipt of the first dose ( 6 . 8 and 8 . 9 respectively in the 14 and 28 day interval arms ) compared to receipt of the second dose ( 4 . 6 and 4 . 7 respectively ) . In children the responses were more pronounced with GMF rise from baseline after first dose in both the arms being 29 . 7 and 20 . 8 respectively . The GMF rise after second dose was 17 . 5 and 10 . 7 respectively . Rise in titers to V . cholerae O1 were higher among individuals with lower baseline vibriocidal titers , as seen in Table 5 . This magnified response was likely due to the lower baseline GMT observed in children , suggesting lower natural exposure . Adults with baseline GMTs lower than the median ( <160 ) demonstrated high GMF rise ( >10 ) and seroconversion ( ~85% ) in both interval groups , which were markedly higher than adults with higher baseline GMTs ( Table 6 ) . Comparable results were noted in children , although median baseline titers were lower ( 80 ) . There was a significantly higher GMF rise in children aged 1–5 years old in the 14 vs 28 day interval groups ( 34 . 7 vs 10 , p = 0 . 01 ) , though no significant difference in seroconversion was noted . This difference is most likely explained by the significantly higher baseline GMT detected in 1–5 year olds between the 14 and 28 day interval group ( 14 . 1 vs 69 . 6 , p = 0 . 01 , S1 Table ) . No other significant differences were noted in any other age group . When controlling for baseline GMT , a multiple linear regression model of log transformed titers did not find any significant difference between the two dosing intervals ( -0 . 13 dosing interval effect comparing the 28 day interval to the 14 day interval , p = 0 . 33 , S2 Table ) . Similar observations were also found for O1 Ogawa . Following the second dose , adults demonstrated GMFr of 4 . 1 with 45% seroconversion and 3 . 8 with 49% seroconversion to O1 Ogawa in 14 and 28 day interval groups ( Tables 4 , 5 ) . Children exhibited GMFr of 11 . 1 with 73% seroconversion and 7 . 8 with 72% seroconversion . As with previous trials in Vietnam , India , and Bangladesh , immunogenicity against O139 was poor in both schedules [4–6] .
The results of our study support flexibility in dosing Shanchol in endemic settings , where strict schedules may be difficult to adhere to . As with any immunogenicity findings , vibriocidal antibodies do not truly reflect a protective response and , at best , are an indirect correlate of protection that is not absolute . While only a field trial can determine true effectiveness of altered dosing regimens , interpreting this data in light of the existing immunogenicity and clinical efficacy data in the same setting may provide a foundation for policy makers to ease implementation of OCV as part of a control strategy for cholera . Both schedules were well tolerated by all recipients with comparable safety profiles between either group . Our findings were compatible with previous studies that revealed that high baseline vibriocidal titers were associated with reduced post-vaccination serum vibriocidal antibody responses[5 , 16] . Higher baseline titers found among participants were most likely due to prior exposure to V . cholerae since the area is cholera-endemic and the population had not earlier received cholera vaccine . The first dose of the vaccine may have elicited memory immune responses among previously exposed individuals resulting in a rise in vibriocidal titers with no further rises after the second dose . In children , the baseline vibriocidal titers were lower , suggesting lower earlier exposure in this age group . Lower baseline titers were associated with higher GMF rise increases following vaccination , with a higher percentage of responders in this age group , though the clinical significance of this finding is unclear . Although it is possible that these results reflect chance , a recurrent theme of immune differences in children under five years of age relating to OCV does occur , and it is possible that in this sub-population that there may be a difference between the two regimens . Our study confirms earlier findings that the two-dose regimen of the killed whole-cell OCV is safe , well-tolerated , and immunogenic [17] . Vibriocidal responses to O1 Inaba were higher in both adults and children following the first dose , as compared to the second dose , with GMFr rises higher in children , likely related to the inverse relation of baseline serum titers mentioned above . Whether the lower responses to O139 indicate that the vaccine elicits poorer responses to O139 or if this reflects differences in assay sensitivity remains an aspect that needs to be explored with additional scientific data . V . cholerae O139 continues to be infrequently isolated from environmental samples but has not been responsible for any large outbreak in the past 10 years . The lack of circulating O139 strains could be a possible factor for the lower immune response to O139 antigen in the vaccine . Serum vibriocidal antibody responses were shown to be no higher following a second dose , when compared to levels after the first dose . This contrasts with the older generation killed OCV ( Dukoral ) , for which serum titers increased further after the second dose[18] . The current reformulated killed whole cell vaccine ( Shanchol ) elicits higher serum vibriocidal responses than the older version of Dukoral . It exhibits no augmentation of these responses after the second dose as compared with the first , perhaps because it has an approximately two times higher LPS content than the older vaccine[4] . This marked difference in magnitude and pattern of immune responses motivated the current evaluation of whether extending the interval between doses has an impact on the vibriocidal response . While extending the dosing interval did not raise immune responses to the second dose , the mechanism behind this observed lack of boosting remains unclear . Since the vibriocidal antibody response does not truly reflect protection , and at best is an indirect correlate of protection , our immunogenicity results are not sufficient to support a hypothesis that a single dose regimen may confer similar efficacy as two doses . Nevertheless , the similarity of immune responses to shorter versus longer inter-dose intervals provides some reassurance that flexibility in dosing , particularly extending the intervals beyond 14 days , will not vitiate vaccine response . Comparable immune responses between different dosing regimen schedules would support additional uses of vaccination as part of a comprehensive strategy . In endemic settings , policy makers could entertain extending the dose interval to 1 month , which could ease delivery by facilitating national routine immunization strategies and linking OCV with other health interventions to populations in high risk regions . These results may be of particular interest in complex outbreaks , such as those seen following a natural disaster . A reactive vaccination strategy provides vaccine following a cholera outbreak to prevent further disease transmission with hopes of shortening outbreak duration . It relies on getting the first dose to affected populations as soon as possible . After the first dose is distributed , one month interval could allow the focus to return to stabilization of infrastructure and water sanitation . This is pertinent to a post disaster context in resource limited areas , which can be a common scenario for cholera outbreaks in both endemic and non-endemic areas ( Indonesia tsunami , Haitian earthquake , Pakistan floods ) . Since this study was conducted in an endemic area and a population with pre-existing vibriocidal antibodies , the results may be different than what can be expected from non-endemic areas . Evaluations of a longer dosing interval in these settings are needed since immunogenicity and overall vaccine impact may likely be impacted by recurrent exposure , or ‘natural boosting’ . With no further rise in seroconversion rates after a second dose , efforts to evaluate a efficacy of a single dose regimen in a clinical field trial is underway to evaluate its potential use in an epidemic setting [19] . From a programmatic standpoint , additional exploration into serum and gut responses when spacing out the dosing interval even further may broaden our knowledge on public health benefits with regards to the amount and duration of clinical protection offered by this OCV . Cholera remains a major global health concern and is an important threat to most developing countries , especially in areas where overcrowding and poor sanitation are common . Large outbreaks often involve populations affected by natural disasters or those displaced by war , where there is inadequate sewage disposal and contaminated water . In spite of current WHO support for use of OCV as part of a prevention and control package for cholera endemic areas , the international community is still exploring the best methods to implement these recommendations . Flexibility with the administration of two doses over one month could ease logistical requirements in a complex outbreak setting , allowing for stabilization of community infrastructure , as well as linking vaccination with other vital community interventions , resulting in the enhanced delivery of OCV . By demonstrating similar immunologic responses to different dosing regimens , with no additional safety risk , further operational research testing even longer inter-dose intervals could provide helpful answers to improve decision making to fill critical knowledge gaps for vaccination in endemic , epidemic , and outbreak scenarios . | The five year efficacy results of the bivalent , killed whole cell oral cholera vaccine was shown to offer 65% protection in cholera endemic Kolkata . Currently , two oral cholera vaccines ( OCV ) are prequalified by the World Health Organization: the whole cell recombinant cholera toxin B subunit vaccine ( Dukoral ) , and the bivalent killed whole cell only OCV ( Shanchol ) . Shanchol , which is less expensive and possibly associated with longer protection , is recommended in a two dose schedule to be given at two weeks apart . Large scale cholera outbreaks often affect vulnerable populations with limited access to care . Strict dosing schedules can create further logistical barriers , hindering proper vaccine delivery to affected residents returning for their second OCV dose . In this study , 356 participants aged 1 year or older were randomized to receive two doses of OCV at 14 or 28 day intervals , for which vibriocidal immune responses were compared . Similar immune responses were demonstrated between a two and four week OCV dosing schedule , which can increase flexibility when offered as part of a targeted vaccination program . This can further serve to increase adherence and completion of the recommended dosing regimen , as well as providing a platform to increase coverage of other beneficial non-vaccine interventions . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [] | 2015 | Flexibility of Oral Cholera Vaccine Dosing—A Randomized Controlled Trial Measuring Immune Responses Following Alternative Vaccination Schedules in a Cholera Hyper-Endemic Zone |
Dihydropyridines ( DHPs ) are L-type calcium channel ( Cav1 ) blockers prescribed to treat several diseases including hypertension . Cav1 channels normally exist in three states: a resting closed state , an open state that is triggered by membrane depolarization , followed by a non-conducting inactivated state that is triggered by the influx of calcium ions , and a rapid change in voltage . DHP binding is thought to alter the conformation of the channel , possibly by engaging a mechanism similar to voltage dependent inactivation , and locking a calcium ion in the pore , thereby blocking channel conductance . As a Cav1 channel crystal structure is lacking , the current model of DHP action has largely been achieved by investigating the role of candidate Cav1 residues in mediating DHP-sensitivity . To better understand DHP-block and identify additional Cav1 residues important for DHP-sensitivity , we screened 440 , 000 randomly mutated Caenorhabditis elegans genomes for worms resistant to DHP-induced growth defects . We identified 30 missense mutations in the worm Cav1 pore-forming ( α1 ) subunit , including eleven in conserved residues known to be necessary for DHP-binding . The remaining polymorphisms are in eight conserved residues not previously associated with DHP-sensitivity . Intriguingly , all of the worm mutants that we analyzed phenotypically exhibited increased channel activity . We also created orthologous mutations in the rat α1C subunit and examined the DHP-block of current through the mutant channels in culture . Six of the seven mutant channels examined either decreased the DHP-sensitivity of the channel and/or exhibited significant residual current at DHP concentrations sufficient to block wild-type channels . Our results further support the idea that DHP-block is intimately associated with voltage dependent inactivation and underscores the utility of C . elegans as a screening tool to identify residues important for DHP interaction with mammalian Cav1 channels .
Calcium influx into the cell can make both an immediate and a long-term impact on many cellular processes , including membrane excitability , neurotransmitter release , muscle contraction , transcription , and proliferation [1] . It is therefore not surprising that calcium current through calcium-permeable channels is tightly regulated . Calcium channels of the Cav family , for example , regulate current through voltage dependent mechanisms and exist in three major biophysical states: a resting ( non-conducting ) state , an activated ( conducting ) state that is triggered by a rapid change of voltage , and an inactivated ( non-conducting ) state that follows the conducting state but is refractory to activation . Cav channels are composed of multiple subunits that regulate calcium current through the pore-forming and voltage-sensing α1 subunit , which has four repeating domains ( called I–IV ) , each with six transmembrane segments designated S1–S6 [2] , [3] ( Figure 1 ) . The most widely-expressed Cav channels are the L-type , otherwise known as Cav1 channels . In humans , alterations in Cav1 abundance or function result in a variety of diseases , including hypertension [4] , Timothy Syndrome [5] , and congenital stationary night blindness [6] . A better understanding of Cav1 function and how this function can be modulated by drug-like molecules may therefore lead to better treatments of Cav1-related diseases . A distinguishing feature of Cav1 channels is their sensitivity to 1 , 4-dihydropyridines ( DHPs ) , which bind to the α1 subunit and are commonly prescribed to reduce hypertension in humans [7] . Exactly how DHPs physically interact with the α1 subunit is incompletely understood , in large part because the crystal structure of any Cav1 α1 subunit remains elusive . However , Cav1 residues that facilitate DHP interaction have been identified through the analyses of candidate domains and residues . For example , swapping domains IIIS5 , IIIS6 , and IVS5-IVS6 from an L-type to a non-L-type channel is sufficient to confer DHP-sensitivity to otherwise insensitive Cav channels [8] , [9] , [10] . Individual Cav1 residues present in DHP-sensitive , but not insensitive channels , have also been extensively investigated to determine if they are necessary and sufficient for DHP-sensitivity [11] , [12] , [13] , [14] . In this way , four residues in domain IS6 , two residues in IIIS5 , eight residues in IIIS6 , and six residues in IVS6 were shown to be important for DHP binding to the α1 subunit of the Cav1 channel in culture [11] , [14] , [15] , [16] , [17] , [18] ( summarized in Table 1 ) . DHPs are thought to block calcium current by inducing allosteric structural changes in the channel that engages the channel in the inactivated state [18] , [19] and promotes high affinity interaction with a single calcium ion within the pore [20] , [21] , [22] , thereby preventing further ion flow . Calcium ions are bound by the channel through four negatively charged glutamate residues ( called the calcium selectivity filter ) , which reside within four membrane-embedded intersegmental loops between S5 and S6 ( called SS1–SS2 ) of each domain ( Table 1 ) . Together , the four SS1–SS2 domains form the extracellular-facing ‘outer pore’ of the channel ( see Figure 1 ) . The relationship between the DHP and the selectivity filter is bidirectional: DHP-binding facilitates a high affinity interaction between the selectivity filter and a calcium ion , and the coupling of the selectivity filter to calcium promotes a high-affinity interaction between the DHP and the channel [21] , [23] , [24] . A candidate analysis of residues surrounding the glutamate selectivity filter revealed several residues that are also important for high-affinity DHP interaction with the channel in culture [21] , [24] , [25] , [26] ( Table 1 ) . Clearly , the candidate approach has been fruitful in identifying residues important for DHP antagonism of Cav1 channels . However , as the exact molecular mechanism of DHP block remains incompletely understood , the search for additional residues involved in DHP-sensitivity might add to our understanding of DHP-Cav1 interaction . We therefore screened random mutants of the tiny nematode worm C . elegans for those animals that are resistant to the effects of a novel DHP analog called nemadipine in the hopes of identifying new Cav1 residues important for DHP-sensitivity . We recently discovered nemadipine through a screen for new biologically active small molecules [27] , [28] . Through a variety of genetic approaches , we demonstrated that the major target of nemadipine is EGL-19 [27] , which is the sole Cav1 channel α1 subunit encoded by the C . elegans genome [29] . We also found that nemadipine can antagonize vertebrate Cav1 channels with a similar potency as an FDA-approved DHP in chick ciliary neurons [27] . However , none of the FDA-approved Cav1 antagonists examined , including nifedipine , nicardipine , felodipine , and nimodipine , elicit robust phenotypes in whole worms . Worm insensitivity to these DHPs is not likely due to Cav1 divergence since nifedipine can antagonize EGL-19 in dissected worms [29] . Instead , we discovered that most FDA-approved DHPs fail to accumulate in whole worms [27] , consistent with the idea that C . elegans has extensive xenobiotic defenses . Thus , nemadipine is a unique reagent to genetically probe interactions between DHPs and Cav1 channels in vivo . Here , we present our genetic screen of 440 , 000 mutant C . elegans genomes for animals that are resistant to nemadipine . Strikingly , 30 of the 35 mutants that we characterized have missense polymorphisms in the egl-19 gene . Only one of these mutations is in a non-conserved residue . Eleven of the mutations correspond to residues of the Cav1 channel that are known to be required for DHP interaction with mammalian channels . The remaining mutations are in nine residues not previously associated with DHP interaction , eight of which are conserved . Behavioral and developmental analyses of 12 of our worm mutants revealed that all exhibit phenotypes consistent with increased channel activity . For seven of these worm mutations , we created orthologous mutations in the rat brain Cav1 . 2 channel and examined the ability of DHPs to block ion current through these mutant channels using whole-cell voltage clamp analysis in culture . Two of the mutant channels have decreased sensitivity to the DHP block , and six disrupt the ability of the DHP to block ion current completely . Thus , screening for worm mutants that are resistant to nemadipine is a viable and novel approach to identify mammalian Cav1 residues that are important for DHP sensitivity .
Previous work demonstrated the utility of using C . elegans calcium channels as a model to better understand mammalian channel function [30] . We therefore performed a forward genetic screen for C . elegans mutants that are resistant to nemadipine-induced growth retardation to identify residues in the Cav1 calcium channel required for DHP sensitivity . A total of 440 , 000 haploid mutagenized wild type genomes were screened and 55 candidate mutants were isolated . Homozygous lines were established for 35 of these mutants and polymorphisms were identified in egl-19 for 30 ( Figure 1 and Table 2 ) . Four of the remaining mutants are genetically linked to egl-19 on chromosome IV , but no polymorphism in egl-19 could be found ( data not shown ) . The remaining mutant is not linked to egl-19 and is not discussed further here . Upon out-crossing select mutants , we found that all of them exhibited increased resistance to nemadipine in two separate dose-response assays ( Table 2 , Figure S1 ) , confirming that they are indeed DHP-resistant mutants . The DHP binding site on the Cav1 channel pore-forming α1c-subunit of vertebrate channels is defined by at least nine residues that are both necessary and sufficient to confer DHP-sensitivity to a non-L-type calcium channel ( Figure 1 , Table 1 ) . Two of these residues reside in helix S5 of domain III ( called IIIS5 ) , three reside in IIIS6 and four in IVS6 [12] , [13] , [14] . Other studies have shown the involvement of additional residues required for DHP interactions ( Table 1 ) [15] , [18] . From our screen for nemadipine-resistant mutants , we found ten mutants with polymorphisms within the IIIS6 domain of EGL-19 . Six of these mutate a single EGL-19 residue , M1056 ( rat α1c residue M1161 ) , critical for high-affinity DHP binding to the vertebrate channel [11] , [14] . We found a seventh mutation in EGL-19 residue M1055 ( rat α1c residue M1160 ) that is also necessary for DHP binding to the vertebrate channel [11] . An eighth polymorphism was detected in the EGL-19 residue A1041 ( rat α1c residue S1146 ) . Previous work demonstrated that a rat α1c S1146A mutation does not disrupt DHP binding [25] , which is perhaps not surprising given that the alanine is conserved in some Cav1 channels [25] . However , no one has yet tested if other residues in position 1146 , such as the valine substitution found in our mutant , would disrupt interactions with DHPs . Finally , two polymorphisms were identified in the EGL-19 residue V1060 ( rat α1c residue V1165 ) within IIIS6 that has yet to be tested for its role in DHP block . Our finding that seven of the ten mutations that we identified within the IIIS6 region are required for DHP-interaction in mammalian Cav1 . 2 channels validates our genetic approach for identifying residues required for DHP-sensitivity , and suggests that the new residues we uncovered are also important determinants of DHP sensitivity . High-affinity DHP binding is coupled to the interaction between the selectivity filter in the outer pore of the α1 subunit and calcium ions ( see introduction ) . Mutating the glutamate residues and several of the surrounding residues of the selectivity filter not only dramatically reduces the binding of calcium ions , but disrupts DHPs calcium-dependent affinity for the channel [21] . For example , the S1115 residue within SS1-SS2 of the rat α1c channel that precedes the key glutamate in domain III has been shown to be necessary for DHP-sensitivity ( Figure 1 , Table 1 ) [25] , [26] . We found that two of the nemadipine-resistant mutants are polymorphic within the EGL-19 residue S1010 that is equivalent to the rat α1c residue S1115 ( Figure 1; Table 2 ) . We also isolated a mutant with a polymorphism in a similar position in domain II ( EGL-19 residue I652; rat α1c residue I702 ) . However , this mutant had a second polymorphism in V679 ( rat residue V729 ) , confounding the attribution of any mutant property to either polymorphism . Nevertheless , these results suggest that disrupting residues within the outer pore can reduce DHP-sensitivity in vivo . DHP antagonists have high affinity for Cav1 channels in the inactive state [31] . After voltage dependent opening of the channel , the channel is coordinately inactivated by both calcium-dependent and voltage dependent mechanisms . Previous work has shown that voltage dependent slow inactivation of Cav1 . 2 channels is mediated by hydrophobic residues in the cytoplasmic end of the S6 transmembrane helices of domains I-IV [32] , [33] , [34] . Together , these residues are thought to form a hydrophobic voltage-sensitive annulus , the disruption of which leads to prolonged activation of the channel . We found 8 nemadipine-resistant mutants with a polymorphism in EGL-19 A702 within IIS6 , which is orthologous to the residue in vertebrates ( rat α1c residue A752 ) that is most important in slow inactivation [32] . We also found two mutants with a polymorphism in the hydrophobic residue V1060 in the IIIS3 domain of EGL-19 , which precedes an orthologous residue that contributes to slow inactivation by three positions ( rat α1c residue V1168 ) [33] ( Figure 1 , Table 1 ) . We also found three mutants with a polymorphism in the hydrophobic residue G365 ( rat α1c residue G402 ) of domain IS6 that is also required for voltage dependent slow inactivation [5] , [35] . The same G365R mutation that we isolated was found independently by Michael Hengartner and shown to prolong the activation phase of the mutant EGL-19 channel by Raymond Lee and colleagues in vivo [29] . Importantly , the isolation of at least eleven mutants within the voltage-sensitive annulus suggests that DHP-blockage is linked to voltage dependent slow inactivation in vivo . Finally , we identified seven other mutants that have polymorphisms in six residues within regions of the Cav1 channel that have little functional annotation ( A236 , A615 , V679 , T1066 , E1079 , and G1594 ) ( Figure 1 , Table 2 ) . Of note , two of these mutant residues are in analogous positions in IS5 and IIS5 ( A236 and A615 , respectively ) . Five of the six mutant residues are in highly conserved regions of the channel . The sixth mutant residue ( G1594D ) is in the carboxy-terminal tail , whose sequence is evolutionarily divergent ( Figure 1 ) . We have further investigated three of these mutants through both phenotypic and electrophysiological analysis as described below . In contrast to traditional electrophysiological approaches that rely on transgenics to investigate DHP interactions , our genomic mutations provided us with a unique opportunity to investigate the consequences of these polymorphisms on channel activity within the context of a whole animal . We reasoned that phenotypic analysis could be used to approximate the level of Cav1 mutant channel activity in vivo because the level of Cav1 EGL-19 channel activity has dramatic consequences on the development and behavior of the animals [27] , [29] . For example , reduction-of-function mutations or hypomorphs , such as egl-19 ( n582 ) , are long and flaccid . By contrast , mutant channels that have a prolonged activation phase in electrophysiological studies , such as the egl-19 ( n2368 ) hypermorphic mutant , are myotonic and consequently shorter than wild type ( Figures 2 and 3 ) [29] . The defects in body size and rigidity in egl-19 mutants are likely due to altered EGL-19 function in the 95 striated muscles that line the body wall of the worm . The more active the channel , the more pronounced the body size phenotypes become [27] , [29] . Thus , if the nemadipine-resistant mutants have unaltered behaviors compared to the wild-type , we might infer that these residues might directly alter DHP binding without altering channel activity . Alternatively , if the nemadipine-resistant mutants behaved more like hypermorphs , then the mutant residues might prolong the activation phase of the channel and confer DHP-resistance indirectly . Hypomorphic mutations were neither expected nor observed , since even weak egl-19 reduction-of-function mutations dramatically decrease the viability of animals on nemadipine and would not be recovered in our screen [27] . We began our phenotypic analysis by examining the body length of twelve of our mutant strains ( Table 2 , Figure 3 ) . We chose these 12 mutants because some represent residues that are known to be required for DHP interaction in other systems ( worm alleles tr92 , tr69 , tr89 , and tr71 ) , while the remaining ones represent potentially new residues not previously implicated in DHP-sensitivity . We measured the length of 20 young adults for each strain and compared their lengths to wild type worms and previously isolated egl-19 hypermorphs and hypomorphs . None of our twelve mutants were as short as either the weak ( ad695 ) or the strong hypermorph ( n2368 ) ( Figure 3 ) . However , two of our mutants were significantly shorter than wild type animals ( p<0 . 001 ) , suggesting that they might be hypermorphic . Next , we examined behaviors that have a greater dynamic range between wild type animals and previously characterized egl-19 mutants , as these may be more sensitive to changes in EGL-19 activity . For example , egl-19 hypomorphs accumulate two-fold or more embryos than the wild-type . Conversely , egl-19 hypermorphs retain approximately two-fold fewer embryos than the wild-type control ( Figures 2 and 3 ) . egl-19 plays important roles in several cells of the egg-laying circuit , including the eight vulval muscle cells that directly control vulva opening during embryo deposition [27] , [29] , [36] . Two thirds of our nemadipine-resistant mutants behaved similar to the egl-19 hypermorphs , retaining significantly fewer embryos than the wild-type ( p<0 . 001 ) , consistent with the idea that these mutants are also hypermorphic ( Figure 3 ) . Although all of our 12 egl-19 alleles that we examined exhibit the same trends in the length and egg-retention assays , it is unclear to us why tr81 and tr86 behave slightly differently . It may be that background mutations have a mild influence on these two behaviors in these two mutants . As expected , we found a positive correlation ( 0 . 64 ) between the strength of the egg-laying phenotype and the degree of resistance to nemadipine ( Table S1 ) . Finally , we examined how our mutants affect the contraction of a specialized set of muscles in the male tail that regulate spicule protrusion . The spicules are a pair of sclerotized cuticle structures that are housed in the male tail and used during the transfer of sperm to hermaphrodites [37] ( Figure 2 ) . The protrusion of each spicule is controlled by a pair of protraction and retraction muscles . egl-19 is a key regulator of spicule-associated muscles and functions autonomously to regulate their contraction [38] . Nearly all males of a previously characterized egl-19 hypermorph have protruded spicules , called a Prc phenotype , while egl-19 hypomorphic males cannot protrude their spicules ( Figure 2 and 3 ) [29] , [38] . We found that all of our nemadipine-resistant mutants were obviously Prc ( Figure 3 ) . The Prc phenotype of all egl-19 alleles examined was suppressed by nemadipine ( Figure 3 ) . Nemadipine similarly suppressed the shortened length and constitutive egg-laying phenotypes of our egl-19 mutants ( Figure 3 ) . This demonstrates that the observed spicule protrusion , constitutive egg-laying and shortened length phenotypes are all mediated by increased egl-19 activity . It is noteworthy that the egl-19 mutations that correspond to the rat α1c residues M1160 , M1161 , and V1165 , which are necessary for high-affinity DHP binding [12] , [13] , [14] , also show increased EGL-19 channel activity . Together , our phenotypic analyses suggest that our nemadipine-resistant mutants have increased channel activity in vivo . We tested if the mutant residues of EGL-19 , which confers DHP-resistance and increased channel activity in worms , also alter DHP-sensitivity of a mammalian Cav1 . 2 channel . We created orthologous mutations in the α1C subunit of the rat Cav1 . 2 channel and transiently expressed these in tsA201 cells along with auxiliary Cav1 . 2 channel components β2a and α2-δ [39] , [40] . Ion currents through both wild-type and mutant channels were then measured in the presence of increasing concentrations of nifedipine , a popular DHP , using whole-cell voltage-clamp electrophysiological techniques ( see methods for details ) . In this way , we could determine if the orthologous rat mutations affect the ability of DHPs to block Cav1 . 2 channel-mediated ionic current . We first examined polymorphisms in domain IIIS6 residues previously shown to be required for DHP binding . As a positive control , we created a rat α1C M1161A mutation that is known to disrupt the high-affinity interaction between DHPs and the Cav1 . 2 channel [11] , [14] . The M1161A mutation increased the IC50 of nifedipine block ∼100-fold compared to the wild type channel , confirming the importance of this residue in DHP sensitivity ( Figure 4 ) . The nearby V1165L mutation also increases the IC50 of nifedipine block by about 30-fold , consistent with its role in high-affinity DHP interactions [18] . Curiously , a leucine substitution in residue M1160 , which is also required for DHP interaction [11] , did not increase the IC50 of nifedipine block ( Figure 4 ) . T1171I , which is cytoplasmic and follows the IIIS6 transmembrane domain , also did not exhibit a shift in the IC50 . However , both M1160L and T1171I exhibited large residual currents at 100 µM nifedipine , which is sufficient to completely block the ion current through wild type channels ( Figure 4 ) . The validity of the residual current was further supported by a comparison to the complete block of current through the respective wild-type and mutant channels by 10 µM cadmium , which is a non-selective pore blocker commonly used to block all current through Cav1 channels [41] . Next , we examined two mutant residues positioned analogously in IS5 ( A283V ) , and IIS5 ( A/S666V ) . Although no significant shift was observed in the IC50 response to nifedipine for either channel , the S666V mutant left a significant residual current at saturating concentrations of nifedipine . It is currently unclear why the A283V polymorphic channel behaves similar to the wild-type channel in our assay . Finally , we examined a pair of mutations in a single residue in IIS6 ( A752T/V ) that contributes to the hydrophobic annulus required for voltage dependent slow inactivation [32] . Consistent with these previous observations , we found that both A752T and A752V yield a significant residual current of about 6-8% , likely reflecting the importance of this site for slow inactivation . Surprisingly , we also found that the IC50 of nifedipine block for A752V , but not A752T , increases about five-fold ( Figure 4 ) . As expected , we found positive correlations between the IC50 values of the mutant rat channels on nifedipine and the degree of resistance by the respective worm mutants on nemadipine ( 0 . 83 ) , and between the amount of residual current of the mutant rat channels and the degree of resistance of the worm mutants ( 0 . 63 ) ( Table S1 ) . In total , two of the seven mutant channels exhibit decreased DHP-sensitivity as indicated by an increased IC50 for the DHP-block , and six of the seven mutant channels resist a completely blocked state . These electrophysiology results therefore demonstrate that the identification of Cav1 . 2 residues that are required for DHP-sensitivity in the worm can reveal residues that are also important for mammalian DHP-sensitivity .
Here , we set out to identify Cav1 residues that are important for DHP sensitivity in vivo through a forward genetic screen . Our approach was made possible through our discovery of nemadipine , which is the only DHP known to robustly antagonize Cav1 channels within the context of undissected C . elegans worms [27] . By screening for mutants that are resistant to the effects of nemadipine , we identified 30 independent Cav1 polymorphisms that reduce DHP-sensitivity . Upon further investigating 12 of these , we found that all likely increase channel activity in whole worms . Further investigation of how these mutant residues alter the response of a mammalian ortholog to DHP-treatment not only led us to new residues required for DHP sensitivity , but provides in vivo support for a previously established idea that the DHP block likely requires voltage dependent inactivation of the channel ( see below ) . Evidence suggests that at least five of the 14 Cav1 polymorphic residues that we identified are required for physical interactions with DHPs . Four of the residues we identified ( corresponding to rat α1c residues S1115 , M1160 , M1161 , and V1165 ) were previously shown to decrease DHP affinity for the channel when mutated [14] , [18] , [25] . Additionally , two mutant residues ( corresponding to rat α1c residues A752V and V1165L ) showed a significant increase in the IC50 response to nifedipine in our electrophysiology assay . The observed shift in the IC50 is consistent with a decrease in the affinity of the DHP for the mutant channels . Previous biochemical analyses identified residues within IS6 [15] , IIIS5 , IIIS6 , and IVS6 [11] , [16] that are important for DHP binding . However , our genetic screen did not reveal polymorphisms in many of these residues . One reason for this is that our screen was not done to saturation; only six of the 14 residues identified are represented by multiple alleles . It is therefore likely that several residues required for DHP sensitivity in vivo remain unidentified , which is proven by the absence of the previously identified A906V nemadipine-resistant mutation [27] , [29] from our screen . Second , many of the residues not identified in the genetic screen play only ancillary or weak roles in DHP binding , including those of IS6 [15] and IVS6 [11] and may not confer DHP-resistance by themselves in vivo . Third , polymorphisms in each of several IVS6 residues involved in DHP-binding significantly disrupt ion flux through the channel [11] and these types of loss-of-function mutations would not be recovered in our forward genetic screen . Fourth , nemadipine may interact with EGL-19 in slightly different ways compared to how canonical DHPs interact with mammalian channels . If true , this difference might prevent the recovery of mutations of some residues required for high affinity DHP interactions with mammalian channels in our worm screen . However , it remains possible that many of the Cav1 . 2 residues shown to be required for high-affinity DHP interactions in culture play little or no role in DHP interactions in vivo . Regardless , our finding that polymorphisms in S1115 , M1160 , M1161 and V1165 do confer DHP-resistance in vivo adds further weight to existing biochemical and electrophysiological evidence that these residues are critical for DHP interactions . Several lines of evidence from our work support the previously established idea that DHP-sensitivity of the channel is intimately associated with voltage dependent inactivation [18] , [19] . First , our behavioral analyses of mutant worms reveals that polymorphisms within or near the predicted hydrophobic annulus ( corresponding to rat α1C polymorphisms G402R , A752T , A752V , V1165L , and T1171I ) , which are normally required to block the channel through voltage dependent slow inactivation , decrease DHP-sensitivity and increase channel activity in vivo . Second , electrophysiological analysis on four of these polymorphisms ( A752T , A752V , V1165L , and T1171I ) reveal significant residual current in the presence of DHP concentrations that is sufficient to block the wild-type channel in culture and two of these polymorphisms ( A752T and T1171I ) do not show any change in the IC50 for the DHP-block . This suggests that these residues are required to engage the DHP-block without being involved in DHP-sensitivity . Because the A752T mutation is known to disrupt voltage-gated slow inactivation [32] ( and T1171 is very close to V1168 and I1169 that are also required for voltage-gated slow inactivation [33] ) , our observations provide additional support for the previously established idea that the DHP block may be mediated through the same mechanism employed by voltage-gated slow inactivation [18] , [19] . Together , our work demonstrates that a forward genetic screen for worms resistant to DHPs is a viable approach for the discovery of residues that are important for mammalian DHP-sensitivity .
Candidate mutants were isolated and mapped through their resistance to nemadipine-induced population growth defect as described previously [27] , [28] . For this work , we used only the nemadipine-A analog , simply referred to as nemadipine here . To determine the IC50 and IC10 of the mutants with respect to the egg-laying-retention ( Egl ) phenotype and the variable abnormal ( Vab ) morphological defects induced by the compound , a population enriched for adults for each mutant strain was chunked onto MYOB plates containing either nemadipine or only the DMSO solvent as the negative control [28] . Approximately 24 hours later , young adults were picked on to a new nemadipine or DMSO plates . For the Egl assay , adults were scored as either being Egl or non-Egl approximately 24 hours later . For the Vab assay , the L1 and L2 larvae laid by the adults were scored for morphological defects approximately 24 hours later . To measure length and count embryos in utero , roughly 100 L4s of each strain were picked onto plates containing either DMSO or 5 µM nemadipine . Approximately 24 hours later , resulting young adults were photographed at 20× using a Leica MZFLIII microscope with a Retiga 1300 digital camera ( Q Imaging ) . The lengths were measured using Openlab software ( Improvision , Inc . ) . Embryos in utero were counted approximately 24 hours after the L4s were place on the plates using a Leica MZFLIII microscope at approximately 80× magnification . For the protruding spicule counts , him-5 ( e1490 ) doubles were created for each strain . L4 males were picked and put onto either DMSO or 5 µM plates . Approximately 24 hours later , the males were scored as either having protruding spicules ( Prc ) or non-Prc . The rat brain L-type α1C subunit in pMT2 and β2a and α2δ were generous gifts of Dr . Terry Snutch ( University of British Columbia , Vancouver , Canada ) . The α1C subunit was divided into 3 fragments based on endogenous restriction sites: KpnI in pMT2 , SalI in α1C , SpeI in α1C , and a second SpeI in the 3′ untranslated region of α1C . The first fragment of size 2 . 5 kB from KpnI ( pMT2 ) -SalI spanned the N-terminus up to domain II S6 of the channel . The last fragment of size 4 kB from SpeI-SpeI ( in the 3′ untranslated region ) spanned domain IIIS1 to the end of the C-terminus . Both of these fragments were separately subcloned into pBluescript SK+ ( Stratagene ) . Specific mutagenic primers ( Table S2 ) were designed and used in PCR-directed mutagenesis on the subcloned fragments using Stratagene's Site-directed Mutagenesis Kit . M1160L and M1161A were created using the GeneEditor Site Directed Mutagenesis Kit ( Promega ) with the following primers: M1160L ( 5′- CATTGCCTTCTTCCTGATGAACATC-3′ ) and M1161A ( 5′-CATTGCCTTCTTCATGGCCAACATCTTCGTGGG-3′ ) . Fragments were subcloned back into the parent α1C subunit in pMT2 , and the mutations were confirmed by DNA sequencing . Cell culture and transfection using tsA201 cells are presented in the SI materials and were done as previously described [39] , [40] . Whole-cell ( ruptured ) recordings were performed on an Axopatch 700A amplifier linked to a personal computer equipped with pClamp9 ( Axon Instruments , Foster City , CA ) . Patch pipettes ( Sutter , BF 150-86-15 ) were pulled using a Sutter P-87 microelectrode puller and polished with a Narashige microforge [39] , [40] . Pipettes ( 2–4 MΩ ) were filled with ( in mM ) 140 Cs-methanesulfonate , 4 MgCl2 , 9 EGTA , 9 HEPES ( pH 7 . 2 adjusted with CsOH ) . Bath recording solution comprised of ( in mM ) 20 BaCl2 , 85 CsCl , 40 TEA-Cl , 1 MgCl2 , 10 HEPES , 10 glucose ( pH 7 . 3 adjusted with CsOH ) . Perfusions consisted of bath solution with various concentrations of DHP from 0 . 01 µM to 100 µM or 10 µM CdCl2 , and applied using a gravity-driven system . In some cases , 0 . 1% DMSO was added to the control bath solution ( no drug ) . Data were filtered at 1 kHz ( −3 dB , 4-pole Bessel ) and digitized at 2 kHz . Currents were elicited by stepping from −100 mV to the indicated test potentials . Drug-response curve were determined by depolarizing to elicit maximal inward Ba2+ currents ( usually 0 to +10 mV ) every 5 seconds , first perfusing with control solution and then with various concentrations of DHP compounds until the current amplitude reached a steady-state level . Recordings were done with Ba2+as Ca2+ may greatly affect current through mechanisms such as Ca2+-dependant inactivation or Ca2+-dependant facilitation [42] . Channel behavior was otherwise normal in the mutants as Ba2+ current through the mutant channels were elicited at the range of applied voltages from −30 to +40 mV , which is similar to that of the wild-type channel . Dose-response curves were fitted to the equation:Where IBa is the current recorded either in the absence ( IBa ( control ) ) or the presence of drug; Fresidual is the fraction residual current at saturated concentrations of DHP; C is the concentration of DHP; IC50 is the DHP concentration required to elicit half the maximal inhibition; and n is the Hill coefficient ( see Table S3 for more information ) . As DHPs do not physically occlude the pore [43] Fresidual was included in the analyses so that the estimated IC50 accurately represents the interaction with the channel and not confounded by the efficacy of the block . Fits were performed in Origin 7 ( OriginLab Corp . , Northampton , MA ) . All experiments were performed at room temperature ( ∼22°C ) . Data are presented as mean±sem . Statistical analysis was done using one-way analysis of variance ( ANOVA , Holm-Sidak post-hoc ) . Differences were considered significant if p<0 . 05 . All chemicals used in the cell culture were purchased from GIBCO ( Invitrogen ) . Chemicals used for physiological recordings were purchased from Sigma . Nifedipine was purchased from Alamone Labs ( Jerusalem , Israel ) and from Sigma , and nemadipine was purchased from Chembridge Corp . ( Chembridge ID#: 5619779 ) . | L-type calcium channels are important drug targets because they regulate many physiological processes throughout the body . For example , L-type calcium channels regulate cardiac myocytes and vascular smooth muscle contraction . Antagonists are therefore commonly used to lower blood pressure and treat other related ailments . Despite their medical importance , the mechanism by which L-type antagonists inactivate calcium channels is not fully understood , due in large part to the lack of a channel crystal structure . Here , we present the first large-scale genetic screen for L-type calcium channel residues that are important for sensitivity to a new drug analog that we discovered called nemadipine . We performed the screen using nematodes , and then recreated similar mutations in a mammalian channel to investigate how the mutant residues alter interactions with the antagonists using electrophysiological techniques . Together , our analyses revealed eight new L-type calcium channel residues that are important for DHP-sensitivity and highlight the utility of using a simple animal model system for understanding how drugs interact with their targets . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"physiology/cardiovascular",
"physiology",
"and",
"circulation",
"non-clinical",
"medicine",
"pharmacology/drug",
"resistance",
"physiology/muscle",
"and",
"connective",
"tissue",
"physiology/neuronal",
"signaling",
"mechanisms",
"chemical",
"biology"
] | 2008 | A Genetic Screen for Dihydropyridine (DHP)-Resistant Worms Reveals New Residues Required for DHP-Blockage of Mammalian Calcium Channels |
Chromatin organization plays a major role in gene regulation and can affect the function and evolution of new transcriptional programs . However , it can be difficult to decipher the basis of changes in chromatin organization and their functional effect on gene expression . Here , we present a large-scale comparative genomic analysis of the relationship between chromatin organization and gene expression , by measuring mRNA abundance and nucleosome positions genome-wide in 12 Hemiascomycota yeast species . We found substantial conservation of global and functional chromatin organization in all species , including prominent nucleosome-free regions ( NFRs ) at gene promoters , and distinct chromatin architecture in growth and stress genes . Chromatin organization has also substantially diverged in both global quantitative features , such as spacing between adjacent nucleosomes , and in functional groups of genes . Expression levels , intrinsic anti-nucleosomal sequences , and trans-acting chromatin modifiers all play important , complementary , and evolvable roles in determining NFRs . We identify five mechanisms that couple chromatin organization to evolution of gene regulation and have contributed to the evolution of respiro-fermentation and other key systems , including ( 1 ) compensatory evolution of alternative modifiers associated with conserved chromatin organization , ( 2 ) a gradual transition from constitutive to trans-regulated NFRs , ( 3 ) a loss of intrinsic anti-nucleosomal sequences accompanying changes in chromatin organization and gene expression , ( 4 ) re-positioning of motifs from NFRs to nucleosome-occluded regions , and ( 5 ) the expanded use of NFRs by paralogous activator-repressor pairs . Our study sheds light on the molecular basis of chromatin organization , and on the role of chromatin organization in the evolution of gene regulation .
Regulatory differences affecting gene expression can play a major role in species evolution [1] and can help elucidate the functional mechanisms that control gene regulation [2] , [3] . Although specific examples of regulatory divergence are known in bacteria [4] , fungi [5] , , flies [9] , and mammals [10] , a general understanding of the evolution of gene regulation is still lacking . The recent availability of many sequenced genomes and accessibility of genomic profiling approaches open the way for comparisons of gene regulation across multiple species . Among eukaryotes , the Hemiascomycota yeasts ( Figure 1A ) , which span over ∼250 million years of evolution , are particularly suitable for studying evolution of gene regulation . This is due to the genetic tractability of yeasts , the wealth of knowledge about the model organism Saccharomyces cerevisiae , the large number of sequenced genomes , and the diversity of yeast lifestyles [3] . Notably , Hemiascomycota yeasts diverged before and after a whole genome duplication event ( WGD , Figure 1A ) [11] , which marked a shift from using respiration for energy production in pre-WGD species to primarily using fermentation in post-WGD species [12] . Nucleosomes modulate eukaryotic gene regulation by affecting the accessibility of other proteins to the DNA , which can impact gene activation and repression [13] . In particular , many genes have nucleosome-depleted “Nucleosome Free Regions” ( NFRs ) in their proximal promoters ( Figure 1B , top ) , providing access to sequence specific transcription factors ( TFs ) and to the basal transcription machinery [14] , [15] , [16] , [17] . Three major determinants have been proposed to impact nucleosome depletion at NFRs: ( 1 ) active transcription by RNA polymerase II results in eviction of the −1 nucleosome [18] , [19] , ( 2 ) intrinsic “anti-nucleosomal” DNA sequences such as Poly ( dA:dT ) bind histones with low affinity and can “program” NFRs constitutively [20] , [21] , [22] , [23] , [24] , and ( 3 ) trans-acting proteins can move nucleosomes away from their thermodynamically preferred locations [25] , [26] . Recent studies in yeast suggest a broad role for chromatin organization in regulatory evolution . Most regulatory divergence between closely related S . cerevisiae strains is associated with divergence in unlinked ( trans ) chromatin remodelers [27] , [28] . Conversely , many transcriptional differences between S . cerevisiae and S . paradoxus ( Last Common Ancestor ( LCA ) ∼2 million years ago ( MYA ) ) are due to linked cis polymorphisms predicted to affect nucleosome occupancy [29] , [30] . Furthermore , a recent study suggested that changes in the regulation of mitochondrial ribosomal protein ( mRP ) genes between the distant species C . albicans and S . cerevisiae ( LCA ∼200 MYA ) were associated with a change in nucleosome organization [31] , [32] . In particular , the higher expression of mitochondrial genes in respiratory C . albicans is accompanied by enrichment for the PolyA-like “RGE” binding site in the mRP gene promoters [31] , which appears to “program” the constitutive presence of wider , more open NFRs at these genes [32] . All of these are absent from the promoters of mRPs in the fermentative S . cerevisiae . Finally , a recent study [33] compared genome-wide nucleosome positioning in S . cerevisiae and S . pombe ( LCA ∼300M–1 BYa ) , finding changes in global nucleosome spacing and in the apparent sequences that intrinsically contribute to nucleosome positioning in vivo . While these examples are intriguing , they are limited in their phylogenetic coverage ( a pair of species ) or their functional scope ( one regulon ) . Thus , we understand little about the evolutionary interplay between gene expression , regulatory sequence elements , and chromatin organization . How does chromatin organization change over evolutionary time scales ? Are the mechanisms underlying chromatin packaging of functional gene modules conserved ? If not , how do they evolve and what is the role of different factors in this divergence ? Are changes in chromatin organization related to changes in gene regulation ? Can phylogenetic comparisons shed light on the distinct mechanisms that help establish chromatin organization ? Here , we present the first large-scale experimental and computational study of chromatin organization across a eukaryotic phylogeny . We measured genome-wide nucleosome locations and mRNA abundance in 12 Hemiascomycota yeast species , spanning over 250 million years of evolution ( Figure 1A ) . We developed an analysis framework that integrates the experimental data with genome sequences , functional gene sets , and TF binding sites across the 12 species . Our analysis uncovers several major principles that govern the evolutionary and functional relationship between chromatin organization and gene regulation in this phylogeny . ( 1 ) While qualitative features of chromatin organization are conserved in all species , quantitative features such as nucleosome packing , NFR length , and NFR to ATG distance have substantially diverged; ( 2 ) promoter chromatin organization and gene expression levels of “growth” and “stress” genes follow distinct patterns , and this dichotomy is conserved in all species; ( 3 ) evolutionary divergence in gene expression is often accompanied by transition of chromatin organization from a “growth” to a “stress” pattern; ( 4 ) changes in transcription levels , gain/loss of anti-nucleosomal sequences , and gain/loss of binding sites for “general regulatory factors” ( GRFs ) all play substantial and complementary roles in divergence of chromatin organization; ( 5 ) the loss of anti-nucleosomal sequences and parallel gain of binding sites for GRFs drive shifts from intrinsic to trans-regulated chromatin organization; ( 6 ) regulatory divergence can also occur by re-positioning of binding sites relative to nucleosome positions or by expanding the use of accessible sites by paralogous TFs . These mechanisms played a role in the evolution of respiro-fermentation , as well as in the evolution of regulation of other key regulons at different phylogenetic points , including mating , meiosis , RNA polymerase subunits , proteasomal , and splicing genes . Together , they uncover novel insights into the general roles for chromatin in regulating genomic access and in the evolution of regulatory programs , and provide a rich resource for future investigation .
We mapped nucleosome positions genome-wide in 12 Hemiascomycota species ( Figure 1A ) [34] by Illumina sequencing of mononucleosomal DNA [19] , [21] , [35] isolated from mid-log cultures ( Materials and Methods , Figures 1A and S1 ) . To minimize condition- and stress-related differences , we grew all species in the same rich medium , where the growth rate of each species was at least ∼80% of its maximal measured rate in any of over 40 tested media formulations . In order to compare our data to transcriptional output , we also used species-specific microarrays to measure mRNA abundance in all species in the same mid-log cultures used for nucleosome mapping ( Table S2 , Materials and Methods ) . Aligning nucleosome reads to each genome and averaging over all genes showed remarkably similar profiles in all species studied ( Figures 1A , S2 , S3 ) . All gene-averaged profiles are dominated by a pronounced depression upstream of the ATG that corresponds to the NFR [14] , [15] , [16] , [17] , [36] . To quantitatively compare chromatin structure between various genes , we first called nucleosome positions , identified 5′ and 3′ NFRs , and measured a number of nonredundant features that describe the chromatin organization at each gene ( Materials and Methods , Figures 1B and S4 ) . Below , we will study each feature at three levels: ( 1 ) globally , averaged across all genes in a genome; ( 2 ) functionally , averaged across all genes in a functional category; and ( 3 ) locally , at a single gene . Several qualitative chromatin features have previously been identified in all eukaryotes studied [14] , and these are conserved across all 12 species ( Figures 1A , S2 , and S3 ) . These include an abundant 5′NFR , a common 3′NFR , a well-positioned +1 nucleosome ( Nuc+1 ) , and increasing nucleosome fuzziness over the body of genes ( Figures S2 and S3 , Table S3 ) , which is consistent with statistical positioning of nucleosomes [23] , [37] , [38] . In contrast , quantitative global features were often variable between species ( Figures 1C–F and S5 , Table S3 ) . Our measurements recapitulated previous predictions or bulk assays in the few cases where these were available , thus validating our dataset and analytical methods . For example , nucleosome spacing in coding regions was variable between species ( Figure 1C , D ) , consistent with observed nucleosome laddering on gels [39] , [40] . This leads to variation in the specific coding sequences exposed in linker DNA and could affect patterns of sequence variation [41] , [42] , [43] and higher-order packaging into the 30 nm fiber [44] . The distance between the NFR and a gene's start codon ( Figures 1E , F and S5 ) is also variable between species , consistent with prior computational predictions [45] . Other evolutionary variations in global features were not previously described , showing that additional major aspects of chromatin architecture can substantially diverge . Most notably , the median NFR width was highly variable between species ( Table S3 ) , ranging from 109 to 155 nucleotides . This likely reflects the variation in the length and abundance of anti-nucleosomal Poly ( dA:dT ) tracts between species ( discussed below ) . Shorter NFRs may constrain regulatory information into more compact promoters . We next explored possible functional implications of chromatin organization in specific sets of genes with related function . Prior studies in S . cerevisiae and C . albicans have shown that in both species , “growth” genes , defined by their co-expression with cytoplasmic ribosomal proteins ( cRPs ) , have a more open chromatin organization on average [32] . Conversely , “stress” genes , whose expression is anti-correlated to that of growth genes , have a more closed chromatin organization in both species . To assess the generality of this observation , and identify additional trends , we tested in each species thousands of functional gene sets for enrichment of each of 22 distinct chromatin parameters . We used gene orthology [34] to project functional gene sets defined in S . cerevisiae across species ( Materials and Methods ) . For a given gene set in each species we calculated whether its constituent genes tended to have high or low values of each of the chromatin features ( Figure 1B ) , relative to the background of that feature's overall distribution in that species ( Kolmogorov-Smirnov ( K-S ) test , Figure 2A , B ) . This provides a comprehensive overview of chromatin organization at 5′ promoters and 3′ ends for each functional gene set across the 12 species ( Figure 2C–J , middle panels , Figures S6 and S7 , and Tables S4–S5 ) . In order to compare chromatin changes to gene expression levels , we also calculated the enrichment for high or low mRNA expression in all gene sets for each species ( K-S test , Figure 2C–J , left panels ) . We confirm a strong dichotomy in the promoter chromatin architecture of most “stress” and “growth” genes in S . cerevisiae [19] , [46] , [47] , [48] , [49] and C . albicans [32] and find that it is conserved across all 12 species ( Figures 2C , D and S6–S8 ) . Promoters of “growth” genes ( e . g . , ribosomal , proteasomal , and nuclear pore proteins , Figure 2C , E , G ) exhibit long and deep ( low occupancy ) 5′NFRs . Conversely , those of “stress” genes ( e . g . , toxin-response genes , integral membrane proteins , Figure 2D ) exhibit a more variable chromatin architecture , with shallower ( higher occupancy ) and narrower 5′NFRs . A host of other chromatin features also distinguish between the two functional groups ( Figure S6 ) . Thus , the separation of the “growth” and “stress” axes is a hallmark of Hemiascomycota gene regulation [2] , [3] and imposes strong constraints at all levels from evolution of gene content [34] to chromatin organization . There are , however , several exceptions to this rule . Most notably , several key “growth” genes , including glycolysis genes and endoplasmic reticulum genes , are highly expressed , yet do not exhibit deep NFRs in any species ( Figure 2F ) . We identify a range of additional conserved patterns of chromatin architecture associated with other specific functions , which were not previously reported . For example , a number of gene sets ( e . g . , reproduction , cell wall , inositol phosphate , benzoate , and nicotinamide metabolism genes ) have conserved long NFR to ATG distances ( Figure S6 ) , but have few other hallmarks of stress genes , and are expressed at average levels . In S . cerevisiae , these genes have long 5′ untranslated regions ( 5′UTRs ) [50] , suggesting that relatively long 5′UTRs are conserved at their orthologs in all 12 species . This may indicate a conserved role for translational control in the regulation of these functions [51] . On this backdrop of conservation , we find that coordinated changes have occurred in chromatin organization of specific functional gene sets , consistent with major phenotypic changes . Most notably , respiration and mitochondrial genes have switched from a “growth”-like chromatin pattern in pre-WGD species ( where they are highly expressed ) to a more “stress”-like pattern post-WGD ( Figures 2H and S6 ) . We confirm the previously reported change between S . cerevisiae and C . albicans for genes involved in respiratory metabolism [32] . We further extend these results across the full phylogenetic scope and to several other gene sets of related function ( Figures 2H and S6 ) . This change corresponds to a major change in lifestyle from respiration to respiro-fermentation after the WGD [12] , [31] , [32] , [52] . We also discover the converse evolutionary pattern ( Figure 2I ) : a number of gene sets involved in cytoskeletal organization are packaged into deeper NFRs in post-WGD species than in pre-WGD species . Surprisingly , the expression level of these genes has not substantially changed with this transition . Changes in chromatin organization have also occurred at other phylogenetic points of phenotypic evolution , suggesting a general evolutionary mechanism . For example , we discovered that in Yarrowia lipolytica spliceosome genes are associated with long and deep NFRs , but in all other species they are enriched for short and shallow NFRs ( Figure 2J , middle panel ) . This switch between deep and shallow NFRs is accompanied by a decrease in expression of these genes ( Figure 2J , left panel ) and is consistent with the much larger number of introns in Yarrowia lipolytica genes [53] and with the loss of introns and reduction of splicing in the subsequently diverged species . We next asked what mechanisms contribute to conservation and variation in chromatin organization across species . Three determinants have been previously implicated in establishing NFRs in S . cerevisiae [14]: ( 1 ) the expression level of the gene , as RNA polymerase recruitment affects NFR width; ( 2 ) the presence of intrinsic anti-nucleosomal sequences such as Poly ( dA:dT ) tracts in the gene's promoter; and ( 3 ) the binding of proteins such as chromatin remodelers that actively evict or move nucleosomes . We first consider these three determinants independently , and then assess their relative contributions . In some cases , variation in chromatin organization in a gene set , both within and between species , correlates with gene expression level . Within each species , many highly expressed “growth” genes ( e . g . , RP genes ) are packaged with wide and deep NFRs , while many poorly expressed stress genes have shorter , occupied NFRs ( Figures 2C , D , S6 ) . Between species , evolutionary shifts from high to low expression levels were sometimes accompanied by corresponding changes in chromatin organization ( e . g . , mitochondrial RP and splicing genes , Figure 2H , J ) . However , transcription level is insufficient to solely explain the NFR occupancy measured across the 12 species . Globally , expression level alone explains only 1 . 7%–13 . 1% of the variation in NFR occupancy in each of the 12 species ( Lowess fit , Figure S9A , C , E , Materials and Methods ) . Furthermore , when we use Lowess subtraction to correct for the relationship between mRNA level and each chromatin feature , the enrichments of most gene sets for high or low values of chromatin features were maintained ( Figure S10 , Materials and Methods ) . Within species , the discrepancy is prominent in some of the gene sets ( e . g . , glycolysis , gluconeogenesis ) that are highly expressed in all species but do not exhibit the expected deep NFRs ( Figure 2F ) . Between species , cytoskeleton and nuclease-related gene sets have shifted from shallow to deep NFRs at the WGD , often without a concomittant change in expression levels ( Figure 2I ) . The failure of transcript levels to fully explain NFR width and depth is consistent with recent experimental results in S . cerevisiae , where the distinctive chromatin organization of growth and stress genes was largely maintained even after genetically inactivating RNA Pol II [19] . We next tested an alternative hypothesis that chromatin organization at the NFR is determined by intrinsic “anti-nucleosomal” sequences with low affinity for the histone octamer , such as Poly ( dA:dT ) tracts [20] , [21] , [22] , [24] , [54] , [55] . We estimated the average extent of nucleosome depletion over a variety of Poly ( dA:dT ) elements ( Materials and Methods ) for each species ( Figures S11 , S12 ) . We then tested if functional gene sets in each species were enriched or depleted for strongly anti-nucleosomal sequences in their NFRs . Finally , we compared this pattern to their chromatin organization ( Figure 2C–J , right versus middle panels ) . In some cases , the variation in chromatin organization within and between species is associated with variation in intrinsic “anti-nucleosomal” Poly ( dA:dT ) tracts . Within each species , Poly ( dA:dT ) sequences are enriched upstream of many highly expressed , nucleosome-depleted , “growth” gene sets , consistent with previous observations in S . cerevisiae [48] , [49] . Between species , we found that gain and loss of polyA sequences is associated with changes in chromatin organization at several gene sets and phylogenetic points , suggesting that this is a common evolutionary mechanism used more than once in this phylogeny . We confirmed a prior observation [32] that the change in chromatin organization at mitochondrial ribosomal protein ( mRP ) genes in post-WGD respiro-fermentative species is accompanied by the loss of PolyA-like sequences from these promoters ( Figure 2H ) . In addition , we found that the deeper and wider NFRs at splicing genes in Y . lipolytica are associated with a greater length and number of PolyA sequences at these genes ( Figure 2J ) . Conversely , the relatively shallow NFRs of gluconeogenesis genes observed in S . castellii are associated with concomitant depletion of polyA sequences in this species ( Figure 2F ) . Nevertheless , intrinsic anti-nucleosomal sequences explain only 8 . 6%–25 . 7% of the global variation in NFR occupancy within a given species ( Figure S9 ) . Even when combining expression levels and sequence information together , these can only explain 13%–29% of the global variation in nucleosome organization in the 12 species ( Figure S9E ) . Similar results are obtained when considering other measures of intrinsic anti-nucleosomal sequences , such as those based on computational models [21] , [48] derived from in vitro data ( unpublished analysis ) . Thus , anti-nucleosomal sequences and expression patterns are insufficient to fully explain either conservation or divergence in chromatin organization across species . For example , proteasomal genes are highly expressed and have deep NFRs conserved in all species , but are not associated with intrinsic anti-nucleosomal sequences ( Figure 2E ) . Furthermore , RNA Polymerase II subunits , RNA export , and nuclear pore genes are highly expressed with deep NFRs conserved in most species , but are enriched for intrinsically anti-nucleosomal sequences in only a subset of species ( Figure 2G , see below ) . Conversely , peroxisome genes are highly expressed in D . hansenii , C . albicans , and Y . lipolytica , where they are packaged with long ( but not deep ) NFRs , despite no enrichment for Poly ( dA:dT ) tracts ( see below ) . In these and other cases , even when we consider expression levels , much of the depletion in NFRs remained unexplained ( Figures S9 , S10 ) . We therefore wished to explore the role that the third mechanism—nucleosome eviction by chromatin remodelers—plays across the 12 species . We hypothesized that changes in chromatin remodeling would be accompanied by variation in the cis-regulatory elements bound by GRFs that likely recruit chromatin remodelers [25] , [56] , [57] . Unlike intrinsic anti-nucleosomal sequences that establish constitutively programmed NFRs , binding sites for GRFs likely establish regulated NFRs that can change based on trans inputs . We first assessed the potential contribution of chromatin remodelers to chromatin organization based on the presence in NFRs of the known binding sites for the two best-studied S . cerevisiae GRFs: Abf1 and Reb1 ( Figure S9E , Materials and Methods ) . Together , the two motifs explain 1 . 2%–15 . 1% of the observed variation in nucleosome organization in the 12 species . Furthermore , Abf1 and Reb1 can explain up to 12 . 6% of the residual variation after accounting for the contribution of expression levels and intrinsic sequences ( Successive Lowess , ) . Thus , GRFs can play an important role in explaining global chromatin organization . Notably , the Abf1 and Reb1 sites explain little of the variation in D . hansenii , C . albicans , and Y . lipolytica—the species from the two clades most distant from S . cerevisiae . In particular , the Abf1 binding site explains less than 1% of the variation in each of these species , consistent with the absence of the Abf1 ortholog from their genome , and validating the specificity of our approach . Furthermore , although the Reb1 ortholog is present in each of these species , its contribution is substantially reduced ( compared to , e . g . , S . kluyveri ) . This loss of predictive power by Abf1 and Reb1 sites at increasing phylogenetic distance led us to hypothesize that other GRFs , with distinct binding specificity , are active in these species . To identify novel GRF cis-elements , we therefore searched for short sequence elements that are depleted of nucleosomes in vivo but not in vitro [21] . We calculated the extent of nucleosome depletion over every 6- and 7-mer sequence in each of our species ( Table S6 , Materials and Methods ) and identified those sequences whose depletion score in vivo in at least one species is significantly greater than expected from published in vitro data ( Figure 3A , Figure S13–S14 ) [21] . This procedure automatically identified in vivo-specific depletion over 7-mers consistent with the binding sites for known S . cerevisiae GRFs such as Reb1 ( Figure 3A , orange ) [58] , [59] and the Rsc3/30 components of the RSC ATP-dependent chromatin remodeling complex ( Figure 3A , green ) [25] , [58] , [59] , validating our approach . Consistent with our hypothesis , it also revealed a number of sequence motifs that were specifically nucleosome-depleted in vivo in some species but not in S . cerevisiae , such as the CACGTG motif that serves as the binding site for Cbf1 in S . cerevisiae and C . albicans ( Figure 3A , blue ) [8] , [58] , [59] , [60] , [61] . We therefore propose that these sites are candidates for putative GRF function in these species . When we compared the GRF sequences between species we discovered extensive divergence that largely conforms to phylogenetic distance ( Figures S13 , S14 ) . The extent of nucleosome depletion over short sequence elements is well conserved between closely related species , such as S . cerevisiae and S . mikatae ( ∼2–5 MYA , Figure S13B ) . In contrast , there are much more dramatic differences in the in vivo depleted sites ( e . g . , Rsc3/30 , Cbf1 ) between the more distant S . cerevisiae and K . lactis ( ∼150 MYA , Figures S13C , S14 ) . Finally , there are also gradual changes in the specific Rsc3/30 CGCG-containing motifs that were nucleosome-depleted in each species ( Table S6 ) , consistent with co-evolution of a GRF and its binding site , as previously observed for TFs [3] , [5] . The use of different GRF sites often follows strong phylogenetic patterns , allowing us to trace transitions from the dominant use of one repertoire of GRFs to that of another , and suggesting compensatory evolution of GRF use . Most notably , we find a major and gradual transition from the use of Cbf1 as a major GRF in pre-WGD species to the use of Reb1 as a GRF in post-WGD species ( Figure 3B ) . The Cbf1 binding sequence CACGTG is nucleosome-depleted in vivo in most pre-WGD species ( except Y . lipolytica and C . albicans ) but not in post-WGD species ( except C . glabrata ) ( Figure 3B ) . Conversely , Reb1 sites are nucleosome depleted in all post-WGD species but not in most pre-WGD species ( except K . lactis ) ( Figure 3B ) . This complementary phylogenetic pattern suggests an evolutionary scenario where Cbf1 was a major ancestral GRF , Reb1 emerged as a GRF before the WGD , and gradually “took over” Cbf1's global functionality . Similar evolutionary patterns were previously observed for TFs [3] , [7] , [61] , [62] , and this is the first demonstration to our knowledge of such a “mediated replacement” for GRFs . Evolutionary transitions in GRF usage are sometimes limited to one or a few species . For example , we found a set of novel motifs that were nucleosome-depleted only in Y . lipolytica ( Table S6 ) , the earliest diverging species in our panel . Finally , we observe changes in the relative balance between nucleosome depletion via GRFs and constitutively programmed depletion via Poly ( dA:dT ) sequences , suggesting a global mode of compensatory evolution . Most notably , A7/T7 is less nucleosome-depleted at D . hansenii promoters than at promoters of any other species , whereas Cbf1-like and Rsc3/30-like sites are strongly nucleosome-depleted in D . hansenii ( Figure S14 ) . This transition is likely due to the shorter lengths of Poly ( dA:dT ) stretches in D . hansenii ( Figure S11C , Table S7 ) , a sequence change that may be an adaptation to the high salt concentrations in this species' ecological niche ( secondary to increased DNA flexibility in high salt ) . As noted above , D . hansenii has a very short average NFR width ( Table S3 , Figure S11D ) , consistent with diminished nucleosome repulsion at its shorter Poly ( dA:dT ) sequences . We hypothesize that the expansion in use of the Cbf1 and Rsc3/30 GRFs is a mode of compensatory evolution needed to adapt to a change in genome sequence in a unique niche; it also suggests that D . hansenii NFRs may be more responsive to environmental signals . We next hypothesized that the identified GRFs are important for the observed chromatin organization in functional gene sets across species . To test this hypothesis , we assessed the enrichments of GRF motifs in the NFRs of each gene set across the 12 species ( Table S8 ) . In some cases , GRF motifs ( but not Poly ( dA:dT ) tracts ) were enriched in a gene set across multiple species , strongly indicating a conserved regulatory mechanism . For example , the Abf1 site is enriched in RNA polymerase genes across the clade spanning S . cerevisiae and S . kluyverii ( Figure S15D ) . However , since the spectrum of GRFs is species-specific ( Figures 3B , S14 ) , we found no gene set associated with the same GRF site across the entire phylogeny . Instead , we found a number of cases where a single gene set has a conserved chromatin architecture but is associated with distinct GRF sites in different species , consistent with changes in the global GRF repertoire . This is most notable in proteasome genes , which are uniformly associated with wide/deep NFRs but are depleted of Poly ( dA:dT ) tracts ( Figure 2E ) . The establishment of NFRs at these genes has likely transitioned from a mechanism dependent on the CACGAC sequence in the Candida clade to an Abf1-dependent mechanism in later lineages , with additional contribution from Reb1 and Rsc3/30 sites , as these GRFs gained dominance in specific species and clades ( Figures 3C and S15E ) . Although the specific GRF mechanism underlying NFRs in proteasome genes has diverged , the establishment of wide/deep NFRs by a GRF-regulated mechanism ( rather than polyA/constitutive mechanism ) is conserved in all species . We hypothesize that GRF-regulated NFRs at proteasome genes may be related to the unusual transcriptional regulation of proteasome genes: these are among the few highly expressed “growth” genes ( with open accessible promoters ) that are further upregulated ( rather than downregulated ) during stress responses [63] . Could promoters evolve from having constitutively programmed NFRs to regulated ones ? To test this , we searched for gene sets where chromatin organization is conserved , while the underlying anti-nucleosomal sequences have diverged in a phylogenetically coherent pattern . We found that genes encoding RNA polymerase subunits exhibit deep NFRs across most of the phylogeny ( Figure S15D ) . These genes' promoters are associated with Poly ( dA:dT ) tracts in Y . lipolytica and the species of the Candida clade , with both Poly ( dA:dT ) and the site for the Abf1 GRF in species from S . kluyveryi to S . bayanus , and only with Abf1 in the clade spanning S . mikatae , S . paradoxus , and S . cerevisiae ( Figures 3D and S15D ) . Similar behavior is seen at a number of other gene sets , such as those encoding nuclear pore components ( unpublished analysis ) . This profile suggests an evolutionary scenario where the ancestral mechanism relied on Poly ( dA:dT ) . With the emergence of Abf1 in the LCA of the pre- and post-WGD species [34] , it gained additional control of the NFRs in this gene set , alongside Poly ( dA:dT ) tracts . Then , after the divergence of S . bayanus , Poly ( dA:dT ) tracts were lost from the genes' promoters , leading to a complete switch from a constitutively programmed to a regulated NFRs . This compensatory evolution is consistent with patterns observed for TF binding sites in functional regulons [3] , [62] and with the global transitions in GRFs described above . In some cases , the gain or loss of binding sites for GRFs can contribute to divergence in chromatin organization , coupled to phenotypic changes . Most notably , peroxisomal genes are associated with wider NFRs in Y . lipolytica , C . albicans , and D . hansenii , and shorter NFRs in subsequently divergent species ( Figures 3E and S15F ) , but are not associated with intrinsic anti-nucleosomal poly ( dA:dT ) tracts in any of the 12 species . Instead , we find that these genes' promoters are enriched for PolyG and Rsc3/30-like sites in Y . lipolytica , C . albicans , and D . hansenii , but not in other species . This suggests an evolutionary scenario where either a Rsc-like motif or PolyG-based nucleosome depletion was the ancestral mechanism controlling peroxisomal genes , and was subsequently lost in the LCA of the clade spanning S . kluyverii and S . cerevisiae . This scenario is consistent with the higher expression of peroxisomal genes in Y . lipolytica ( where peroxisomes are particularly central for carbon metabolism ) and C . albicans ( where peroxisomes play a key role in virulence ) . Even when NFR positions and their underlying mechanisms are largely conserved , they can play an important role in regulatory divergence . Nucleosomes are generally inhibitory to TF binding [13] , and in S . cerevisiae most functional TF binding motifs are found in NFRs [23] . Precise positioning of TF binding sites relative to nucleosomes has regulatory consequences such as changing signaling thresholds [64] or logic gating [65] . We therefore hypothesized that an evolutionary change in the location of TF-binding motifs relative to the nucleosomes in a gene's promoter can lead to regulatory divergence between species . To test this hypothesis , we examined the location of known TF binding motifs ( from S . cerevisiae; [58] , [59] , [60] , [66] ) relative to nucleosome positions in each of the 12 species ( Materials and Methods ) . Consistent with our expectations , in S . cerevisiae ( Figure 4A , B ) , up to 90% of the binding sites for growth-related TFs are localized to NFRs ( e . g . , REB1 , ABF1 , RAP1 , and FHL1 ) , whereas as few as 25% of sites for stress-related TFs are at NFRs ( e . g . , HSF1 , YAP6 , HAP2/3/5 , GZF3 , and CRZ1 ) . Thus , sequences that are mostly occluded by nucleosomes tend to be the binding sites for inactive TFs , and we can use chromatin information to infer TF activity under our growth conditions in each species . We therefore calculated for each motif the fraction of its instances located in NFRs in each of the 12 species ( Figures 4C and S16 ) . The NFR positioning of many key motifs is strongly conserved . For example , sites for growth-related factors such as SWI4/6 and GCN4 were similarly NFR-exposed in all species in this phylogeny . Notably , this conservation is observed despite the fact that many motifs , which were experimentally defined for S . cerevisiae proteins , were globally less NFR-localized in distantly related species ( Figure 4C , Figure S16B ) . This can be attributed in some cases to divergence of binding site preferences of the cognate TFs , and in other cases to the absence of the TF's ortholog from the genome ( Figure 4C , white ) . Nevertheless , many motifs showed robust conserved positioning in NFRs . Conversely , the motifs for key TFs associated with regulation of respiration and carbohydrate metabolism have repositioned relative to NFRs at the WGD , consistent with regulatory divergence in these functions ( Figure 4D ) . For example , the sites for the HAP2/3/4/5 complex ( a regulator of respiration genes ) and for YAP6 ( a regulator of oxidative functions ) have re-positioned from NFRs to nucleosome-occluded positions post-WGD , consistent with the reduction in expression of respirative genes . In contrast , the sites for the carbon catabolite repressor MIG2 and for the glucose-responsive TF RGT1 have repositioned from nucleosomes into NFRs in post-WGD species , consistent with these factors' role in establishing a fermentative strategy through gene repression . Motif re-positioning has also occurred at other phylogenetic points and gene sets , suggesting that this is a general regulatory and evolutionary mechanism ( Figure 4E , F ) . For example , the mating-related STE12 motif is significantly enriched upstream of reproduction and mating-related genes in species from S . cerevisiae to S . kluyverii , including C . glabrata . Although STE12 sites are found in NFRs at mating genes for most of these species , they are largely nucleosome-occluded in C . glabrata ( Figure 4E ) , an organism which has never been observed to mate [67] . We speculate that occlusion of STE12 sites under nucleosomes may contribute to this species' reluctance to mate , but the continued enrichment of STE12 upstream of mating genes and the retention of many meiosis-related genes [34] in C . glabrata suggests that it may still be capable of mating under special conditions . We therefore predict that conditions ( environmental or perhaps genetic ) that either mobilize or destabilize the nucleosomes covering STE12 sites at pheromone-response genes might enable mating in this species . Similarly , motifs for UME6 , a major regulator of meiosis genes in S . cerevisiae [68] , are globally NFR-positioned in all species except C . glabrata ( Figure 4F ) , despite the fact that UME6 sites are enriched upstream of orthologs of meiosis-related genes in C . glabrata . Thus , the relative re-positioning of NFRs and TF binding sites may help explain the molecular underpinnings of dramatic changes in regulatory and phenotypic evolution . Finally , we asked whether chromatin information could be used to infer the regulatory effect of exposed TF binding sites from the expression level of their target genes . We expect exposed TF binding sites to have different regulatory consequences depending on whether or not the TF is active and whether it acts as an activator or a repressor . We reasoned that an NFR-positioned site for an active positive regulator will be associated with a higher expression of the target genes . Conversely , an NFR-positioned site for an active negative regulator will be associated with a lower expression of the target genes . We therefore compared the expression level of all genes where a given TF motif was located within nucleosomes versus those in which the motif was located within promoter linkers ( largely the NFR , Figure 5A ) . Consistent with our expectation , in S . cerevisiae , transcriptional activators known to be active in mid-log phase , such as RPN4 or PBF1 , were associated with higher expression levels at genes carrying an accessible , linker-positioned motif . In contrast , NFR-positioned motifs for transcriptional repressors known to be active in mid-log ( e . g . , MIG1 , SUM1 , NRG1 , DIG1 , STB1/2 , or RIM101; Figure 5A ) were associated with lower downstream gene expression . Thus , we devised a novel approach to predict whether a given motif is associated with an activator or repressor in vivo in the growth condition tested . When we extended this analysis to all 12 species ( Figure S17 ) , we found substantial divergence in the regulatory logic of the same NFR-positioned motif , most notably at the WGD ( Figure 5B ) . We found a host of motifs which , when present in NFRs , were associated with differences in RNA expression levels between pre- and post-WGD species . Many of those ( ∼100 ) appeared to shift from activator-like behavior in pre-WGD species ( higher target expression when in NFR ) to repressor-like behavior in post-WGD species ( lower target expression when in NFR ) . These included sites for a surprisingly large number of TFs involved in repression of metabolic genes in S . cerevisiae , including MIG1 , GIS1 , RGT1 , and GAL80 . Interestingly , several of these genes are found in a single copy in pre-WGD species but were retained as duplicates [34] with similar DNA-binding specificity following the WGD ( e . g . , GIS1/RPH1 , RGT1/EDS1; Figure 5B , C ) . This suggests that widespread usage of competing activator/repressor pairs in S . cerevisiae may have been facilitated by the generation of such TF pairs at the WGD . Such duplication of trans-factors can serve as an alternative evolutionary mode to expand and evolve regulatory capacity [69] even when NFRs and motif positioning may be conserved .
What establishes the nucleosomal organization of a genome ? While it has been argued that intrinsic DNA sequence can almost fully explain nucleosome organization [21] , recent analysis of in vitro reconstitution data showed that the major intrinsic contributor to nucleosome positioning in budding yeast is the anti-nucleosomal behavior of Poly ( dA:dT ) and related sequences [21] , [24] , [70] . Conversely , recent reports indicate that in S . pombe Poly ( dA:dT ) plays only a minor role in nucleosome exclusion in vivo [33] , indicating that even the best-understood sequence contributor to chromatin organization plays variable roles in chromatin structure in different species . Our analysis provides several lines of evidence that expression levels , intrinsic anti-nucleosomal sequences , and binding sites for GRFs that may recruit chromatin modifiers all play a role in establishing promoter chromatin architecture , and that the balance between these three contributors changes in evolution and between functional groups of genes . ( 1 ) We show that a sequence-based model based on in vitro depletion alone [21] can only account for 8 . 6%–25 . 7% of variance in NFR depth within any of the 12 species , including S . cerevisiae ( 10 . 6% ) . Similarly , expression levels alone can only account for 1 . 7%–13 . 1% of the variation in each species . Even when combining both the expression and intrinsic models we can only explain 13%–29% of the variation within any single species . ( 2 ) Although changes in intrinsic sequences and expression levels can explain changes in chromatin across species for some gene sets ( e . g . , mRPs or splicing genes; Figure 6A and B ) , they are insufficient to explain conserved chromatin behavior across the phylogeny ( e . g . , RNA Polymerase subunit genes; Figure 6D ) , nor do they explain changes in chromatin organization across species in other groups of genes ( e . g . , peroxisome genes; Figure 3E ) . Thus , these two determinants ( alone or in combination ) are insufficient to explain both intra- and inter-species variation . ( 3 ) In contrast , by comparing our in vivo data in each species to two in vitro datasets [21] , [32] , we find in each species a host of sequences that exhibit significantly greater nucleosome depletion in vivo than in vitro . Many of these correspond to binding sites for known GRFs that play an active role in nucleosome eviction in S . cerevisiae [14] , [25] , [56] , [58] , whereas others represent novel candidate GRF sequences ( Figures 3 and 6C ) . ( 4 ) The relative contribution to nucleosome organization from GRFs , intrinsic sequences , and expression levels varies between different gene sets ( in all species ) . For example , we show that intrinsic anti-nucleosomal sequences are enriched at NFRs in cytoplasmic RPs ( in all species; Figure 2C ) , whereas GRFs fulfill this role in proteasome genes ( in all species; Figure S15 ) . ( 5 ) We also show that the relative contribution of one mechanism versus another can change in evolution ( across species ) , both globally ( as in the halophile D . hansenii , that relies more on GRFs ) and in specific gene sets ( as in the RNA polymerase gene set that shifted from intrinsic to regulated NFRs; Figure 6D ) . ( 6 ) Globally , even when we consider only the binding sites for the two best-characterized GRFs from S . cerevisiae ( Abf1 and Reb1 ) , GRFs alone can explain 5 . 2%–15 . 1% of the variation in nucleosome organization ( in species where their orthologs are present ) , and 3 . 7%–12 . 6% of the residual variation after considering the contribution from expression and Poly ( dA:dT ) . Taken together , this analysis points to a complex interplay between the different factors that control nucleosome positions , allows us to assess their contributions , and recognizes the plastic and evolvable nature of all the determinants . Our study also discovers an intricate and intimate relationship between conservation and divergence of chromatin organization and evolution of gene regulation . At one extreme , we found a broad functional dichotomy in chromatin organization between “growth” and “stress” genes , which is largely conserved . At the other extreme , we found that chromatin organization has diverged at a major evolutionary scale , as has happened during the evolution of respiro-fermentation , and at other points of phylogenetic and phenotypic divergence . We found five major mechanisms by which chromatin organization can be associated with divergence of gene expression . Each of these was “used” more than once in the phylogeny , and is associated with more than one phenotypic or regulatory change , including the changes described in carbon metabolism , mating , meiosis , and splicing genes . These include ( 1 ) gain or loss of intrinsic ( PolyA ) sequences can open or close NFRs ( Figure 6A , B ) [32]; ( 2 ) conserved NFRs can be controlled by different GRF determinants , through compensatory evolution ( Figure 3C ) ; ( 3 ) NFRs can shift between constitutive and regulated determinants by compensatory ( “balanced” ) gain/loss of intrinsic anti-nucleosomal sequences and GRF binding sites ( Figure 6D ) ; ( 4 ) motifs can re-position relative to NFRs to change transcriptional output ( Figure 6E–G ) ; and ( 5 ) duplication and divergence of trans-factors can expand the regulatory behavior of conserved NFRs and binding sites ( Figure 6H ) . The evolution of the respiro-fermentative lifestyle following the WGD required a major reprogramming of the yeast transcriptional network and involved all of the mechanisms we describe . The shift thus included loss of intrinsic Poly ( dA:dT ) anti-nucleosomal sequences in some functional modules ( e . g . , mitochondrial RP genes ) , and the loss or switch of putative GRF sequences in others ( e . g . , oxidation-reduction genes ) . Furthermore , sites for certain respiratory TFs ( e . g . , HAP2/3/5 , YAP1/3/6 ) have re-positioned out of NFRs , and those for glucose repression TFs have re-positioned into NFRs ( e . g . , RGT1 , MIG1 ) . In yet other cases , the WGD has resulted in the retention of paralogous activator-repressor pairs that control several modules in carbohydrate metabolism . Notably , each of these mechanisms has acted also at other phylogenetic points , suggesting that they point to general principles , and emphasizing the utility of the WGD as a model to study regulatory evolution . Our work provides a general framework for the study of chromatin organization , function , and evolution . This includes a comprehensive genomics resource ( http://www . broadinstitute . org/regev/evolfungi/ ) and a host of analytical approaches with broad applicability . Future studies can use our resource and methods to decipher more detailed models of the relationship between sequence elements , trans-factors , and gene expression , as well as on the evolution of regulatory systems . Finally , our comprehensive study in the emerging field of comparative functional genomics demonstrates how to combine the power of functional assays with extensive phylogenetic scope , to shed light both on mechanistic and evolutionary principles .
We used the following strains in the study: Saccharomyces cerevisiae , BY4741 , Saccharomyces cerevisiae , Sigma1278b L5366 , Saccharomyces paradoxus , NRRL Y-17217 , Saccharomyces mikatae , IFO1815 , Saccharomyces bayanus , NRRL Y-11845 , Candida glabrata , CLIB 138 , Saccharomyces castellii , NRRL Y-12630 , Kluyveromyces lactis , CLIB 209 , Kluyveromyces waltii , NCYC 2644 , Saccharomyces kluyveryii , NRRL 12651 , Debaryomyces hansenii , NCYC 2572 , Candida albicans , SC 5314 , Yarrowia lipolytica , CLIB 89 . All cultures were grown in the following medium: Yeast extract ( 1 . 5% ) , Peptone ( 1% ) , Dextrose ( 2% ) , SC Amino Acid mix ( Sunrise Science ) 2 grams per liter , Adenine 100 mg/L , Tryptophan 100 mg/L , and Uracil 100 mg/L . This in-house recipe was designed to mitigate differences in growth rates between species . Overnight cultures for each species were grown in 450 ml of media at 220 RPM in a New Brunswick Scientific air-shaker at 30°C until reaching mid log-phase ( OD600 = 0 . 5 , WPA biowave CO 8000 Density Meter ) . Before formaldehyde fixation , 50 ml of the culture were transferred to a 50 ml conical and spun down immediately . The isolated cell pellets were then placed in liquid nitrogen , stored at −80°C , and were later archived in RNA later for future RNA extraction . Nucleosomal DNA isolation was carried out as previously described [23] with the following slight modifications . For different species , cells were spheroplasted with zymolase between 30 and 40 min , depending on how much time was necessary to fully remove each species' cell wall . MNase digestion levels for all samples were uniformly chosen across species to contain a slightly visible tri-nucleosome band ( Figure S1 ) . Mononucleosomes were size-selected on a gel and purified using BioRad Freeze-N-Squeeze tubes followed by phenol-chloroform extraction . Selected DNA was prepared for sequencing using the standard Illumina protocol that includes blunt ending , adaptor ligation , PCR amplification , and final size selection plus gel purification [35] . Libraries were sequenced on an Illumina 1G Analyzer , to generate 36 bp reads . Total RNA was isolated using the RNeasy Midi or Mini Kits ( Qiagen ) according to the provided instructions for mechanical lysis . Samples were quality controlled with the RNA 6000 Nano ll kit for the Bioanalyzer 2100 ( Agilent ) . Genomic DNA was isolated using Genomic-tip 500/G ( Qiagen ) using the provided protocol for yeast . DNA samples were sheared using Covaris sonicator to 500–1000 bp fragments , as verified using DNA 7500 and DNA 12000 kit for the Bioanalyzer 2100 ( Agilent ) . Independently sheared samples labeled with Cy3 and Cy5 were highly correlated ( R> . 97 in each of 4 independent hybridizations ) , indicating that the shearing procedure is reproducible and unbiased . Total RNA samples were labeled with Cy3 ( cyanine fluorescent dyes ) and genomic DNA samples were labeled with Cy5 using a modification of the protocol developed by Joe Derisi ( UCSF ) and Rosetta Inpharmatics ( Kirkland , WA ) that can be obtained at www . microarrays . org . Between three and four biological replicates of Cy3-labeled RNA samples were mixed with a reference Cy5 labeled genomic DNA sample and hybridized on two-color Agilent 55- or 60-mer oligo-arrays . We used the 4×44 K format for the S . cerevisiae strains ( commercial array; 4–5 probes per target gene ) or a custom 8×15 K format for all other species ( 2 probes per target gene , designed using eArray software , Agilent ) . After hybridization and washing per Agilent's instructions , arrays were scanned using an Agilent scanner and analyzed with Agilent's feature extraction software version 10 . 5 . 1 . 1 . For each probe , the median signal intensities were background subtracted for both channels and combined by taking the log2 of the Cy3 to Cy5 ratio . To estimate the absolute expression values for each gene , we took the median of the log2 ratios across all probes . The experiments were highly reproducible; most biological replicates correlated at R = 0 . 99 and replicates with R<0 . 95 were removed . Different biological replicates were combined using quantile normalization to estimate the absolute expression level per gene per species . We used BLAT [71] to map sequenced reads from each experiment to the corresponding reference genome , keeping only reads that mapped to a unique location and allowing for up to 4 mismatches . Each uniquely mapped read was then extended to a length of 100 bp . To generate a genomic nucleosome occupancy landscape , we summed all extended reads covering each base pair . We then masked all repetitive regions along each track , defining repetitive regions as locations in the genome that cannot be uniquely defined by the length of a read ( 36 bp ) . We also masked all regions of nucleosome occupancy greater than 10 times the median occupancy , to remove outlier effects that occur in places such as the rDNA locus . To normalize for sequencing depth for each genomic nucleosome track , we divided the occupancy at each location by the mean nucleosome occupancy per base pair . These normalized maps were used to generate the average nucleosome occupancy plots ( Figures 1 , 2A , and S2–S3 ) . To infer the location of nucleosomes from the data , we used a Parzen window approach similar to that previously described [35] , [46] . Our modified approach uses three parameters—the average DNA fragment length , the standard deviation of the Parzen window , and the maximum allowable overlap between nucleosomes . To estimate the mean DNA fragment length in each experiment , we shifted reads from one strand and then correlated them with the reads of the opposite strand . For each species , we observed a peak in the cross-correlation at a shift between 127 and 153 bp , which we used to estimate the mean DNA fragment length per experiment . We chose a standard deviation of the Parzen window of 30 bp for all species , since it closely matched the observed standard deviation around the cross-correlation peak of each experiment . Finally , we set the maximum allowable overlap between nucleosomes to 20 bp . We then shifted all read start locations by half of the mean DNA fragment length in the direction towards the dyad of the nucleosome they represent . Our approach places a normal distribution with a standard deviation of 30 bp at each read's shifted location . Summing all individual curves for all loci leads to a smoothed probability landscape of nucleosome occupancy . We next identify all peaks along the landscape , which represent nucleosome centers . The algorithm then places nucleosomes along the genome in the order of decreasing peak heights ( greedy approach ) and iteratively masks out these regions to prevent more than 20 bp overlap between nucleosomes . We define 5′ and 3′ NFRs as the linker DNA of “significant length” closest to the 5′ and 3′ end of each gene , respectively . To find NFRs , we first created a nucleosome call landscape for each genome , normalized for sequencing depth in the same manner as the nucleosome occupancy maps ( above ) . NFR boundaries were often obscured by very low occupancy nucleosome calls . We therefore removed all nucleosome calls with occupancy less than 40% of the average nucleosome occupancy from the map . We searched for 5′ or 3′ NFRs within 1 , 000 bases upstream/downstream of the 5′ or 3′ end of each gene , truncated when neighboring ORFs overlapped this region . We then defined an NFR as the linker DNA longer than 60 bp closest to the 5′ or 3′ end of each gene . If no linker longer than 60 bp was found in this search , we defined the NFR as the first linker from the 5′ or 3′ end . Our method was highly predictive of transcription start sites ( TSSs ) in S . cerevisiae [50]—the NFR boundary closest to the 5′ end of the gene was able to predict 84% of TSSs within 50 bp . Linker lengths of 50 bp or 70 bp and occupancy thresholds of 30% or 50% produced highly similar results ( unpublished data ) . Since 5′NFR-ATG distances vary substantially between species , an analysis of nucleosome organization that relies on alignment by ATG can be misleading . For example , the average nucleosome organization of C . glabrata and S . castellii look similar when aligned by the +1 nucleosomes ( Nuc+1 ) but very different when aligned by ATG ( unpublished data ) . A previous study [32] defines a promoter nucleosome depleted region ( PNDR ) score as mean nucleosome occupancy of the most depleted 100 bp region within 200 bp upstream of the ATG . Since some species have longer 5′NFR-ATG distances we reasoned that the NFR of some genes may not be contained within a 200 bp window ( e . g . , only a third of C . glabrata NFRs are contained within 200 bp , while 90% are contained within 500 bp ) . To avoid such pitfalls and analyze nucleosome organization consistently in all species we aligned the data by Nuc+1 , which is consistent with alignment by TSS . For S . cerevisiae we used functional gene sets from several sources: KEGG [72] , GO categories [73] , MIPS [74] , and BioCyc [75] , as previously described [34] . For all other species , we project these gene sets based on gene orthologies [34] using the ortholog mapping at www . broad . mit . edu/regev/orthogroups . The chromatin features used in our analysis are listed and defined in Table S1 . To quantify the enrichment for a given feature within a functional category we used the two-sample K-S test . For each K-S test , we defined our two sample sets as genes within a given functional group and all other genes in the genome . The K-S test quantifies the distance between the distributions of a given chromatin feature for the two sets . The K-S statistic KK-S is defined as the maximum absolute difference between the cumulative distribution functions ( CDFs ) of the two samples . We estimated the P value , PK-S , for the statistical significance of this difference as follows: For further analysis , we converted P values to K-S scores , SK-S , where SK-S = ±log10 ( PK-S ) if the difference realizing the statistic KK-S is positive/negative , respectively . To account for multiple hypotheses testing , we only considered PK-S as significant if it was below the P value threshold for a False Discovery Rate of 5% [76] . This analysis was also applied to absolute expression levels , Poly ( dA:dT ) strength in NFRs , trans factor motif affinity scores in NFRs , and comparison in expression of sites located in NFRs versus sites located in nucleosomes . To subtract the effect of expression on observed chromatin features , we used robust Lowess smoothing . We smoothed the scatter data of each chromatin feature versus expression level using a Lowess linear fit and a smoothing window set to 10% of the span of expression level values . We assigned zero weight to outliers , defined as data more than six standard deviations from the mean . To remove the effect of expression , the Lowess fit was subtracted from its corresponding chromatin feature value . K-S functional enrichments for the Lowess subtracted chromatin features were calculated as described above . We assessed transcriptional activity by absolute RNA expression . We assessed intrinsic anti-nucleosomal sequence by Poly ( dA:dT ) strength in NFRs , since it explains the vast majority of the intrinsic sequence information and generalizes to all species in an unbiased manner . Other models of intrinsic sequence contribution [21] , [48] yielded similar results ( unpublished data ) . We assess the contribution of chromatin modifiers based on the Abf1 and Reb1 motif affinity scores in NFRs . This is a conservative estimate , since we only considered the two most established GRFs . To quantify the contribution of each of these factors on NFR occupancy , we used robust Lowess smoothing as described above . To compute the percent of variance explained by the robust Lowess fit , each NFR occupancy was assigned a “fitted” value Fi from the Lowess fitting line based on each of the three determinants . Then the variance of the residuals σ2R = var ( {Fi−Zi} ) was compared to the variance of the original data σ2D = var ( {Zi} ) . The percent of variance explained is defined as ( 1− ( σ2R/σ2D ) ) *100 . To find the percent variance explained by all determinants we first fit NFR occupancy versus one determinant , then iteratively took the residual , and fit it against the next determinant . For the figures , we first fit expression , then fit the successive residual versus Poly ( dA:dT ) tracts , and then fit the residual versus Abf1 and Reb1 motif affinity scores . Changing the order of the successive fits did not significantly reduce the total percent variance explained . Promoter sequences for each gene were defined as 1 , 000 bases upstream , truncated when neighboring ORFs overlapped with this region . We collected a library of Position Weight Matrices ( PWMs ) for several hundred S . cerevisiae DNA-binding proteins as previously defined [58] , [59] , [60] , [66] . Motif targets were identified via the TestMOTIF software program [77] using a 3-order Markov background model estimated from the entire set of promoters per genome . We considered all motif instances with P value <0 . 05 as significant . Since a few motifs had thousands of instances for this cutoff , we also limited the number of promoters with significant sites to the top 1 , 000 . The upper bound was chosen to exceed the maximal number of promoters bound ( 866 , P value <0 . 05 ) by any TF in S . cerevisiae , as measured by ChIP-chip [60] . For all subsequent motif analyses , we used the above criterion to define two sets of sites: ( 1 ) all significant sites within allowed promoters and ( 2 ) the best sites per allowed promoters . All motif instances were binned into five regions ( Nuc+1 , 5′NFR , Nuc−1 , Nuc−2 , and NFR2—the linker between Nuc−1 and Nuc−2 ) if their centers overlapped with the defined regions . In addition , sites were also split into two categories: Linkers ( 5′NFR and NFR2 ) and Nucs ( Nuc+1 , Nuc−1 , and Nuc−2 ) . We assigned the expression level of each gene to each site in the upstream promoter of that gene . We used a two-sample K-S test ( as described above ) to quantify the difference in expression levels between sites in Linkers versus Nucs . To quantify the preference of a motif for nucleosome depleted regions , we compared the mean log2 normalized nucleosome occupancy at all sites ( x ) against the mean log2 normalized nucleosome occupancy over the corresponding promoters ( y ) . To estimate the significance of the difference of the two vectors ( x-y ) , we used the paired Wilcoxon signed rank test that assigns a P value for rejecting the null hypothesis that x-y comes from a continuous , symmetric distribution with a zero median . To estimate the probability that k or more elements intersect subsets of n and m members at random in a superset of size N ( or the P value for overlap of k , PHG ) we summed over the right tail of a hypergeometric distribution: Using the hypergeometric P values , we estimated the significance of k overlaps between n genes with sites in their upstream promoter and m genes within a GO category , for a species with N genes . We represent each motif of length L by a position specific scoring matrix ( PSSM ) P , or the probability distribution P ( S1 , … , SL ) of that motif occurring over any sequence S1…SL . This is a standard approximation to a factor binding energy for sequence S1…SL . We also learned the 0th-order Markov background probability distribution B ( S1 , … , SL ) for each sequence S1…SL , set to the frequency of the four nucleotides in the promoter regions of a given species . We calculate A ( P , S ) , a motif's affinity score for an NFR sequence S , by summing the contributions of P ( S1 , … , SL ) /B ( S1 , … , SL ) over all allowable positions k in S as follows: Here , b ( Sk+j-1 ) is the background probability of the nucleotide Sk+j-1 of sequence S , and p ( Sk+j-1 , j ) is the probability for nucleotide Sk+j-1 in position j of the motif's PSSM . For the results in this study , we combined the contributions of both forward and reverse strands of each NFR . Also , normalizing the affinity by the length of each NFR sequence did not affect our results significantly . Prior to analysis , we log2-transformed the normalized nucleosome occupancy data ( Data post-processing , above ) , subtracted the mean , and divided by the standard deviation . Hence , the global nucleosome occupancy data for each species is approximately normal with zero mean and unit variance . We also used the same procedure for processing published in vitro data [21] . For each N-mer , we define the in vivo depletion score as the mean −log2 normalized nucleosome occupancy across all instances and all instances of the reverse complement . We also defined the depletion score relative to in vitro as power 2 of the difference between the in vivo depletion scores in each species and the in vitro depletion scores in S . cerevisiae ( also repeated for in vitro data from C . albicans [32] ) . The analysis was done for N = 5 , 6 , 7 , 8 and also repeated for N-mers found only in coding regions and only in upstream promoter regions . To annotate all Poly ( dA:dT ) tracts in each species and determine their nucleosome repelling strength we used an approach similar to a previously described one [48] . In summary , for each species' genome we found all PolyA or PolyT tracts of length L of 5 bp or more . We define the depletion score for a tract of length L as the mean of the −log2 normalized nucleosome occupancy across all instances of that length . This was calculated both using in vitro data from S . cerevisiae [21] and the in vivo data from each species . For long Poly ( dA:dT ) tracts with very few occurrences in a given genome we noticed a larger variation in the depletion score , likely due to small sample size . To mitigate this problem , we fit a line for depletion scores versus L using a weighted linear least squares fit with weights proportional to the number of occurrences for tracts of length L . We then used the line as an estimate for long tracts with fewer than 100 occurrences in a given genome . We iterated this procedure for all maximal Poly ( dA:dT ) tract with k allowed mismatches , k = 1 , … , 20 . The depletion score increases linearly with L for tracts with different k , confirming that a linear fit is appropriate ( Figure S11 ) . To aggregate all non-overlapping Poly ( dA:dT ) tracts within a given genome , we first quantized the strengths for each L . We define the fold depletion score of all tracts of length L as power 2 of the depletion score . We then quantized all Poly ( dA:dT ) tract fold depletion scores to the highest fold depletion level exceeding 2 , 4 , 8 , 16 , and 32 . For example , a tract with a depletion score of 3 . 5 is 23 . 5 = 11 . 3-fold depleted in nucleosomes relative to average , and would be assigned a fold depletion score of 8 . We next iterated over all Poly ( dA:dT ) tracts with mismatches k = 0 , … , 20 , replacing overlapping tracts only if the tract with more mismatches had a higher quantized fold depletion score . Data have been submitted to GEO , accession #GSE21960 .
Supplementary website http://www . broadinstitute . org/regev/evolfungi/ | Divergence in gene regulation plays a major role in organismal evolution . Evidence suggests that changes in the packaging of eukaryotic genomes into chromatin can underlie the evolution of divergent gene expression patterns . Here , we explore the role of chromatin structure in regulatory evolution by whole-genome measurements of nucleosome positions and mRNA levels in 12 yeast species spanning ∼250 million years of evolution . We find several distinct ways in which changes in chromatin structure are associated with changes in gene expression . These include changes in promoter accessibility , changes in promoter chromatin architecture , and changes in the accessibility of specific transcription factor binding sites . In many cases , changes in chromatin architecture are coupled to physiological diversity , including the evolution of a respiration- or fermentation-based lifestyle , mating behavior , salt tolerance , and broad aspects of genomic structure . Together , our data will provide a rich resource for future investigations into the interplay between chromatin structure , gene regulation , and evolution . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods",
"Supporting",
"Information"
] | [
"genetics",
"and",
"genomics/genomics",
"genetics",
"and",
"genomics/comparative",
"genomics",
"genetics",
"and",
"genomics/functional",
"genomics",
"genetics",
"and",
"genomics/gene",
"expression",
"genetics",
"and",
"genomics/nuclear",
"structure",
"and",
"function",
"molecular",
"biology/chromosome",
"structure",
"genetics",
"and",
"genomics/chromosome",
"biology",
"evolutionary",
"biology/genomics"
] | 2010 | The Role of Nucleosome Positioning in the Evolution of Gene Regulation |
The Sonic hedgehog ( Shh ) signaling pathway regulates developmental , homeostatic , and repair processes throughout the body . In the skin , touch domes develop in tandem with primary hair follicles and contain sensory Merkel cells . The developmental signaling requirements for touch dome specification are largely unknown . We found dermal Wnt signaling and subsequent epidermal Eda/Edar signaling promoted Merkel cell morphogenesis by inducing Shh expression in early follicles . Lineage-specific gene deletions revealed intraepithelial Shh signaling was necessary for Merkel cell specification . Additionally , a Shh signaling agonist was sufficient to rescue Merkel cell differentiation in Edar-deficient skin . Moreover , Merkel cells formed in Fgf20 mutant skin where primary hair formation was defective but Shh production was preserved . Although developmentally associated with hair follicles , fate mapping demonstrated Merkel cells primarily originated outside the hair follicle lineage . These findings suggest that touch dome development requires Wnt-dependent mesenchymal signals to establish reciprocal signaling within the developing ectoderm , including Eda signaling to primary hair placodes and ultimately Shh signaling from primary follicles to extrafollicular Merkel cell progenitors . Shh signaling often demonstrates pleiotropic effects within a structure over time . In postnatal skin , Shh is known to regulate the self-renewal , but not the differentiation , of touch dome stem cells . Our findings relate the varied effects of Shh in the touch dome to the ligand source , with locally produced Shh acting as a morphogen essential for lineage specification during development and neural Shh regulating postnatal touch dome stem cell maintenance .
The Hedgehog ( Hh ) pathway is conserved across the Metazoa subkingdom , and is one of a small number of intercellular signaling pathways that regulate the differentiation and pattering of morphologically diverse structures during development [1 , 2] . Postnatally , Hh ligands regulate tissue specific stem cell , homeostasis , and wound healing [3] . The basic molecular mechanisms of Hh signaling are still being investigated , and even less is known about how activation the pathway can result in such pleiotropic functions . Timing and distribution of Hh ligand delivery , ligand concentration , and duration of exposure can all influence signaling outcomes [4] . Multiple additional mechanisms have been proposed to alter Hh signaling , including ligand modification and sequestration , regulation of the primary cilium , modulation of Smo function by kinases , redundancy and cross regulation of Gli transcription factors , altering target gene expression by transcriptional co-regulators , and cross regulation by other signaling pathways [1–5] . In the present study , we discovered that Shh is the final critical element in a signaling cascade that specifies the touch dome lineage in developing mouse skin . Contrasting these findings with the role of Shh in regulating postnatal touch dome stem cells [6] , we found the changing function of Shh was accompanied by a change in the source of the ligand , suggesting an additional contextual mechanism that influences the results of Shh signaling . The functional diversity of vertebrate skin depends greatly on the variety of ectodermal appendages it produces . The development of ectodermal appendages including hair follicles , teeth , sweat glands , and mammary glands is precisely regulated by networks of signaling pathways including Wnt/β-catenin , Eda/Edar , Shh , and BMP [7] . Hair follicle development is particularly well studied as a model of ectodermal appendage development [8] . Identifying and comparing the developmental networks that control the specification and differentiation of ectodermal lineages can provide insights into developmental disorders and genetic diseases . The touch dome ( TD ) is a specialized epidermal sensory structure composed of K8+ Merkel cells ( MCs ) arrayed among columnar basal keratinocytes that express the hair follicle keratin K17 . Cells of the TD are morphologically and molecularly distinct from the adjacent interfollicular epidermis . The TD arises from the K14+ ectoderm [9 , 10] and is maintained as a distinct epidermal lineage by resident stem cells [6 , 11–13] . In postnatal skin , self-renewal of TD stem cells is regulated by Shh from sensory neurons that innervate the MCs [6] . MC development requires the transcription factor Atoh1 [14] and is regulated by levels of Sox2 expression [15 , 16] . TD MCs can be identified in the epidermis based on their expression of Atoh1 , Sox2 , or K8 [17] . TD MC development in the embryonic ectoderm is spatially and temporally associated with that of primary hair follicles . In embryonic day 14 ( E14 ) mice , a primary wave of hair follicle placode induction takes place . These primary follicles produce guard hairs that ultimately comprise ~2% of the adult mouse coat . Secondary ( E16 ) and tertiary ( E18 ) waves of hair follicle induction are responsible for forming the three remaining types of hair follicles [18] . Developing MCs are first detected at ~E15 in association with nascent primary hair germs . Just before birth , TD MCs surround the infundibulum at the top of guard hair follicles . Shortly after birth , the touch dome has migrated into a crescent-shaped domain just caudal to the guard hair follicle ( Fig 1A ) . Planar cell polarity signals regulate the reorganization of the forming TD [19] . Loss of Eda function results in abnormal guard hair formation and an absence of TD MCs , however the mechanism by which Eda signaling impacts MC specification is unclear [20] . Other signals involved in TD MC development are largely unknown . To elucidate the developmental requirements for TD MC formation , we used genetically modified mice to disrupt signaling pathways known to be important in the development of other ectodermal appendages . We found that embryonic deletion of dermal β-catenin prevented TD MC formation . Similarly , Edar mutant skin failed to generate TD MCs . The mechanism of MC loss in these mice was due to failure of Shh expression by primary hair follicles , as evidenced by an absence of TD MC formation in Shh-null skin . Moreover , independent deletion of Shh or Smo in the embryonic epidermis reveled that intraepithelial Shh signaling from primary hair germs was necessary for TD MC specification . Notably , the loss of MC specification in Shh-deficient skin was observed before any disruption in hair follicle development was apparent . The importance of Shh signaling in TD MC formation was further demonstrated by using a Smo agonist to rescue MC specification in ex vivo-cultured Edar mutant skin . In contrast , Fgf20 was dispensable in TD MC development , demonstrating that the signaling cascades required for guard hair follicle development and TD formation diverge at the level of Fgf20 signaling . Distinction from the forming guard hair follicle was further demonstrated when fate mapping of the follicle lineage showed that TD MC progenitors predominantly arise in the epidermis outside the hair placode . Thus , like other ectodermal appendages , the touch dome is a distinct epidermal lineage whose specification and development requires Wnt-dependent mesenchymal-epithelial interactions and reciprocal signaling within the developing ectoderm , including Eda signaling to primary hair placodes and subsequent Shh production by primary hair germs . The critical role for Shh signaling in embryonic TD specification is dependent on locally produced ligand , whereas the regulation of postnatal TD stem cells requires Shh transported to the skin by sensory neurons . These observations suggest that ligand source influences the differential effects of Shh signaling in the TD .
Mesenchymal-epithelial interaction is critical in hair follicle morphogenesis [21] . In mice with dermal β-catenin conditional knockout , hair follicle development is arrested at a very early stage [22] . We investigated the effect of dermal β-catenin deletion on TD MC development using En1Cre/+; β-cateninflox/Δ mouse embryos where Cre recombination occurs in the dorsal trunk mesenchyme that forms the dermis [22] . As expected , dorsal trunk skin of E17 . 5 affected mice had no evidence of developing hair follicles on histological sections ( Fig 1C ) . Using whole mount K8 immunostaining to detect MCs , we found K8+ TD MC adjacent to developing guard hair follicles in dorsal trunk skin of control ( En1Cre/+; β-cateninflox/+ , n = 5 ) E17 . 5 embryos ( Fig 1B ) . In contrast , we found no K8 staining in dorsal trunk epidermis of E17 . 5 En1Cre/+; β-cateninflox/Δ embryos ( n = 3 , Fig 1B ) , indicating that β-catenin function in the developing dermis is necessary for the formation of TD MCs . En1Cre is not expressed in proximal limb mesenchyme [23] . As an internal control , we examined limb skin of E17 . 5 En1Cre/+; β-cateninflox/Δ embryos and observed normal collections of K8+ MCs in TDs adjacent to hair follicles , further implicating dermal Wnt/β-catenin signaling in TD MC development . Dermal Wnt signaling is necessary for the earliest steps of epidermal patterning , including the upregulation of Edar expression in nascent hair placodes [22] . Edar is the membrane receptor for the TNF family protein Eda . MCs do not form in Eda mutant Tabby mice that have defective guard hair development [20] . We hypothesized that loss of Edar function would also disrupt MC development . To test this , we used point mutant Edardl-J/dl-J ( downless ) mice that lack guard hairs as their abortive primary hair germs fail to form follicles ( n = 3 , Fig 2B ) [24] . No MCs were detected in postnatal day 0 ( P0 ) dorsal trunk skin by whole mount K8 immunostaining ( Fig 2A ) . We also failed to detect TD MCs in adult Edardl-J/dl-J; Atoh1LacZ/+ mouse skin stained with X-gal ( n = 3 , Fig 2C ) [25] , demonstrating that TD MC formation is truly abrogated in the absence of Edar function , and not simply delayed . To examine the early molecular specification of TD MC , we examined gene expression in E15 . 5 Edardl-J/dl-J dorsal trunk skin ( n = 3 ) using quantitative reverse transcription–PCR ( RT-PCR ) . Relative to control skin , Edar mutant skin shows significantly reduced mRNA levels of the MC differentiation markers Sox2 and Atoh1 ( Fig 2D ) , suggesting that MC specification itself is disrupted . A difference in K8 expression was not detected , likely due to the broad expression of K8 in the periderm of embryonic skin [26] . As Sox2 is also expressed in the forming dermal papilla of E15 . 5 primary hair germs [16] , these results further suggest that dermal papilla differentiation is disrupted in Edardl-J/dl-J mouse skin . Together , these results strongly suggest that Edar signaling in the primary hair placode critically regulates TD MC formation and that Edar loss alone can explain the absence of TD MC development in the dermal β-catenin knockout skin . There are a number of signaling molecules expressed downstream of Edar in nascent primary hair follicle that could potentially influence MC specification and development . Shh is a morphogen expressed by forming hair placodes [27] , and Edardl-J/dl-J mice fail to express Shh in their abortive primary hair germs ( Fig 2D ) [24] . As nerve-derived Shh was recently shown to regulate TD stem cell renewal in adult mouse skin [6] , we hypothesized that Shh production was the mechanism by which primary hair follicles regulate TD MC development . Embryos deficient for Shh survive to birth but die postnatally due to holoprosencephaly [28] . Shh-null mice have defective hair follicle development where hair germs form but do not elongate into follicles [29 , 30] . We generated ShhGFPcre/GFPcre mouse embryos ( n = 3 ) to assess TD MC formation in Shh null skin [31] . As expected , E18 . 5 ShhGFPcre/GFPcre hair follicles were arrested as hair germs ( Fig 3C ) . In E18 . 5 control epidermis ( n = 3 ) , K8+ TD MCs were observed in annular clusters around primary hair follicles ( Fig 3A and 3B ) . In contrast , no TD structures were detected in Shh-null mouse embryos , and the epidermis contained very few scattered K8+ cells ( Fig 3A and 3B ) , demonstrating that Shh signaling is essential for TD MC development . Using the Gli1LacZ reporter allele to visualize cells with active hedgehog signaling [32] , we observed diffuse X-gal staining in E17 . 5 and P0 basal epidermis , dermis , and hair follicles ( S1A and S1B Fig ) , demonstrating that there are hedgehog-responding cells throughout the epidermis of developing mouse skin . This finding is distinct from adult epidermis where Gli1 expression is restricted to the mature touch dome [12 , 33] . No LacZ reporter expression was detected in P0 ShhGFPcre/GFPcre; Gli1LacZ/+ trunk skin ( S1C Fig ) , suggesting that Shh is the primary Hh ligand regulating skin development . To test if hedgehog-responding cells can serve as Merkel cell precursors , E15 . 5 Gli1CreER/+; R26YFP/+ embryos ( n = 3 ) were treated with low-level tamoxifen ( 2mg to the gravid dam ) to genetically label a fraction on Gli1+ cells at the time of MC specification [34] . Scattered labeled cells were visible in the primary hair follicles , dermis , and epidermis of P0 sectioned dorsal trunk skin ( S2 Fig ) . GFP staining was found in 27 . 6% of K8+ Merkel cells ( n = 333 ) . This was comparable to GFP labeling in 24 . 3% of primary hair follicle epithelial cells ( n = 1847 ) , a structure derived from precursors that highly express hedgehog response genes at E15 . 5 [27] . These fate mapping results demonstrate that hedgehog-responding precursors give rise to TD MCs . The requirement for Shh in TD MC formation suggests that Shh loss accounts for the MC defect in Edar mutant skin . To test whether Shh signaling was indeed the critical mechanism regulating MC formation downstream of Edar , we used the Smo agonist Hh-Ag1 . 5 to restore hedgehog signaling in embryonic Edar mutant skin during the window of MC specification . We cultured E13 . 5 Edardl-J/dl-J skin ( n = 3 ) , and after two days ( E15 . 5 ) , used RT-PCR to assess gene expression . In untreated cultures ( n = 3 ) , we observed reductions in Sox2 , Atoh1 , Shh , and Gli1 relative to control skin ( n = 3 ) , similar to those seen in uncultured E15 . 5 Edar mutant skin ( Figs 2D and 3D ) . Smo agonist treatment resulted in significant increases of Gli1 transcription in both control and Edar mutant skin , confirming activation of hedgehog signaling . Smo agonist also resulted in a reduction of Shh expression in control skin , suggesting that some form of negative feedback downregulates Shh transcription . Similarly , Smo agonist further reduced Shh levels in Edar mutant skin . Interestingly , Smo agonist reduced Atoh1 mRNA levels in control skin . This result is consistent with the observation that Shh signaling in neurons prevents Atoh1 degradation [35] and suggests that elevated Atoh1 protein levels in the setting of a Smo agonist can feedback to reduce Atoh1 transcripts . Most importantly , Smo agonist rescued expression of the MC differentiation markers Sox2 and Atoh1 to normal levels in Edar mutant skin ( Fig 3D ) . Together , these results suggest that Shh is both necessary and sufficient for MC formation in embryonic trunk skin and is the critical factor lost in Edar mutant skin . In embryonic skin , Shh is initially expressed in hair follicle placodes and continues to be expressed in developing hair follicle bulbs [36] . However , Shh is also delivered to the skin by sensory nerves [33] . To test the requirement for follicle-produced Shh , we deleted Shh from the embryonic epidermis using K14-Cre; Shhflox/flox mice [37 , 38] . Unlike control skin ( n = 3 ) , no TD MCs were detected in P0 mutant ( n = 3 ) epidermis by immunostaining ( Fig 4A and 4B ) , indicating that epidermal Shh is necessary for TD MC production . Similar to the Shh null mice , hair follicles failed to develop in K14-Cre; Shhflox/flox skin . Abortive K17+ germ-like structures formed at sites of follicle induction but failed to elongate ( Fig 4B , 4C and 4E ) . At the time of TD MC specification in E15 . 5 control skin ( n = 2 ) , we observed Sox2+ and K8+ cells in the epidermis above and adjacent to primary hair germs . In E15 . 5 K14-Cre; Shhflox/flox skin ( n = 2 ) , neither Sox2 nor K8 staining was observed , suggesting that hair placode/germ-derived Shh is necessary for MC specification and that MC loss occurs prior to , and independent of , the follicle downgrowth defect in Shh mutant skin ( Fig 4D and 4E ) . Sox2+ dermal papillae were observed under the abortive primary hair germs in E15 . 5 K14-Cre; Shhflox/flox skin ( Fig 4E ) , suggesting epidermal Shh is not necessary for dermal papilla specification . Next , we used K5-tTA; TRE-Cre; Smoflox/flox mice in the absence of doxycycline to delete the obligate Shh signaling mediator Smo in the developing epidermis [39 , 40] . At E18 . 5 , these mice ( n = 2 ) completely lacked TD MCs that were readily detected by K8 immunostaining in control epidermis ( n = 9 , Fig 4F ) , demonstrating that the cells that require Shh signaling for TD MC development reside in the epidermis . Together , these results strongly suggest that intraepithelial Shh signaling from primary hair placodes/early hair germs to epidermal target cells is required for the specification and development of TD MCs . Although K8+ cells seem to arise within forming primary hair follicles [20] , our observations that Sox2+ and K8+ cells first appear in the epidermis above and adjacent to primary hair germs made us question whether TD MC progenitors develop within the hair placode or simply in close proximity to the hair follicle lineage . We used ShhGFPcre/+; R26YFP/+ mice to fate map the hair follicle lineage originating from the Shh-expressing hair placode [41] . To determine whether TD MCs arise from within the hair follicle lineage , we used immunostaining to assess co-labeling of Sox2+ and K8+ MCs with GFP staining in ShhGFPcre/+; R26YFP/+ skin at E14 . 5 ( n = 3 ) , E16 . 5 ( n = 3 ) , and E18 . 5/E19 . 5 ( n = 3 , Figs 5A , 5B , S3 and S4 ) . In total ( n = 189 MC ) , over 89% of MCs failed to stain with GFP , indicating that TD MCs primarily originate from extrafollicular cells . A breakdown of staining characteristics for MCs associated with hair follicles at different developmental stages [42] is shown in Fig 5C . A similar lack of GFP+ K8+ TD MCs was seen in adult ShhGFPcre/+; R26YFP/+ skin ( Fig 5D ) , indicating that the stem cells maintaining TD MCs [13] also principally originate outside of the hair follicle lineage . Together , these results indicate that TD MCs predominantly arise from a developmental linage outside of primary hair placodes , although Shh+ ectodermal cells can give rise to a minor fraction of TD MCs . The presence of Sox2+ K8- MCs that are more abundant in E14 . 5 skin versus later stages of development , and the complete absence of Sox2- K8+ cells ( Figs 5C and S5A ) , suggests that Sox2 expression is an early event in MC specification and that K8 upregulation occurs later in MC differentiation . This finding is consistent with prior observations [16] and the fact that adult TD MC progenitors express very low levels of K8 and upregulate K8 expression upon differentiation to mature MCs [6 , 13] . In E14 . 5–16 . 5 skin , TD MCs were scattered in the epidermis adjacent to primary hair germs . In late embryonic and early postnatal skin ( E18 . 5-P4 ) , TD MCs were organized around the infundibulum of primary hair follicles , but were also found clustered on the upper caudal side of primary hair follicles , in a region known to express NCAM ( Figs 5A , 5B , S3 , S4 , S5B , S5C ) [20] . Interestingly , the MCs and NCAM+ keratinocytes found in ShhGFPcre/+; R26YFP/+ upper follicles infrequently expressed YFP ( Figs 5A and S3 ) . This observation suggests that the caudal side of the upper developing primary follicle does not form by downgrowth of the hair follicle placode but primarily by expansion of extraplacodal cells into the forming follicle . MCs do not persist in the upper region of primary follicles , as K8+ MCs are not detected within P13 guard hair follicles [6] . The purpose of the transient population of follicular MCs during late embryonic and early postnatal development is unclear . We used mice with dermal β-catenin deletion , Edar mutation , Shh mutation , epidermal deletion of Shh , and epidermal deletion of Smo to show that paracrine Shh signaling within the epidermis , downstream of Wnt and Eda signaling , is essential for TD MC development . However , in all of these mice , there is either a global defect in hair follicle patterning and development , a developmental defect in all hair follicles , or a defect in primary hair follicles . To separate Shh signaling from normal guard hair development , we used Fgf20-null ( Fgf20LacZ/LacZ ) mice [43] . Like Shh , Fgf20 is expressed by hair placodes and is regulated by Wnt/β-catenin and Eda/Edar signaling . In Fgf20 mutant skin , there is a defect in the formation of hair follicle-associated dermal papillae and a failure in the downgrowth of primary hair germs [43] . Although the abortive primary hair germ phenotype of Fgf20 mutant skin resembles that of Eda and Edar mutant skin , Shh expression is preserved in the defective primary hair follicles [43] . We used RT-PCR to confirm that Shh is expressed in E15 . 5 Fgf20LacZ/LacZ skin ( n = 3 ) and found Shh levels elevated compared with those in control skin ( n = 3 , Fig 6A ) . A small , nonsignificant increase in Gli1 mRNA levels was also observed . In P0 Fgf20 mutant skin ( n = 3 ) , we occasionally observed larger follicles that were comparable to guard follicles in control skin ( n = 3 , Fig 6B ) , suggesting that after an initial arrest , primary follicle downgrowth can occur . Nonetheless , the primary follicles were not normal , as guard hairs were absent in the coats of juvenile and adult Fgf20LacZ/LacZ mice [43] . In E15 . 5 Fgf20LacZ/LacZ skin , there were decreased levels of Edar , Sox2 , and Atoh1 expression , although only the Atoh1 reduction reached statistical significance ( Fig 6A ) , and a Sox2 reduction was expected based on the dermal papilla defect in this mouse . Despite the reduction in MC factors at the time of MC specification , the pattern and distribution of K8-immunostained TD MCs was normal in P0 Fgf20LacZ/LacZ skin ( n = 4 mice , 752 TD , 11 , 393 MC , Fig 6B and 6D ) . We did observe a 17% reduction in the mean number of MC per TD relative to control epidermis ( n = 2 mice , 348 TD , 6 , 338 MC , Fig 6B and 6D ) . Normal-appearing TDs persisted into adulthood with a normal TD density and a continued reduction ( 14% ) in MC/TD observed in adult ( P50-P103 ) Fgf20LacZ/LacZ mice ( n = 3 mice , 131 TD , 1685 MC ) relative to control ( n = 3 mice , 136 TD , 1979 MC , S6 Fig ) . These results demonstrate that Fgf20 is required for normal guard hair follicle development but is not necessary for TD MC specification; our results also illustrate that when Shh expression is preserved , even abnormal primary hair germs are capable of supporting TD MC development .
Vertebrate skin provides barrier , mechanical , defensive , communicative , sensory , metabolic , and homeostatic functions . This diversity of function is achieved , in part , by specialization of ectodermal appendages that form through a series of mesenchymal-epithelial interactions in the embryo . By regulating common inductive events with similar yet distinct sets of morphogens , great diversity of structure and function is achieved [7] . Although some types of ectodermal appendages are specific to mammals ( hair follicles , mammary , and sweat glands ) , structures that enhance sensation of the outside world are common adaptations across all classes of vertebrates . We have used multiple genetically modified mouse models to elucidate for the first time the developmental signaling cascade required for the formation of sensory TDs in embryonic skin . Only recently was the TD discovered to be a distinct skin lineage , maintained by its own resident stem cells . Here , we find that the mechanism for establishing the TD lineage within the developing ectoderm requires many of the morphogens that drive formation of ectodermal appendages . We discovered that TD MC specification requires Wnt-dependent mesenchymal signals to establish reciprocal signaling within the developing ectoderm , including Eda signaling to primary hair placodes , and subsequent Shh signaling from primary follicles to extrafollicular MC progenitors ( Fig 7 ) . Our identification of primary follicle Shh as a critical regulator of MC specification is consistent with TD development being spatially and temporally associated with the first wave of hair follicle induction in embryonic trunk skin , whereas our fate mapping results confirm that the TD lineage is separate from adjacent hair follicles . Together , our findings identify the TD as a distinct ectodermal touch receptor whose development is critically regulated by Wnt , Eda/Edar , and Shh signaling . Our results support a model of TD MC development in which MC specification is tied to primary hair follicle patterning and morphogenesis up to the point of Shh production by the early follicle and can diverge from follicle development with signals such as Fgf20 that act in parallel or downstream to Shh in the hair follicle . Similarities in the developmental signaling pathways between the TD and primary hair follicles include the importance of mesenchymal Wnt signaling and ectodermal Eda/Edar signaling . Accordingly , there is no Shh expression by primary hair placodes in En1Cre/+; β-cateninflox/Δ skin , Edardl-J/dl-J skin , or Eda mutant Tabby skin [44] , and TD MCs fail to form in all these mice . Although Shh signaling is critical in the development of both hair follicles and TDs , the role of Shh appears different for each structure . In the hair follicle , Shh signaling is dispensable for placode induction but is needed for the subsequent growth of the follicle . In the TD , loss of Shh signaling by either deleting Shh production or removing epidermal Smo results in complete loss of MC specification and development—a phenotype that is evident even before changes are seen in the primary hair germ . Another contrasting feature is that a hedgehog signaling agonist was sufficient to rescue MC differentiation in Edar mutant skin , whereas transgenic expression of Shh in Eda mutant Tabby mice failed to restore primary hair follicle development [45] . Downstream of Eda/Edar signaling , Fgf20 expression in the developing hair placode is needed for proper dermal papilla and guard hair formation , while Shh is expressed and TD MCs are able to form . Thus , although TDs and hair follicles share early developmental requirements for Wnt and Eda signaling , they arise from distinct ectodermal compartments , are maintained as distinct lineages , and differ in their specific requirements for Shh and Fgf20 . The coats of Fgf20 mutant mice lack guard hairs , and yet typical-appearing TDs were found at normal density and only a slightly reduced average number of MCs per TD . Interestingly , TD MCs are maintained in adult hairless mice , where hair follicles develop normally but undergo cystic degeneration after the first month of life [6] , demonstrating that once established , the TD lineage can persist without an affiliated guard follicle . Thus , the primary hair follicle is a necessary source of Shh for TD MC formation in embryonic skin but is dispensable in postnatal MC maintenance . In postnatal mouse skin , maintenance of TD stem cells requires Shh signaling from sensory neurons that innervate MCs . However , TDs are specified at a time prior to epidermal innervation [46] , and embryonic deletion of Shh from sensory neurons had no impact on TD MC formation [6] . It is noteworthy that Shh critically regulates different TD functions during development and in postnatal skin . In the embryo , hair follicle Shh is required to establish the MC lineage . Postnatally , loss of neural Shh blocks the maintenance of TD stem cells , but MC differentiation continues until the progenitor pool is exhausted [8] . This is analogous to the developing telencephalon , where Shh from the prechordal mesoderm and ventral forebrain is a critical morphogen for neural patterning and development [47] , however postnatal neural stem cells are maintained by Shh from local niche neurons [48 , 49] . Shh signaling often shows pleiotropic effects within a given organ system with roles in patterning , specification , and proliferation during development and later functions in stem cell regulation , tissue regeneration , and cancer formation . Our findings and the observations from the central nervous system suggest that altering the source of ligand is an important contextual component influencing the function of Shh signaling within a tissue . The formation of TDs adjacent to primary hair follicles requires Shh signaling from the nascent follicles . Secondary and tertiary hair follicles also express Shh during their formation , and yet TDs do not form in association with induction of those hair follicles . Moreover , based on the broad expression of Gli1 within the developing epidermis , many embryonic epidermal cells receive Shh signaling . Thus , even though a hedgehog signaling agonist was sufficient to rescue MC differentiation in Edar mutant skin , Shh signaling alone is not sufficient to induce TD MC specification in any and all developing ectoderm . Further experimentation will be necessary to identify the factors that establish the temporal and spatial competency of developing epidermis to respond to Shh signaling with MC specification . However , Polycomb repressive complex 2 ( PRC2 ) activity appears to be important in preventing MC specification around secondary and tertiary hair follicles ( personal communication , E . Ezhkova ) . Notch signaling may also play a role in limiting MC specification , as ectopic expression of Atoh1 is able to induce MC production in some epithelial compartments of the skin , and impeding Notch signaling facilitates this process [50] . Fate mapping showed that MCs arise predominantly from ectoderm outside the forming hair follicle . This is consistent with our observations that the first MCs appeared as Sox2+ cells in the E14 . 5/15 . 5 epidermis above and adjacent to forming primary hair germs . We confirmed that late embryonic and early postnatal guard hair follicles contain a discrete population of MCs , however these cells also tend to arise from outside the hair follicle lineage . Although the purpose of the MCs in the upper regions of primary hair follicles during development is unclear , it has been proposed that the NCAM expressed on keratinocytes around these MCs may facilitate MC innervation [20] . Coincidently , the early postnatal window corresponds to a period when TD MC innervation undergoes pruning and maturation [46][51] , and the onset of perineural influence on MC maintenance [6] . Currently , it is not known whether the MCs in the follicle during this period eventually die , change their fate , or migrate to TDs in the epidermis . We have determined that intraepithelial Shh signaling from the developing hair follicle to MC precursors is a critical factor in TD MC production . However , the precise location of the target cells for Shh remain undefined . Because the Hh response gene Gli1 is broadly expressed in the developing epidermis , it cannot be used to identify the specific K5+ Shh-responding cells required for TD MC development . Just as adult TD stem cells are Gli1+ [6] , our Gli1 fate mapping experiments in the embryo show that TD MC progenitors are a direct target of Shh signaling . Because MC development within the TD anlage is dependent on Shh signaling from developing hair follicles , it is reasonable to assume the TD anlage forms in proximity to primary hair placodes . Moreover , the appearance of early MCs in the epidermis around primary hair germs and the observation that a minor portion of TD cells originate from the hair placode lineage suggest that the TD anlage is likely the ectoderm immediately adjacent to , and slightly overlapping , the primary hair placode . It is uncertain whether the TD forms from its own placode and associated mesenchymal condensate; however , this is unlikely , as no such structures have been observed . It is more likely that the TD is induced in the adjacent ectoderm by inchoate primary hair follicles or by the same signals that induce the follicles . Nonetheless , as an independent epidermal lineage requiring mesenchymal induction during development , and being absent in mouse models of ectodermal dysplasia ( Eda and Edar mutant mice ) , TDs can be considered ectodermal appendages , or at least accessory structures to ectodermal appendages . This work elucidates the developmental signaling requirements for the sensory TD and illustrates the commonalities and contrasts that exist between TD development and development of the closely associated guard hair follicle . These results further our functional understanding of skin patterning and development , the repertoire of ectodermal appendage formation , and how specialized sensory structures form prior to interfacing with the sensory nervous system . Intriguingly , along with prior observations , these findings indicate that the distinct functions of Shh signaling in TD development and maintenance correspond to changes in the source of the Shh ligand required for the varied effects .
Mice were housed and bred on an outcrossed Swiss Webster background in a pathogen-free facility at the National Cancer Institute ( NCI ) , Bethesda , MD . Genotyping of mice was performed by allele-specific PCR on DNA extracted from tail tissue . All experiments were performed in accordance with institutional guidelines according to IACUC-approved protocols . En1Cre/+; β-cateninflox/Δ and control samples were provided by Dr . Radhika P . Atit . Some ShhGFPcre/+; R26YFP/+ samples were provided by Dr . Sunny Y . Wong . Edardl-J/+ , Atoh1LacZ/+ , ShhGFPcre/+ , Gli1LacZ/+ , Gli1CreER/+ , K14-Cre , Shhflox/+ , K5-tTA , TRE-cre , Smoflox/+ , R26YFP/+ , R26LacZ/+ , and Fgf20LacZ/+ mice were described previously as cited in the text . For all embryonic and neonatal observations , littermate animals with a wildtype copy of the targeted allele and/or lacking cre recombinase activity were used as controls . K5-tTA; TRE-cre; Smoflox/flox mice were bred and maintained on a standard rodent diet without doxycycline during embryonic development . Tamoxifen ( Sigma ) was dissolved in corn oil ( 20mg/ml ) and administered ( 2mg intraperitoneal injection per gravid mouse ) to induce CreER . Skin was fixed in 4% paraformaldehyde for 15 minutes ( for X-gal staining ) or overnight ( for immunostaining ) . Tissue was whole mount-stained or cryoprotected overnight in 30% sucrose , embedded in OCT , and frozen; 12-μm sections were obtained . Standard and whole mount immunostaining procedures were performed . Tissue sections on glass slides were fixed in 4% paraformaldehyde for 15 minutes before incubation in 10% serum in 0 . 1% PBT ( 0 . 1% Triton X-100 in PBS ) for 1 hour and then in primary antibody ( in 5% serum/0 . 1% PBT ) overnight at 4°C . The primary antibodies used were: rat anti-K8 ( 1:50 , University of Iowa ) , rabbit anti-K17 ( 1:200 , Epitomics ) , chicken anti-GFP ( 1:1000 , Abcam ) , rabbit anti-GFP ( 1:500 , Abcam ) , rabbit anti-Sox2 ( 1:500 , Stemgent ) , and rabbit anti-NCAM ( 1:500 , Millipore ) . Alexa Fluor-conjugated secondary antibodies ( 1:2000 , Invitrogen ) were used to detect the signals . Whole mount immunostaining followed the online protocol as described [52] . Concomitant staining of littermate control tissue and control staining where the primary antibody was omitted were used to confirm the specificity of experimental staining . Confocal images were acquired with the Zeiss LSM 710 Confocal system ( Carl Zeiss Inc , Thornwood , NY ) . The Troma-1/K8 antibody developed by Dr . Philippe Brulet and Dr . Rolf Kemler was obtained from the Developmental Studies Hybridoma Bank , developed under the auspices of the National Institute of Child Health and Human Development and maintained by The University of Iowa , Department of Biology , Iowa City , IA 52242 . Merkel cells per touch dome and touch dome per mm2 numbers were assessed by direct visualization of immunofluorescently stained K8+ Merkel cells in whole mount dorsal trunk skin from experimental and control animals . Counting was performed by a blinded observer . In tissue sections , Merkel cells were visualized by K8 and/or Sox2 staining and were scored for co-staining with GFP . Reporter recombination in primary hair follicles was quantified by counting total GFP staining cells and dividing by total number of DAPI stained nuclei within a preselected region of each follicle . Total RNA from embryonic trunk skin was purified using RNeasy Micro Kit ( Qiagen ) and reverse-transcribed into cDNA following the manufacturer’s manual ( Invitrogen #11752 ) . Then quantitative RT-PCR , with actin as control , using SYBR Green was performed to detect RNA expression . Data are presented as means ± SD . The sequences of PCR primers are shown in S1 Table . Embryonic skin culture was performed as described [53] . E13 . 5 embryos were collected , each embryonic mouse was cut through the sagittal midline and eviscerated , half was cultured in media ( DMEM +10% FBS +1% penicillin/streptomycin ) as a control group , and half was cultured in media with 25 nM Smo agonist Hh-Ag1 . 5 ( Xcess Biosciences Inc . ) as the treated group . Media was changed the next day , and skin was harvested 2 days after culture , followed immediately by RNA isolation . Population data sets are shown as the mean values , and error bars represent SD . For comparisons between sets , a two-tailed t-test was applied . | Sonic hedgehog ( Shh ) is one of a limited set of signaling molecules that cells use to drive organ formation during development and tissue regeneration after birth . How Shh signaling achieves different biological effects in the same tissue is incompletely understood . Touch domes are unique sensory structures in the skin that contain innervated Merkel cells . Using mouse genetics , we show that touch domes develop in tandem with , but distinct from , primary hair follicles . Moreover , touch dome specification requires a cascade of cell-cell signaling that ends with Shh signaling from an adjacent primary hair follicle . It was previously shown that Shh signaling from sensory nerves regulates the maintenance of touch dome stem cells after birth . Thus , the critical role for Shh signaling in embryonic touch dome specification is dependent on locally produced Shh , whereas the renewal of touch dome stem cells requires Shh transported to the skin by sensory neurons . These observations suggest that the distinct functions of Shh in touch dome development and maintenance correspond to changes in the source of the Shh signal required for the varied effects . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"skin",
"merkel",
"cells",
"medicine",
"and",
"health",
"sciences",
"integumentary",
"system",
"epithelial",
"cells",
"developmental",
"signaling",
"epidermis",
"research",
"and",
"analysis",
"methods",
"specimen",
"preparation",
"and",
"treatment",
"staining",
"hair",
"animal",
"cells",
"biological",
"tissue",
"hair",
"follicles",
"hedgehog",
"signaling",
"signal",
"transduction",
"anatomy",
"cell",
"biology",
"epithelium",
"biology",
"and",
"life",
"sciences",
"cellular",
"types",
"wnt",
"signaling",
"cascade",
"cell",
"signaling",
"signaling",
"cascades"
] | 2016 | A Cascade of Wnt, Eda, and Shh Signaling Is Essential for Touch Dome Merkel Cell Development |
Cholera remains a substantial health burden in Asia and Africa particularly in resource poor settings . The standard procedures to identify the etiological organism V . cholerae are isolation from microbiological culture from stool as well as Polymerase Chain Reaction ( PCR ) . Both the processes are highly lab oriented , labor extensive , time consuming , and expensive . In an effort to control for outbreaks and epidemics; an effective , convenient , quick and relatively less expensive detection method is imperative , without compromising the sensitivity and specificity that exists at present . The objective of this component of the study was to evaluate the effectiveness of a locally produced rapid diagnostic test ( RDT ) for cholera diagnosis . In Bangladesh , nationwide cholera surveillance is ongoing in 22 hospitals covering all 8 divisions of the country since June , 2016 . In the surveillance , stool samples have been collected from patients presenting to hospitals with acute watery diarrhea . Crystal VCTM ( Span diagnostics , India ) and Cholkit ( locally produced RDT ) have been used to detect V . cholerae from stool samples . Samples have also been sent to the main laboratory at icddr , b where the culture based isolation is routinely performed . All the tests were carried out for both direct and enriched stool samples . RDT sensitivity and specificity were calculated using stool culture as the gold standard . A total of 7720 samples were tested . Among these , 5865 samples were solely tested with Crystal VC and 1355 samples with Cholkit whereas 381 samples were tested with both the RDTs . In comparison with culture , direct testing with Crystal VC showed a sensitivity of 72% ( 95% CI: 50 . 6% to 87 . 9% ) and specificity of 86 . 8% ( 95% CI: 82 . 8% to 90 . 1% ) . After enrichment the sensitivity and specificity was 68% ( 95% CI: 46 . 5% to 85 . 1% ) and 97 . 5% ( 95% CI: 95 . 3% to 98 . 8% ) respectively . The direct Cholkit test showed sensitivity of 76% ( 95% CI: 54 . 9% to 90 . 6% ) and specificity of 90 . 2% ( 95% CI: 86 . 6% to 93 . 1% ) . This evaluation has demonstrated that the sensitivity and specificity of Cholkit is similar to the commercially available test , Crystal VC when used in field settings for detecting V . cholerae from stool specimens . The findings from this study suggest that the Cholkit could be a possible alternative for cholera endemic regions where V . cholerae O1 is the major causative organism causing cholera .
Even with the development of modern established treatments and preventative measures , cholera still remains a major health burden in low-income countries with limited resources , particularly in the developing world . Cholera is a water-borne infectious disease which can be characterized by life-threatening secretory diarrhea , often accompanied by numerous voluminous watery stools and vomiting [1] . Clinical consequences range from asymptomatic to severe disease with massive watery diarrhea which may become fatal if untreated [2] . Globally , an estimated 1 . 3 billion people are at risk of cholera where India and Bangladesh jointly constitute the largest share of population at risk . In Bangladesh , according to estimations , at least 66 million people are at risk of cholera with an estimated 109 , 052 cholera cases annually [3] . While many infections can result in only mild symptoms , at least 300 , 000 severe cases occur annually which are severe enough requiring hospitalisation [4] . The causative agent of cholera at present is Vibrio cholerae O1 , a Gram-negative pathogen . To date , more than 200 V . cholerae serogroups have been identified where most serogroups are non-pathogenic . Only isolates of serogroup O1 ( consisting of two biotypes known as ‘classical’ and ‘El Tor’ and the serotypes Ogawa and Inaba ) and O139 have been reported to be pathogenic and cause cholera epidemics and pandemics . However , in the last decade no epidemics due to V . cholerae O139 have been reported and only sporadic clinical cases have been observed [5] . Stool culture remains the reference method for laboratory surveillance of cholera though the sensitivity of direct stool culture is not 100% and depends on the concentration of V . cholerae ( 106–107 CFU ) in stool specimens [6–10] . Moreover , due to limited facilities in peripheral and field sites , diagnosis is a major hindrance for early detection of cholera in endemic regions or during a cholera epidemic . The routine culture also costs approximately 6–8 USD/case [8] and the procedure involves isolation of the bacteria , routine microbiological and biochemical analyses which is lengthy and requires about 24–72 hours . Additionally , microbiological facilities are often not feasible in remote locations and transport to the closest sufficiently equipped laboratory may add further costs . Various molecular-based techniques have been developed including PCR for the rapid detection of virulence and regulatory genes [11] . Although the specificity of PCR method is relatively high , it requires expensive equipment and technical expertise which may is very often not available in small laboratories or field settings . For the effective control of disease outbreaks , diagnostic methods should be both quick and easy without sacrificing specificity and sensitivity of detection . RDT for cholera could be a potential alternative with advantages such as it is rapid , requires minimum training , easy to use and interpret , can be stored at ambient temperature , reasonably priced and can be deployed widely for early confirmation of cholera outbreaks . One of the most recent cholera RDTs currently available in the market is Crystal VC™ ( Span Diagnostics Ltd , Surat , India ) , a dipstick assay initially developed by the Institut Pasteur which is now being produced by Span Diagnostics ( Surat , Guzarat , India ) . The assay relies on the detection of the lipopolysaccharide ( LPS ) antigen of both V . cholerae O1 and O139 serogroups by monoclonal antibodies based on a one-step vertical-flow immunochromatography principle . Crystal VC has shown sensitivity ranging from 94–100% , and a specificity range of 84–100% [9 , 12–14] . However , the O1 and O139 together in Crystal VC lead to lower specificity . Recently , another RDT named ‘Cholkit’ has been developed by our group . Cholkit is a lateral flow immunochromatography test for the qualitative determination of LPS antigen of only Vibrio cholerae O1 serogroup using monoclonal antibody specific to V . cholerae O1 [15] . The objective of this study was to evaluate the performance of the RDT Cholkit and compare the performance with Crystal VC assay , a commercially available RDT designed to detect V . cholerae O1 and O139 .
The study protocol was approved by the Research Review Committee ( RRC ) and Ethical Review Committee ( ERC ) at the icddr , b . Informed written consent was taken from adult patients , or guardians on behalf of children . Since 2016 , icddr , b has been running a nationwide enteric disease surveillance in collaboration with Institute of Epidemiology Disease Control & Research ( IEDCR ) under the Government of Bangladesh ( GoB ) . The surveillance is being conducted in different districts comprising of 22 sentinel sites ( health facilities ) , covering all 8 divisions across Bangladesh ( Fig 1 ) . Stool samples were obtained from individuals seeking treatment with complaints of acute watery diarrhea . A diarrheal visit was defined as a patient ( age > 2 months ) attending hospital with 3 or more loose or liquid stools in last 24 hours or less than 3 loose/ liquid stools causing dehydration; or at least one bloody loose stool in last 24 hours , as well as ( age < 2 months ) changed stool habit from usual pattern in terms of frequency ( more than usual number of purging ) or nature of stool ( more water than fecal matter ) . Patients presented with acute watery diarrhea were requested to provide a stool sample . Freshly collected stool samples were immediately used for the direct dipstick assay at the sentinel sites . Fecal specimens were concurrently enriched overnight at 37°C in alkaline peptone water ( APW ) ( 1% peptone , 1% NaCl; pH-8 . 5 ) and dipstick assays were performed on the following day . For culture , stool samples were placed into the Cary Blair transport medium and transported to the icddr , b laboratory fortnightly by maintaining the cold chain ( 2−80 C ) . Initially all stool specimens ( n = 381 ) were tested with both Crystal VC and Cholkit simultaneously . After observing similar performance of two RDTs , the kits were separately provided in different field sites . Conventional stool culture was carried out by streaking stool directly on selective TTGA ( taurocholate-tellurite gelatin agar ) plates , and plates were incubated overnight at 37°C . Enrichment was performed in APW overnight at 37°C , followed by plating on TTGA to isolate V . cholerae . Colonies morphologically consistent with V . cholerae were tested for agglutination reaction with monoclonal antibodies specific to V . cholerae serovar O1 ( Ogawa or Inaba ) and O139 . Clinical and sociodemographic data were collected as per the original protocol requirement . Data were checked and then entered into the visual studio version 10 . 0 ( Texas , USA ) . After completing data entry , data were transferred into the SQL server 2008 . Data consistency was checked using SQL query . The primary endpoint was the assessment of the performance of the RDT using microbiological stool culture result as the gold standard for comparison . Sensitivity ( true-positive or TP rate ) was defined as the probability that patients with laboratory-confirmed cholera had a positive RDT . Specificity ( true-negative or TN rate ) was defined as the probability that patients with no laboratory-confirmed cholera had a negative RDT . The positive predictive value ( PPV ) was the probability that patients with a positive RDT had V . cholerae isolated from stool culture . The negative predictive value ( NPV ) was the probability that patients with a negative RDT had no V . cholerae isolated from a stool culture . Proportion test statistics was used for calculating p-values to distinguish the difference between two RDT kits in terms of sensitivity and specificity . Statistical analyses were conducted using STATA version 13 ( USA ) . Sensitivity and specificity were determined based on the comparison of Cholkit and Crystal VC results with the lab culture test and presented as percentages . Along with the percentages of sensitivity and specificity , 95% Clopper-Pearson confidence intervals ( CIs ) were as estimated for better predictions .
From 22 sentinel surveillance sites , a total of 7220 patients who presented with acute watery diarrhea were recruited into the study and analyzed to evaluate the performance of two RDT Kits ( Fig 2 ) . Among them 50% were from <5 years age group and 5% from 5–17 years old and the rest 45% from those who were older . Mean age of the participants was 18 . 75 years , and 55% were male . Among them , 381 stool samples ( both direct and enriched stool ) were tested by using both Cholkit and Crystal VC at the field sites , and the performance was compared with microbiological culture . Amongst 381 stools , V . cholerae was isolated from 25 ( 6 . 6% ) samples by culture . Positivity by Crystal VC with direct and enriched sample was 65/381 ( 17 . 1% ) and 26/381 ( 6 . 8% ) , respectively , whereas Cholkit with direct and enriched sample was positive for 54/381 ( 14 . 2% ) and 37/381 ( 9 . 7% ) respectively ( Table 1 ) . Crystal VC on direct stool showed a sensitivity of 72 . 0% ( 95% CI: 50 . 6% to 87 . 9% ) , specificity of 86 . 8% ( 95% CI: 82 . 8% to 90 . 1% ) and after enrichment the sensitivity and specificity were 68% ( 95% CI: 46 . 5% to 85 . 1% ) and 97 . 5% ( 95% CI: 95 . 3% to 98 . 8% ) respectively . Negative predictive values ( NPV ) of Crystal VC were similar; however , we found different positive predictive values ( PPV ) 27 . 7% and 65 . 4% on direct and enriched stool respectively . Test results on direct sample of Cholkit revealed a sensitivity of 76 . 0% ( 95% CI: 54 . 9% to 90 . 6% ) and specificity 90 . 2% ( 95% CI: 86 . 6% to 93 . 1% ) while enrichment revealed 64% ( 95% CI: 42 . 5% to 82 . 0% ) and 94 . 1% ( 95% CI: 91 . 1% to 96 . 3% ) respectively . The sensitivity and specificity of the RDTs using either direct or enrichment methods , were not found to be different ( p>0 . 05 ) . The PPVs of Cholkit was 35 . 2% and 43 . 2% on fresh and enriched samples , whereas NPVs were similar ( Table 2 ) . A total of 5 , 865 direct stool samples and a subset of 614 enriched stools were tested with Crystal VC . On the other hand , 1 , 355 direct stools and a subset of 424 enriched samples were tested with Cholkit ( Table 3 ) . The sensitivity and specificity of Cholkit with direct stool was 79 . 4% ( 95% CI: 62 . 1% to 91 . 3% ) , and 87 . 4% ( 95% CI: 85 . 5% to 89 . 1% ) respectively , while the sensitivity and specificity of Cholkit with enriched stool was 66 . 7% ( 95% CI: 47 . 2% to 82 . 7% ) , and 94 . 4% ( 95% CI: 91 . 7% to 96 . 5% ) , respectively . PPVs were 13 . 9% on direct stool and 47 . 6% on enriched sample , whereas NPVs showed similar result on both . In contrast , sensitivity and specificity of Crystal VC with direct stool was 72 . 2% ( 95% CI: 64 . 6% to 78 . 9% ) and 77 . 1% ( 95% CI: 75 . 9% to 78 . 2% ) respectively , while the sensitivity and specificity of Crystal VC was respectively 68 . 3% ( 95% CI: 51 . 9% to 81 . 9% ) and 90 . 8% ( 95% CI: 88 . 1% to 92 . 9% ) with enriched stool . The results of NPVs are almost similar , while PPVs are 8 . 2% on fresh stool and 34 . 6% on enriched sample ( Table 4 ) .
Our study demonstrated that the RDT Cholkit , locally developed in Bangladesh is comparable to Crystal VC in terms of sensitivity and specificity and can be used for monitoring cholera hotspots and epidemics . The kit will also be relatively cheaper than the commercially available RDT in the market . The Cholkit has only monoclonal antibody that detects V . cholerae O1 . In a cholera endemic region like Bangladesh where V . cholerae O1 is the only prevalent strain , it is more efficient to have a test available for the O1 serogroup only . The study demonstrates the feasibility of using RDTs for monitoring cholera in resource poor as well as in hard to reach areas . This analysis also demonstrates the presence of cholera hotspots in different parts of Bangladesh in the surveillance carried out in 8 divisions of the country . However , confirmation of the RDT tests with bacteriological culture was also carried out to further strengthen and confirm the results . The information obtained from this study will be useful for planning preventive measures for eliminating cholera in Bangladesh which is an agenda for the global road map of ending cholera by 2030 . In conclusion , our data shows that cholera RDTs will be helpful in predicting the population based incidence of cholera in the country and this information can also be utilized in other countries endemic or having epidemic potentials . | Cholera still remains a burning public health issue in the developing world . Microbiological culture is the gold standard method for cholera diagnosis . However , it requires well equipped laboratories and 24–72 hours’ time for the isolation of pathogens , which may not be feasible for hard to reach areas and during epidemics or seasonal outbreaks . For the effective control of disease outbreaks , detection methods should be both quick and easy without sacrificing specificity and sensitivity . Rapid diagnostic test for cholera could be a potential alternative for early detection of the disease . Addressing this issue in our study , we report the performance of a rapid diagnostic test ( RDT ) , Cholkit for the diagnosis of cholera cases using stool obtained in field settings and the assessment of its performance with those of microbial culture and Crystal VC assay , a commercially available dipstick . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [
"medicine",
"and",
"health",
"sciences",
"immune",
"physiology",
"pathology",
"and",
"laboratory",
"medicine",
"microbial",
"cultures",
"pathogens",
"vibrio",
"biological",
"cultures",
"tropical",
"diseases",
"microbiology",
"geographical",
"locations",
"immunology",
"diarrhea",
"bacterial",
"diseases",
"crystals",
"vibrio",
"cholerae",
"signs",
"and",
"symptoms",
"materials",
"science",
"gastroenterology",
"and",
"hepatology",
"neglected",
"tropical",
"diseases",
"antibodies",
"bacteria",
"bacterial",
"pathogens",
"bangladesh",
"research",
"and",
"analysis",
"methods",
"monoclonal",
"antibodies",
"immune",
"system",
"proteins",
"infectious",
"diseases",
"cholera",
"proteins",
"medical",
"microbiology",
"epidemiology",
"microbial",
"pathogens",
"people",
"and",
"places",
"biochemistry",
"diagnostic",
"medicine",
"asia",
"physiology",
"biology",
"and",
"life",
"sciences",
"physical",
"sciences",
"materials",
"organisms"
] | 2019 | Field evaluation of a locally produced rapid diagnostic test for early detection of cholera in Bangladesh |
Calcium responses have been observed as spikes of the whole-cell calcium concentration in numerous cell types and are essential for translating extracellular stimuli into cellular responses . While there are several suggestions for how this encoding is achieved , we still lack a comprehensive theory . To achieve this goal it is necessary to reliably predict the temporal evolution of calcium spike sequences for a given stimulus . Here , we propose a modelling framework that allows us to quantitatively describe the timing of calcium spikes . Using a Bayesian approach , we show that Gaussian processes model calcium spike rates with high fidelity and perform better than standard tools such as peri-stimulus time histograms and kernel smoothing . We employ our modelling concept to analyse calcium spike sequences from dynamically-stimulated HEK293T cells . Under these conditions , different cells often experience diverse stimulus time courses , which is a situation likely to occur in vivo . This single cell variability and the concomitant small number of calcium spikes per cell pose a significant modelling challenge , but we demonstrate that Gaussian processes can successfully describe calcium spike rates in these circumstances . Our results therefore pave the way towards a statistical description of heterogeneous calcium oscillations in a dynamic environment .
Transient changes in the intracellular calcium ( Ca2+ ) concentration have long been associated with the activation of plasma membrane receptors [1] . Since the seminal work by Woods et al . [2] linking the frequency of cytosolic Ca2+ oscillations in hepatocytes to the concentration of various hormones , both experimental and theoretical studies have provided compelling evidence for encoding extracellular stimuli into intracellular Ca2+ oscillations [3–14] . In whole-cell recordings , Ca2+ oscillations are usually observed as sequences of spikes of the intracellular Ca2+ concentration . A prominent feature of Ca2+ spike sequences is that they are random . Ca2+ spikes only occur with some probability that generally changes over time . For example , there are distributions of inter-spike intervals ( ISIs ) for agonist induced Ca2+ oscillations in HEK293 cells and spontaneous Ca2+ oscillations in astrocytes , microglia and PLA cells instead of a single value [15] . When astrocytes are transiently stimulated with ATP three times , with a recovery period between stimuli , the observed Ca2+ spikes display any number of response patterns , from no spikes to three [16] . To elucidate the principles that govern the translation of extracellular cues into changes of the intracellular Ca2+ concentration therefore requires faithfully capturing the stochasticity in Ca2+ spike generation . To date , the main modelling approach to investigate stochastic Ca2+ spike generation has been based on numerical solutions of differential equations , both ordinary and partial [11 , 12 , 15 , 17–30] . In these studies , the randomness of Ca2+ spikes results from the stochastic behaviour of Ca2+ releasing channels , such as the inositol-1 , 4 , 5-trisphosphate ( InsP3 ) receptor ( InsP3R ) , and their interactions . The random dynamics of the InsP3R is then either described by coupling a Markov chain for the InsP3R to the differential equations , or by assuming a Langevin-type equation . All of these approaches require detailed models of the InsP3R ( with often a considerable number of rate constants ) , and other assumptions such as the number of InsP3Rs per cluster and the spatial distribution of InsP3Rs , Ca2+ pumps and Ca2+ buffers in the case of partial differential equations . However , such mechanistic detail , which has been instrumental in advancing our understanding of Ca2+ spikes , often comes at considerable computational costs . It might be therefore desirable to change the perspective from the mechanistic bottom-up approach to a top-down view , in which cellular Ca2+ spikes are described directly . The mathematical concept that has proven particularly useful for this endeavour is the theory of point processes [31] . Indeed , in [15] a time-dependent conditional rate for the generation of a Ca2+ spike was introduced , and its two parameters ( a time scale and an amplitude ) were determined from experiments on four different cell types . Subsequent work [32 , 33] demonstrated in more detail that for constant stimulation , the time scale was cell type specific , while different cells of the same type could be distinguished by their amplitude . These modelling approaches have not confronted the issue of the dynamic nature of cell stimulation that occurs under physiological conditions . Cells in vivo experience a complex and dynamically-changing environment , where signals frequently arrive in a time-varying manner—such as transient release of neurotransmitters or oscillations in the concentration of circulating hormones . Diffusion of messengers through tissue ( such as away from a blood vessel ) also introduce spatial variation in signal strength , meaning that cell populations encounter a complex spatiotemporal pattern of stimulation . An efficient mechanism for modelling Ca2+ responses in such a heterogeneous population has not yet been devised . Our approach offers a solution to this issue and is based on combining point processes with Bayesian inference . Bayesian concepts have been used with great success across various disciplines ( see e . g . [34] for an overview ) . One advantage of utilising Bayesian ideas is that model parameters can be effectively constrained by observed data and can be estimated in a controlled fashion . For example , each set of parameter values comes with its own probability that informs us about how likely this set represents the observed data . In the past , the combination of Bayesian inference and point processes has been successfully applied to action potential spike trains in neurons [35 , 36] , but to our knowledge , this is the first time that Ca2+ spike sequences have been analysed in this way . While we can draw on these previous results , the substantial differences between action potential spike trains and Ca2+ spike sequences ( e . g . the number of spikes per train or the time scales of spikes ) have required significant attention . A particular characteristic of Ca2+ spikes is that their generation depends on the cellular Ca2+ spike history and hence is non-Markovian . This results from both the stochastic nature of Ca2+ spike formation as well as the dynamic variation in the cellular signalling micro-environment . To model such history dependence , we follow a concept introduced in [37] , which effectively turns a non-Markovian Ca2+ spike sequence into a Markovian one . This is achieved with the help of a so-called intensity function x ( t ) , for which we provide a definition and more details in the Materials and Methods section . Since the intensity function is directly inferred from individual Ca2+ spike sequences , it is specific to each cell . This results from two facts . Firstly , Ca2+ spikes are shaped by the cellular composition of the Ca2+ signalling toolkit , i . e . the expression levels and spatial arrangement of Ca2+ channels , pumps , transporters and buffers . Secondly , each cell experiences a different signalling micro-environment as illustrated above . Autocrine and paracrine signalling modify the original signal further . In addition , any time-dependence of x ( t ) that originates from a time varying signal is compounded by dynamic changes of the Ca2+ signalling apparatus , such as Ca2+ store refilling , adaptation or desensitisation . In the present paper we demonstrate how to estimate x ( t ) in the presence of such single cell variability . This will allow us to make progress in two main directions . Firstly , one of the most highly discussed questions in computational biology is concerned with scaling dynamics from single cells to tissue . Taking into account our current understanding of intracellular Ca2+ signalling , this will require the simulation of spatially extended single cells driven by fluctuating Ca2+ releasing channels . The computational costs for such studies are extraordinary , and how to sample the associated high dimensional parameter space remains an open challenge . On the other hand , generating Ca2+ spike sequences from an intensity function is computationally cheap and hence puts researchers in an advantageous position to obtain high quality statistical insights into tissue level dynamics . Secondly , while our approach is statistical , the Ca2+ ISI distribution that we determine as part of estimating x ( t ) has direct mechanistic interpretations . This may further our understanding of how exactly Ca2+ spikes emerge at the single cell level from the orchestrated action of the Ca2+ signalling toolkit . For instance , one line of argument suggests that Ca2+ spikes result from the co-ordinated interplay of a certain number of Ca2+ puffs . Our statistical analysis may provide quantitative estimates for this .
A main motivation for our study is the need to model the heterogeneity of Ca2+ responses observed in cell populations in a computationally efficient manner . To illustrate the problem , the results in Fig 1 show HEK293T cells challenged with a solution containing 100 μM carbachol . Fig 1B illustrates the large variability of Ca2+ responses observed between different cells even when the stimulus is stationary . While some cells only detect the onset of the stimulus ( third trace ) , other cells exhibit Ca2+ spikes during the entire stimulation period . However , the spike characteristics vary significantly . Some cells show irregular Ca2+ spikes ( first trace ) , while other cells settle into an almost regular pattern ( fourth trace ) . It is also worth noting that the frequency of Ca2+ spikes spans a considerable range ( cf . first and second trace ) and that all cells display a decrease in peak amplitude , which is indicative of adaptation . Fig 1C and S1 Video provide further evidence for the large heterogeneity in the timing of Ca2+ spikes for a constant stimulus . The first step in developing a top-down model that reproduces this heterogeneity is to determine the probability distribution that most accurately describes recorded ISIs . Throughout this study , Ca2+ spikes are treated as all-or-nothing events and any information on the width or amplitude of a Ca2+ spike is excluded . The reason why we can introduce an ISI distribution and hence treat successive ISIs as independent—instead of trying to fit a time-dependent ISI distribution—comes from the introduction of the intensity function x ( t ) . We tested three possible candidate ISI statistics ( see Materials and methods for details ) : an inhomogeneous Poisson ( IP ) , an inhomogeneous Gamma ( IG ) and an inhomogeneous inverse Gaussian ( IIG ) distribution . The IP distribution often serves as starting point for analysing spiking behaviour as it is the most basic statistical distribution . While the parsimony of the IP distribution has undoubtedly helped in establishing a large body of mathematical results , real world data often exhibit more complex statistics . The IG distribution provides a natural extension of the IP distribution , in that it contains the IP distribution as a special case: putting γ = 1 in ( 5 ) recovers ( 10 ) . The shape parameter γ endows the IG distribution with more flexibility , which has proven fruitful in numerous applications . Conceptually , spikes in general and Ca2+ spikes in particular have been described as first passage events [38] . One of the most fundamental models , which contains positive drift and random motion only , gives rise to the IIG distribution . To ascertain which ISI distribution describes the results in Fig 1 best , we transformed the recorded ISIs using the time rescaling theorem and analysed the results in a Kolmogorov-Smirnov plot ( see Material and methods ) . The Kolmogorov-Smirnov plot allows the visual inspection of whether two probability distributions are the same by plotting their respective cumulative distribution functions against one another . If the two distributions are identical , the cumulative distributions functions coincide and plotting one against the other results in a straight line with slope 1 . Fig 2A shows results for the Kolmogorov-Smirnov plot for the data taken from Fig 1 . Here , u corresponds to the cumulative distribution of the transformed data , which depends on the chosen ISI distribution , while s stems from a theoretical prediction based on the time rescaling theorem ( see Materials and methods ) . If we have identified the correct ISI distribution , data points should cluster around a straight line with a 45° slope . Note that in our analysis , each cell was treated individually and no data amalgamation took place . Results for the IP and IIG clearly deviate from a line with slope 1 , while the data for the IG exhibit much less deviation . This suggests that an IG , but not an IP or IIG , better describes the ISI statistics of Ca2+ spikes . The box plots in Fig 2C provide further quantitative evidence . They demonstrate that the slopes obtained from individual cells are concentrated close to 1 and exhibit a significantly smaller variability for the IG compared to the IP and IIG . To further corroborate these findings , we analysed data from cells exposed to a lower concentration of carbachol ( 10μM ) . Fig 2B and 2D illustrate that again the IG distribution captures the ISI statistics most closely . In addition to Kolmogorov-Smirnov plots , we also tested our data with a quantile-quantile plot ( see S1 Fig and Materials and methods ) . As with the Kolmogorov-Smirnov plot the correct ISI distribution leads to data points that accumulate around the 45° line . Panels A and B in S1 Fig show that this is the case for the IG , but not for the IP and IIG , hence confirming our results from the Kolmogorov-Smirnov plot . The box plots in panels C and D of S1 Fig show that again the slopes obtained from single cells for the IG model are much closer to one and exhibit less variability than those for the IP and IIG . Taken together , these results indicate that the patterns of Ca2+ spikes observed across the population of cells can most accurately be reproduced with an IG distribution . This approach will therefore be used as the basis of our analysis . It is worth noting that when testing the different ISI distributions , we also obtain an estimate for the intensity function x ( t ) . As illustrated by Eq ( 11 ) , the probability for a specific ISI depends on x ( t ) . Therefore , all ISI distributions in this study have to be understood as being conditioned on x ( t ) . In the next section , we show how to estimate x ( t ) from measured Ca2+ spike sequences . In addition to the ISI distribution , we also need to know the intensity function x ( t ) to fully describe Ca2+ spike sequences . In the following , we will use three different approaches to estimate x ( t ) : peri-stimulus time histograms ( PSTHs ) , kernel smoothing ( KS ) , and Gaussian processes ( GPs ) combined with Bayesian inference . An illustration of a GP is shown in Fig 3 . In contrast to a deterministic curve , a GP generates infinitely many curves ( 3 possible candidates are shown ) , and the statistics of these curves is the organising principle . At each time point , the values of a GP are Gaussian distributed , and the mean and standard deviation can change over time . To guarantee a controlled comparison between PSTHs , KS and GPs , we will fix an intensity function , generate surrogate Ca2+ spike sequences from it and then estimate how well the above methods recapture the original intensity function ( see Materials and methods for details ) . Before proceeding it is worth noting that the intensity function that we need to estimate is identical to the Ca2+ spike rate that we find from using either PSTHs or KS [36] . In other words , if we can obtain a high quality estimate for the Ca2+ spike rate from either PSTHs or KS , we have a very good estimate for x ( t ) . However , this usually requires a large number of Ca2+ spike sequences , which is an issue that we will address below . Since we can identify the Ca2+ spike rate with the intensity function , we will use both terms interchangeably . Note , however , that this Ca2+ spike rate is different from the conditional intensity function defined in Eq ( 15 ) , which is often used in generating spike trains . A major objective of this model is to reproduce Ca2+ signalling patterns during complex stimulation conditions . For example , physiological patterns of hormone or neurotransmitter release are time-varying , rather than the step-changes used in typical experiments ( such as Fig 1 ) . We therefore tested the performance of candidates for x ( t ) for reproducing dynamically-changing signals—specifically sinusoidal oscillations . As a first choice , we considered a regularly oscillating intensity function xdet ( t ) = 0 . 5 cos ( t ) + 0 . 5 cos ( 0 . 5t ) + 1 . Fig 4A shows Ca2+ spike sequences generated from xdet , and Fig 4C reveals that both PSTH and KS capture the original intensity function very well . By using a specific functional form for x ( t ) as in xdet ( t ) , we make strong assumptions about the intensity function . A more flexible and versatile approach is based on GPs . In Fig 4B we plot Ca2+ spike sequences generated from one xGP candidate , while Fig 4D depicts the estimation of xGP from a PSTH and KS . As with xdet we find very good agreement between the original and estimated Ca2+ spike rate . This strategy supposes that all Ca2+ spike sequences can be combined into a single large dataset , but in a physiological context this will not always be true . To be a more useful tool , the model should be able to simulate the diversity of responses expected in a more complex environment , where the stimulus varies in both space and time . Under these circumstances , the number of cells that receive an equivalent stimulus would be more limited , and so it is important to assess how the number of spike sequences available for parameter estimation affects the accuracy of predicting the Ca2+ spike rate . Consequently , we randomly picked groups of 1 , 2 , 4 and 7 Ca2+ spike sequences and computed the Ca2+ spike rate based on a GP and KS . In Fig 5 we show results for xGP . Since the Ca2+ spike rate estimation obtained from KS depends on the bandwidths σ ( see Eq ( 17 ) ) , we employed different σ values . For a single Ca2+ spike sequence ( Fig 5A ) the estimated Ca2+ spike rates differ visibly from the theoretical one , and the smallest bandwidth leads to spurious oscillations . As we increase the number of Ca2+ spike sequences the estimated Ca2+ spike rates capture the true Ca2+ spike rate more faithfully . The light blue area in each panel delineates the 95% confidence interval , which we obtain as a by-product from the GP optimisation . Overall , all estimates lie within this confidence interval except the one for σ ^ in Fig 5A ( see Materials and methods for the definition of σ ^ ) . We obtain similar results for xdet ( t ) as illustrated with S2 Fig . We quantified the accuracy of predicting the Ca2+ spike rate by computing the normalised L2 norm of the difference between the known and estimated Ca2+ spike rate ( see Materials and methods ) . Fig 6 shows that for a given method , the L2 norm decreases as we increase the number of Ca2+ spike sequences , which corresponds to better predictions . When we fix the number of Ca2+ spike sequences , GPs yield a better estimate . The improvement is particularly evident when comparing the Ca2+ spike rate estimation based on σ ^ at small numbers of Ca2+ spike sequences . We further tested that our results did not depend on the particular choice of Ca2+ spike sequences nor on the details of the surrogate generator . For the latter , we compared three different approaches: inverse sampling , a Bernoulli process and time rescaling . We generated a number of Ca2+ spike sequences with each method and then estimated the Ca2+ spike rate using the same methods as in Fig 6 , i . e . KS with different bandwidths and GPs . S3 Fig shows box plots of the normalised L2 norm between the estimated and the true Ca2+ spike rate xdet . We find that for all three methods , the normalised L2 norm is generally smallest for GP estimates . To test the statistical significance of this result , we computed the corresponding p-values as shown in S1 Table using the non-parametric Mann-Whitney test . Based on the common assumption that a finding is statistically significant if p < 0 . 05 , estimates using GPs perform statistically better than KS since the largest p value was 0 . 0375 . We repeated the analysis for xGP and report the normalised L2 norm in S4 Fig . While the GP performs clearly better than KS with bandwidths of σ ^ and 35 , the distributional results for bandwidths of 52 and 70 look similar to those of the GP . This is also confirmed by the p-values shown in S1 Table , where some exceed the threshold of 0 . 05 . This indicates that KS can approach the performance of GPs . However , since there are no a priori estimates for this optimal bandwidths for a given scenario , GPs provide the more robust estimation method . The results so far provide strong evidence that an intensity function derived from a GP allows accurate prediction of Ca2+ spike patterns even when the estimate is based on small numbers of Ca2+ spike sequences . We next applied our Bayesian approach to a more complex experimental system , designed to reproduce some of the stimulus heterogeneity expected in vivo . We used a microfluidics chamber to deliver sinusoidal changes in carbachol concentration to HEK293T cells ( see Materials and methods ) . The concentration of carbachol varied in both space and time . Fig 7A shows a complex concentration surface throughout the chamber at a fixed time point and illustrates how the sharp interface between high and low agonist concentration on the left side of the chamber widens as the flow progresses through the chamber . In Fig 7B we plot agonist concentration time courses sampled at four positions along the transverse direction of the microfluidics chamber , which demonstrates the stimulation heterogeneity that cells experience depending on their position within the chamber . We also include the corresponding Ca2+ spike sequences , which again display significant variability . The goal of this experiment was to generate an environment in which a population of cells is exposed to agonist in a manner that varies with the cells’ distance from the stimulus source and with a dynamic mechanism of delivery ( by analogy to a circulating hormone diffusing from a blood vessel to underlying cells , for example ) . This scenario is the context in which a Bayesian framework is most useful , as it can predict spiking Ca2+ responses to a complex but physiologically meaningful stimulus profile . In Fig 7C–7E , we show clustered stimulus curves , the corresponding Ca2+ spike sequences and the estimated intensity functions , respectively . We grouped cells that experienced similar agonist concentration profiles to allow for a meaningful comparison of the resultant intensity functions . To determine how similar stimulus traces are , we computed the weights of the three leading principal components ( see Materials and methods ) . Results are shown in Fig 8 , where we employed k-means [39] to detect possible clusters . Data points that belong to the same group are plotted in the same colour , and these colours correspond to those used in Fig 7C–7E . Overall , 4 distinctive groups represented the data best , containing 10 , 13 , 13 and 3 cells , respectively . It is worth noting that while clustering was performed on the stimuli time courses the Ca2+ spike sequences show a consistent pattern in that they are generally more similar within a given group than between groups . The black lines in Fig 7C and 7E denote the mean stimulus and mean intensity function , respectively . We observed that the intensity functions broadly mirror the global behaviour of the stimulus . The mean intensity function for all responses showed a single peak with a similar time course to the stimulus . The amplitude of stimulus and intensity function are also well matched . However , there are also observable differences . For example , individual cells showed intensity functions with a more complex time course than the mean ( such as multiple peaks ) . Furthermore , while the mean stimulus was symmetrical , the mean intensity functions could exhibit notable asymmetries . For example , for weak stimuli , the rising phase of the intensity function may be markedly slower than the falling phase ( Fig 7C and 7E; yellow and green traces ) . This most likely reflects the excitable character of intracellular Ca2+ signalling [30 , 40] . For weaker stimulation , it takes longer to reach the threshold for generating a Ca2+ spike , hence the intensity function grows more slowly . The quicker decrease results from the stimulus dropping below the Ca2+ spike generating threshold quickly after reaching its maximum , hence prohibiting further Ca2+ spikes . The faster increase of the intensity function for stronger stimuli ( grey traces ) lends further support for this interpretation , as the Ca2+ spike generating threshold is reached more quickly . To illustrate the value of the mean intensity functions shown in Fig 7E , we generated surrogate Ca2+ spike sequences from them using the IG distribution and plotted them in Fig 9C . For ease of comparison , we also show the measured Ca2+ spike sequences in Fig 9B . We first note that the simulated Ca2+ spike sequences resemble the measured ones . This is also confirmed by the histrograms in Fig 9D , which exhibit large overlaps between the experimental and theoretical Ca2+ spike sequences . To quantify how similar the two histograms are , we computed the histogram distance given by H ( R , S ) = ∑ i min ( R i , S i ) max ( ∑ i R i , ∑ i S i ) , ( 1 ) where Ri and Si denote the histogram count in the ith bin of the recorded and simulated data , respectively . The histogram distance is bounded between 0 and 1 , and the closer it is to 1 the more similar the histograms are . The histograms coincide if H = 1 . We found that H = 0 . 86 , 0 . 94 , 0 . 91 and 0 . 89 ( from top to bottom ) , which confirms our visual inspection that the histograms vary little between recorded and simulated Ca2+ spike sequences . Moreover , Ca2+ spike sequences generated from one intensity function exhibit a certain degree of heterogeneity , which is consistent with our experimental findings . Taken together , these results show that without explicitly including any information about the stimulus into the estimation of the intensity functions , our approach yields intensity functions that reflect the characteristics of the stimulus and that are consistent with experimentally recorded Ca2+ spike sequences .
In this work we have developed a mathematical framework to quantitatively describe the heterogeneous timing of Ca2+ spikes in a cell population subject to time-varying stimulation . At the heart of this new approach is the use of Bayesian inference to determine the most likely intensity function and hence the most likely Ca2+ spike rate for a given stimulus . As part of this estimation process , we found that the statistics of Ca2+ ISIs are best captured by an IG distribution . Importantly , knowledge of the intensity function and the ISI statistics suffices to completely describe Ca2+ spiking . Since generating Ca2+ spike sequences from an ISI distribution and intensity function is computationally significantly cheaper than solving partial differential equations for cellular Ca2+ transport , this approach is ideally suited for numerically studying large numbers of cells . The estimation of inhomogeneous single cell behaviour also puts us in an ideal position to ascertain whether or not there is signal processing at the cell population level . Indeed , numerous examples exist where the average population behaviour is not shared by any cell ( see e . g . [55] ) . These incongruous dynamics also warrant investigations into population invariances , where cell populations respond consistently in the same manner , albeit with completely heterogeneous single cell behaviour [56 , 57] . By reliably estimating single cell Ca2+ dynamics , the present study provides a stepping stone towards answering these questions for intracellular Ca2+ signalling .
We here follow the exposition in [37] for the definition of the intensity function . Assume that Ca2+ spikes occur at times y1 < y2 < … < yN . Let p ( v ) denote the probability density for a general renewal process on v ∈ ( 0 , ∞ ) , i . e . p ( v ) dv is the probability for an event in [v , v + dv] , and subsequent events are independent . For ya > 0 , let y correspond to a time variable on ( ya , ∞ ) and X be a one-to-one mapping X ( y ) = v of ( ya , ∞ ) to ( 0 , ∞ ) . Conservation of probability then entails that p ( y ) = | d v d y | p ( v ) = | X ′ ( y ) | p ( X ( y ) ) . ( 2 ) In other words , the probability density for a Ca2+ spike at yi can be computed from the renewal probability density p if we know the mapping X . A convenient form of X is X ( y ) = ∫ y a y x ( u ) d u , ( 3 ) which satisfies the conditions above and where x is called the intensity function , which is the object that we need to estimate . Eq ( 3 ) can be interpreted as rescaling the original time y such that Ca2+ ISIs become independent and identically distributed in the new time [58] . Given two subsequent Ca2+ spikes times yi−1 and yi in the original time , the ISI in the new time is X ( y i - 1 , y i ) = ∫ y i - 1 y i x ( u ) d u . ( 4 ) Since it is only through the introduction of the intensity function x ( t ) that Ca2+ ISIs become Markov , we introduce the notation p ( yi , yi−1|x ) , which corresponds to the ISI probability density given x ( t ) . Note that formally the conditional ISI probability density is defined as the joint conditional probability density for spikes at yi and yi−1 ( and hence no spike in [yi−1 , yi] ) given an intensity function x ( t ) . We will employ three different choices for the ISI probability density: an inhomogeneous Gamma distribution p ( y i , y i − 1 | x ) = γ x ( y i ) Γ ( γ ) [ γ X ( y i − 1 , y i ) ] γ − 1 e − γ X ( y i − 1 , y i ) , ( 5 ) where γ > 0 denotes the shape parameter and Γ is the Gamma function; an inhomogeneous inverse Gaussian distribution p ( y i , y i − 1 | x ) = x ( y i ) 2 π X 3 ( y i − 1 , y i ) exp { − ( X ( y i − 1 , y i ) − α ) 2 2 α 2 X ( y i − 1 , y i ) } , ( 6 ) where α > 0 is the location parameter; and an IP distribution p ( y i , y i - 1 | x ) = x ( y i ) e - X ( y i - 1 , y i ) . ( 7 ) The time-dependent intensity function x ( t ) is modelled as a Gaussian Process ( GP ) [34 , 59] . A GP is uniquely defined by its mean μ ( t ) and covariance function Σ ( t1 , t2 ) . While there are many possible choices for Σ [34 , 60] , we employ the widely used squared exponential ( SE ) kernel Σ ( t 1 , t 2 ) = σ f 2 e - κ ( t 1 - t 2 ) 2 2 + δ ( t 1 - t 2 ) σ v 2 , ( 8 ) where κ measures the smoothness of the GP and σf controls its variance . The last term allows us to model additional noise sources . We originally included σ v 2 as a hyperparameter in the optimisation . However , we consistently found small values for σ v 2 and hence decided to fix it at a presentative value of σ v 2 = 10 - 4 . We collect the spike times in a sequence of N Ca2+ spikes in a vector y = {y1 , … , yN} . For consistency , we set y0 = 0 . Through the introduction of an intensity function x ( t ) , the joint probability density for a spike sequence y given x ( t ) factorises and reads as [36] p ( y | x ) = p 1 ( y 1 | x ) p T ( T , y N | x ) ∏ i = 2 N p ( y i , y i - 1 | x ) . ( 9 ) Here , p1 ( y1|x ) represents the conditional probability density of finding the first spike at time y1 . We also take into account that the observation time T usually exceeds the last spike time through the term pT ( T , yN|x ) , which denotes the conditional probability that no spike occurs after yN . The statistics for p1 and pT are often based on an inhomogeneous Poisson ( IP ) process , i . e . p 1 ( y 1 | x ) = x ( y 1 ) e - X ( 0 , y 1 ) , p T ( T , y N | x ) = e - X ( y N , T ) , ( 10 ) where X is given by Eq ( 3 ) . For practical purposes , we discretise time with a time step Δ such that T = nΔ [36] . When working with experimental spike trains , we set Δ equal to the inverse of the recording frame rate . A spiking time yi can then be expressed as yi = li Δ for an appropriate l i ∈ N . By setting xi = x ( iΔ ) and using Eq ( 9 ) with e . g Eq ( 5 ) , we obtain the probability density for a spike sequence for the inhomogeneous Gamma distribution as p ( y | x , θ ) = x l 1 e − X ^ 0 , 1 e − X ^ N , n ∏ i = 2 N γ x l i Γ ( γ ) [ γ X ^ i − 1 , i ] γ − 1 e − γ X ^ i − 1 , i , ( 11 ) where X ^ i , j = Δ ∑ k = l i l j x k and l0 = 0 , ln = n . By introducing θ on the left hand side , we make explicit the dependence of the probability density on the hyperparameters θ , which in this case are θ = {γ , κ , σf} . The most probable intensity function x* ( t ) given a spike train y is determined by x* = argmaxx≥0 p ( x|y ) . Under the assumption that the nodal value x* is close to its mean , we have x * ≈ ∫ x θ * p ( θ | y ) d θ = 1 Z ∫ x θ * F ( y , x θ * , θ ) d θ , ( 12 ) where x θ * = argmax x ≥ 0 p ( x | y , θ ) = argmax x ≥ 0 p ( y | x , θ ) p ( x | θ ) . ( 13 ) To evaluate the first integral in Eq ( 12 ) we note that p ( θ | y ) = p ( θ ) p ( y ) ∫ p ( y | x , θ ) p ( x | θ ) d x = F ( y , x θ * , θ ) p ( y ) , ( 14 ) with F ( y , x θ * , θ ) = p ( θ ) p ( y | x θ * , θ ) p ( x θ * | θ ) / | Λ * + Σ - 1 | and Λ * = - L x 2 log p ( y | x θ * , θ ) , where we used Laplace’s approximation for the integral as shown in S1 Appendix . We further introduced the notation L x 2 f to denote the Hessian of f with respect to x ( t ) and Z = ∫ F ( y , x θ * , θ ) d θ = p ( y ) / ( 2 π ) n / 2 . Let q ( t|yk , x ) , t > yk denote the conditional intensity function , i . e . q ( t|yk , x ) dt is the probability for a spike in [t , t + dt] given an intensity function x ( t ) and the last spike at yk . We can express q ( t|yk , x ) in terms of the ISI probability density as [31] q ( t | y k , x ) = p ( t , y k | x ) 1 - ∫ y k t p ( s , y k | x ) d s . ( 15 ) The time rescaling theorem then states that the rescaled ISIs [31 , 47 , 61 , 62] τ k = ∫ y k - 1 y k q ( s | y k - 1 , x ) d s , ( 16 ) are independent and identically distributed exponential random variables with mean one if y is a realisation from a point process with conditional intensity function q ( t|yk , x ) . Suppose there are K rescaled ISIs . For a quantile-quantile plot [46] , we order the τk in ascending order giving rise to the new ISIs τ ˜ n . We then plot the quantiles of the distribution of the τ ˜ n against the quantiles of an exponential distribution with unit rate , which are given by τ ^ n = - ln ( 1 - s n ) with sn = ( n − 0 . 5 ) /K . For the Kolmogorov-Smirnov , plot [62] , we define the random variable u k = 1 - e - τ k and then plot the ordered set of the uk against the cumulative distribution function of the uniform distribution , i . e . F ( x ) = x for 0 ≤ x ≤ 1 , sampled at sn . The Ca2+ spike rate is estimated from m spike sequences via kernel smoothing ( KS ) through [49 , 63] r = 1 m ∑ j = 1 m ∑ i = 1 N j f ( t - y i j , σ ) , ( 17 ) where y i j denotes the ith spike time in the jth Ca2+ spike sequence yj , and Nj is the total number of spikes in yj . The function f represents the kernel , and we chose a Gaussian of the form f ( t , σ ) = 1 2 π σ 2 exp ( - t 2 2 σ 2 ) . ( 18 ) The parameter σ is referred to as the bandwidth of the kernel . In case we work with a large number of independent Ca2+ spike sequences yj , we can use an optimal bandwith [49 , 50] . To evaluate how well a given method ( e . g . Bayesian inference or KS ) approximates the true Ca2+ spike rate used to generate surrogate data , we evaluated the normalised L2 norm as L 2 = [ ∫ 0 t ( r ^ ( t ) − r ˜ ( t ) ) 2 d t ] 1 / 2 [ ∫ 0 t r ˜ ( t ) d t ] − 1 , ( 19 ) where r ˜ and r ^ denote the known and estimated Ca2+ spike rate , respectively . We arrange the stimuli experienced by individual cells in a matrix X such that each row corresponds to a single stimulus time course . We then compute the singular value decomposition of X , i . e . X = UΣVt , where t denotes transposition . The columns of V correspond to the eigenvectors of XtX , and Σ is a diagonal matrix that holds the singular values of X . The weights of the principal components of the stimuli time courses are the rows of XV = UΣ . The k-means algorithm requires the number k of clusters as input and then determines the members of each cluster by minimising the error function [64] E = ∑ i = 1 k ∑ x ∈ C i ‖ x - μ i ‖ 2 . ( 20 ) Here , x are the data points , C1 , … , Ck are the k disjoint clusters and μi is the centroid of the ith cluster . We varied k and visually inspected the clustering . For consistency , we also clustered the data using other algorithms such as mean shift , spectral clustering and density-based spatial clustering of applications with noise . While there were minor differences between the suggested clusters , the overall clustering structure remained the same . To see which is the most likely ISI statistics , we apply the following protocol to every single cell from the experiment shown in Fig 1: To test the performance of Ca2+ spike rate estimation , we generated surrogate data from an IG for the two different intensity functions xdet ( t ) and xGP ( t ) using inverse sampling , a Bernoulli process based on the conditional intensity function in Eq ( 15 ) and time rescaling [47 , 65 , 66] . A key factor in estimating Ca2+ spike rates from PSTHs and KS is the choice of a bin width and bandwidth , respectively . For a large number of Ca2+ spike sequences , optimal estimates exist [49–51] , and we use them for Fig 4 . In case of only a few Ca2+ spike sequences with a small number of spikes per sequence , as in Fig 5 , no estimates for a bin width or bandwidth exist . We therefore employed a bandwidth that was approximately equal to the optimal bandwidth determined in Fig 4 as well as bandwidths 1 . 5 and 2 times larger than this . In addition , we used the same formal expression as for the optimal value , which resulted in the bandwidth σ ^ . Note that σ ^ differs from the optimal bandwidth in Fig 4 , since it explicitly depends on the number of Ca2+ spike sequences . A bespoke perfusion system connected to a 3-port microfluidics device [67] was used to expose cultured HEK293T cells to varying concentrations of the muscarinic receptor agonist , carbachol . The HEK293T cell line was a gift from Dr N . Holliday , University of Nottingham , that had been frozen after passage 28 of the original stock . After thawing , cells were used for up to a further ten passages . Cells were seeded at a density of 105 cells/ml in the central micro-channel of the microfluidic devices , in DMEM D6429 growth media ( Invitrogen , Paisley , UK ) containing 10% fetal calf serum . Cells were loaded inside the microchannels with 1 μM of the Ca2+ indicator Fluo5F-AM for 30 min , followed by washout with imaging buffer ( 135 mM NaCl , 3 mM KCl , 10 mM HEPES , 15 mM D-glucose , 2 mM MgSO4 and 2 mM CaCl2 ) for at least a further 30 min . To stimulate the cells , the flow rates of two inlet channels into the microchannel were varied , allowing the interface between the two solutions to be shifted laterally across the chamber . One inlet stream contained the agonist ( 100 μM carbachol ) and Alexa Fluor 594 ( AF594 , 2 nM; to allow monitoring of agonist concentration in proportion to AF594 fluorescence ) . The second inlet contained buffer alone . The interface formed between the two solutions due to laminar flow was shifted across the width of the microchannel by controlled changes in the fractional flow rates for each stream , with total flow being constant . In combination with the shifting interface position , the concentration gradient formed by diffusional collapse of the interface as the co-flow progresses through the channel length results in a spatiotemporal gradient in agonist concentration throughout the channel . This method enables the exposure of cells to pre-defined , time-varying changes in agonist concentration , from simple step-changes to complex waveforms . During dynamic stimulation with agonist , AF594 and Fluo-5F AM indicators were excited sequentially ( 100 ms exposure , 1 Hz frame rate ) using a pE2 LED system ( excitation peaks 470 nm and 565 nm; CoolLED , Andover UK ) . Emission was detected at 535 ± 50 nm and 565 ± 20 nm with an ORCA-R2 camera ( Hamamatsu , Welwyn Garden City , UK ) . A time-series analyser plugin to ImageJ ( Wayne Rasband , National Institutes of Health , Bethesda , MD , available at http://rsb . info . nih . gov/ij ) was used to manually define circular regions of interest ( ROI ) centred on each cell . Mean Fluo-5F emission intensity of pixels falling within each ROI was quantified and expressed as the ratio of fluorescence at time t divided by mean intensity from a 25 s window prior to the first increase in stimulus concentration ( F/F0 ) . The baseline window is selected as the window with minimum standard deviation from sliding 25 s windows taken from 0 to 120 s ( before increase in stimulus concentration ) . Fluorescence of AF594 was quantified as the mean fluorescence intensity of pixels falling within each ROI being quantified; therefore each cell has a Ca2+ response measure and an associated stimulation profile . | Upon stimulation a large number of cell types respond with transient increases of the intracellular calcium concentration , which often take the form of repetitive spikes . It is therefore believed that calcium spikes play a central role in cellular signal transduction . A critical feature of these calcium spikes is that they occur randomly , which raises the question of how we can predict the timing of calcium spikes . We here show that by using Bayesian ideas and concepts from stochastic processes , we can quantitatively compute the calcium spike rate for a given stimulus . Our analysis also demonstrates that traditional methods for spike rate estimation perform less favourably compared to a Bayesian approach when small numbers of cells are investigated . To test our methodology under conditions that closely mimic those experienced in vivo we challenged cells with agonist concentrations that vary both in space and time . We find that cells that experience similar stimulus profiles are described by similar calcium spike rates . This suggests that calcium spike rates may constitute a quantitative description of whole-cell calcium spiking that reflects both the randomness and the spatiotemporal organisation of the calcium signalling machinery . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] | [
"carbachol",
"action",
"potentials",
"medicine",
"and",
"health",
"sciences",
"engineering",
"and",
"technology",
"paracrine",
"signaling",
"drugs",
"membrane",
"potential",
"electrophysiology",
"neuroscience",
"endocrine",
"physiology",
"probability",
"distribution",
"mathematics",
"pharmacology",
"statistical",
"distributions",
"fluidics",
"microfluidics",
"autocrine",
"signaling",
"probability",
"density",
"endocrinology",
"probability",
"theory",
"signal",
"transduction",
"cell",
"biology",
"physiology",
"biology",
"and",
"life",
"sciences",
"physical",
"sciences",
"cell",
"signaling",
"neurophysiology"
] | 2017 | A Bayesian approach to modelling heterogeneous calcium responses in cell populations |
Dengue is a major public health problem in tropical and subtropical regions; however , under-reporting of cases to national surveillance systems hinders accurate knowledge of disease burden and costs . Laboratory-confirmed dengue cases identified through the Nicaraguan Pediatric Dengue Cohort Study ( PDCS ) were compared to those reported from other health facilities in Managua to the National Epidemiologic Surveillance ( NES ) program of the Nicaraguan Ministry of Health . Compared to reporting among similar pediatric populations in Managua , the PDCS identified 14 to 28 ( average 21 . 3 ) times more dengue cases each year per 100 , 000 persons than were reported to the NES . Applying these annual expansion factors to national-level data , we estimate that the incidence of confirmed pediatric dengue throughout Nicaragua ranged from 300 to 1000 cases per 100 , 000 persons . We have estimated a much higher incidence of dengue than reported by the Ministry of Health . A country-specific expansion factor for dengue that allows for a more accurate estimate of incidence may aid governments and other institutions calculating disease burden , costs , resource needs for prevention and treatment , and the economic benefits of drug and vaccine development .
Dengue is a mosquito-borne illness that is a major public health problem in tropical and subtropical regions , causing millions of cases annually [1] . However , under-reporting of dengue cases to national surveillance systems hinders accurate local , regional and global calculations of disease burden and costs , which in turn impact appropriate resource allocation and availability of reliable data for vaccine and drug development . Under-reporting has been attributed to a range of factors , including misdiagnosis , limited laboratory capabilities , poor application of the World Health Organization ( WHO ) case definition , and limitations of national reporting systems , among others [2] , [3] . Furthermore , scarce data on inapparent dengue virus ( DENV ) infections limits estimates of risk for more severe secondary infections . To address the lack of a country-specific estimate of disease burden based on surveillance data , we calculated the difference in dengue case capture rates between a pediatric dengue cohort study ( PDCS ) and the Ministry of Health dengue surveillance program ( “expansion factor” ) in Managua , Nicaragua , which can be applied to national statistics for more accurate estimations of dengue burden .
The PDCS is a community-based , prospective cohort study that was initiated in 2004 in a low-to-middle-income district of Managua and is based in a municipal clinic , the Health Center Sócrates Flores Vivas ( HCSFV ) , which is the principal source of health care for the district's population . The study captured possible dengue cases through “enhanced” passive surveillance by study physicians and nurses at the HCSFV and periodic home visits for follow-up and monitoring of study compliance [4] . Cohort participants , initially aged 2–9 years old in 2004 , were followed closely for all illnesses , and children who presented with fever were screened for signs and symptoms of dengue . Those who met WHO criteria for suspected dengue ( acute febrile illness with two or more of the following symptoms or signs: headache , retro-orbital pain , myalgia , arthralgia , rash , hemorrhagic manifestations , leukopenia , or platelets ≤150 , 000/mm3 ) as well as those with undifferentiated fever were evaluated for acute DENV infection . A dengue case was considered laboratory-confirmed when 1 ) DENV was isolated , 2 ) DENV RNA was demonstrated by reverse-transcriptase polymerase chain reaction ( RT-PCR ) , 3 ) seroconversion was observed with paired acute and convalescent phase sera by IgM capture ELISA or Inhibition ELISA [5] , [6] , or 4 ) a ≥4-fold increase in antibody titer in paired acute and convalescent sera was observed by Inhibition ELISA [7] . Overall , 118 ( 74% ) of the confirmed DENV infections in the PDCS complied with the WHO case definition ( 15 ( 88% ) , 44 ( 68% ) , 8 ( 62% ) , and 51 ( 80% ) in years 2004–5 , 2005–6 , 2006–7 , 2008–9 , respectively ) , while 41 ( 26% ) had undifferentiated fever ( 2 ( 12% ) , 21 ( 32% ) , 5 ( 38% ) , and 13 ( 20% ) in years 2004–5 , 2005–6 , 2006–7 , 2007–8 , respectively ) ( 8 ) . There were 2 , 2 , 1 , and 16 hospitalized cases in the cohort in 2004–5 , 2005–6 , 2006–7 , 2007–8 , respectively , all of which were compliant with the WHO case definition . Inapparent DENV infections , presented here separately from symptomatic cases , were identified through serological testing of paired annual blood draws from healthy subjects , which were collected every year in July . Specifically , children with paired annual serum samples demonstrating seroconversion or ≥4-fold increase in DENV-specific antibody titer , but who had not presented to the HCSFV with acute DENV illness , were considered to have experienced inapparent DENV infections [4] , [8] . All confirmatory testing was conducted at the National Virology Laboratory ( NVL ) of the Nicaraguan Ministry of Health , the same laboratory responsible for confirmatory testing of routine , non-PDCS suspected dengue cases in Managua ( see below ) . This study was approved by the Institutional Review Boards of the Nicaraguan Ministry of Health; the study hospital , Hospital Infantil Manuel de Jesús Rivera ( HIMJR ) ; the University of California , Berkeley; and the International Vaccine Institute . Parents or legal guardians of all subjects provided written informed consent , and subjects over 5 years of age gave verbal assent . Between 2004 and 2008 , a total of 4 , 742 children participated in the PDCS , with a yearly active cohort of 3 , 693-3 , 795 children with equal distribution by gender , age and neighborhood . Compliance and participation were consistently high [4] . Over 90% of participants sought medical care at the HCSFV at least once during the study period , and loss to follow-up – primarily due to moving outside of the study district – was low ( 4 . 3–7 . 1% annually ) . Possible dengue cases presented early to the HCSFV ( 94% presenting within the first 72 hours since symptom onset ) and 94% provided a convalescent sample [4] . The National Epidemiologic Surveillance program ( NES ) collects information on suspected and confirmed dengue cases at the primary and secondary care level . Physicians in both the public and private sectors are required to report suspected cases meeting WHO criteria ( see definition above ) via a standardized MOH reportable disease form and to send blood samples ( ≥5 days since onset of fever ) to the NVL at the MOH for testing . All febrile cases presenting at public facilities are reported , but only for those meeting the WHO definition is a suspected dengue report filled out and sample collected . Approximately 60% of suspected dengue cases in Managua have blood samples sent to the NVL ( A . Balmaseda , unpublished data ) . A suspected case is considered positive ( referred to here as a “confirmed” case ) when the sample tests positive for DENV-specific IgM antibodies via IgM capture ELISA or , in rare cases , if the sample is obtained during the first 5 days after symptom onset and DENV RNA is detected . Data on confirmed and suspected dengue cases by epidemiologic week among children 2–14 years old ( and children 1–14 years old in 2004–5 , before the NES began reporting 1 year-olds separately ) were obtained from the NES for the years 2004–8 , corresponding to the first four years of the PDCS . NES data include both confirmed and suspected cases reported by primary and secondary health facilities in Managua ( primarily public , as private facilities do not comply as frequently with reporting guidelines ) , and are presented here by year , age group , and the health center corresponding to the district of residence of the child . The population of 1–14 year-olds ( 2004 ) and 2–14 year-olds ( 2005–8 ) in each health center's district was also obtained from the NES . These data were used to calculate incidence of confirmed dengue per 100 , 000 persons corresponding to each health center's district and overall in Managua . The expansion factor for confirmed cases was calculated by dividing the annual incidence of laboratory-confirmed symptomatic dengue among PDCS participants by the annual incidence of laboratory-confirmed dengue in Managua according to NES statistics . In addition , a sub-analysis was performed to calculate expansion factors using only data from 2–9 year-old children in both the PDCS and Managua ( NES ) in 2005–8 in order to compare the results with expansion factors calculated using the complete data sets , which could not be precisely age-matched ( age-disaggregated data is not available from the NES: in 2004 , NES dengue case reports are aggregated into 1–14 year-old age groups , and in 2005–8 , into 1 , 2–4 , 5–9 , and 10–14 year-old age groups ) . We also calculated an expansion factor for suspected dengue cases by dividing the annual incidence of suspected dengue cases in the PDCS meeting WHO criteria by the annual incidence of suspected dengue cases according to NES statistics . In addition , we calculated the ratio of inapparent DENV infections ( defined above ) to symptomatic dengue cases in the cohort . Annual incidence and expansion factors were calculated in relation to the annual dengue season , beginning in July of each year .
In the PDCS cohort , between 13 and 65 confirmed cases were recorded annually ( Table 1 ) , and between 0 and 51 confirmed cases were reported to the NES in individual health districts ( data not shown ) . This translates to incidence of dengue ranging from 343 to 1 , 759 cases per 100 , 000 persons in the cohort study , as compared to 21 to 77 cases per 100 , 000 persons across all Managua's health centers ( Table 1 ) . The HCSFV , where the cohort study is based , reported greater numbers of confirmed and suspected dengue cases among its non-PDCS patients to NES than most other health centers ( 51 to 206 cases per 100 , 000 , data not shown ) , though still only a fraction of the incidence observed in the cohort ( Figure 1 ) . The expansion factor ranged from 14 to 28 dengue cases in the cohort study for every confirmed case reported by the NES . Thus , despite year-to-year variation in the numbers of dengue cases , the cohort study consistently identified approximately 15- to 30-fold ( average 21 . 3 ) more cases than were reported in Managua via national surveillance . Applying these annual expansion factors to national-level data , the estimated incidence of laboratory-confirmed dengue throughout Nicaragua ranged from approximately 300 to 1 , 000 cases per 100 , 000 persons . In a sub-analysis using only 2–9 year-olds in the PDCS and Managua NES data sets in 2005–8 , we calculated expansion factors of 26 , 19 and 21 – very similar to the numbers ( 24 , 14 and 22 ) obtained using the full data sets in 2005–8 . Expansion factors were also calculated based on PDCS and NES suspected dengue cases , yielding an average of 21 ( range 16–28 ) times more suspected cases per 100 , 000 persons in the PDCS compared to those reported by NES throughout Managua . The ratio of inapparent to symptomatic DENV infection in PDCS participants also varied year-to-year [8] from 16 in 2004–5 and 2006–7 , to 5 and 3 in 2005–6 and 2007–8 , respectively ( Table 1 ) . Combining the ratio of inapparent to symptomatic DENV infection with the calculated expansion factor each year , we estimate 68–347 ( average 193 ) inapparent DENV infections for every one symptomatic DENV infection reported to the NES . During the four years analyzed , 42% of inapparent DENV infections in the PDCS were primary , and thus at greater risk of more severe disease in a second DENV infection .
In this study , we have shown much higher incidence of symptomatic DENV infection in a pediatric cohort study than is reported to the national surveillance system from comparable urban public health centers . Annual expansion factor calculations indicate that up to 28 times more symptomatic DENV infections may occur than are reported to the NES , and under-reporting may be even greater among older age groups . This expansion factor , albeit a rough approximation , provides an estimate of the actual impact of dengue on this urban Latin American population , which may be of great use to governments and other institutions involved in dengue prevention and research . Furthermore , estimates of inapparent DENV infections and their immune status ( primary vs . secondary ) such as we provide here , especially when coupled with information about circulating DENV serotypes each year , may be useful for understanding the level of protective immunity in the population as well as help assess population risk for more severe dengue , since the single greatest risk factor for severe disease is a previous infection with a distinct DENV serotype [9] . This data also illustrates that total DENV transmission is distinct from transmission that can be observed clinically or reported through national surveillance systems . This was not a classic capture-recapture study , but rather an ecological study comparing incidence rates in a cohort to national surveillance rates in the surrounding urban areas . The application and calculation of an expansion factor based on PDCS data has several limitations . The expansion factor cannot be applied to adult populations as only children , who are most affected by dengue in the study area , were included in the study . Limitations in NES reporting precluded calculations of incidence in precisely the same age groups as PDCS participants ( e . g . 2–9 in 2004–5 , 2–10 in 2005–6 , etc ) ; the larger NES population used ( through 14 years old ) may have led to slightly lower calculated expansion factors . However , in a sub-analysis restricted in both the PDCS and NES data to only 2–9 year-old children in 2005–8 , expansion factors were found to be very similar to the numbers obtained with the full data set in 2005–6 and 2007–8 ( 26 vs . 28 and 21 vs . 22 , respectively ) . The slightly higher expansion factor in 2006–7 for 2–9 year old children only ( 19 vs . 14 ) is due to the higher incidence that year in 10–14 year-olds in Managua; this was not apparent in the older PDCS participants , which included only 10 and 11 year-old children that year ( Table 2 ) . Additionally , the HCSFV district , which borders Lake Managua , may have higher dengue rates than other health centers , as there was more reported dengue among the non-study population of the HCSFV . However , another plausible explanation is that these higher numbers are due to the impact of the PDCS study protocol and increased awareness of dengue among both non-PDCS medical staff at the HCSFV and in the general population of the HCSFV catchment area . A greater expansion factor for pediatric dengue may be estimated in our study due to the inclusion of undifferentiated febrile illnesses that do not fit the traditional WHO definition , which account for approximately 25% of symptomatic cases identified in the cohort [8] . However , we also calculated an expansion factor based on suspected dengue cases ( which only reflect the WHO criteria ) in the PDCS compared to suspected cases in Managua health centers , and obtained very similar results as those calculated using only confirmed cases . While we were unable to restrict the estimates of national incidence to only urban areas due to limitations of the NES reporting system , since over 56% of Nicaragua is urban , using the entire Nicaraguan population could underestimate the expansion factor by up to 2-fold . However , all our calculated expansion factors ( ∼20-fold ) are an order of magnitude greater than this potential correction factor; thus , we believe our overall estimates and conclusions to be valid and useful . Despite these limitations , our expansion factors falls in the same range as the only published expansion factors for ambulatory dengue . Meltzer et al . [10] calculated that 10 and 27 times more DENV cases occur in Puerto Rico than are reported in pediatric and adult populations , respectively . It is to be expected that expansion factors may vary somewhat in different countries with distinct surveillance programs , health care systems , and degrees of under-reporting . More sophisticated analysis is needed to calculate more precise expansion factors , for instance by controlling for socioeconomic and other district-level factors , and additional data is needed in order to calculate similar expansion factors for adult populations . Nonetheless , an expansion factor such as we present here should allow for more accurate estimations of dengue burden and economic costs . Such estimates can benefit government surveillance programs , aid in allocation of resources to medical care and prevention , and facilitate calculation of the economic benefits of developing vaccines and drugs against dengue . | National public health and epidemiology programs are often tasked with tracking certain infectious diseases , but many barriers lead to under-reporting . In tropical and subtropical countries where dengue fever is endemic , under-reporting may be due to misdiagnosis , limitations of the WHO case classification , and lack of laboratory infrastructure or resources , among other issues . In Nicaragua's capital , Managua , we compared the number of dengue cases identified in a cohort study of 2- to 12-year-old children in one health center to the number of pediatric cases reported from all other municipal health centers in Managua to the National Epidemiologic Surveillance ( NES ) program . In the years 2004–2008 , between 13 and 65 dengue cases were identified in the cohort ( approximately 3 , 700 participants ) , and between 0 and 51 cases were reported by individual health centers in Managua . When the incidence per 100 , 000 people was calculated and compared , an average of 21 times more dengue cases were identified in the cohort study compared to the number reported to NES . Application of such an expansion factor may help governments and other health institutions to estimate the actual number of dengue cases in a population and therefore better allocate resources for prevention , treatment , and drug and vaccine development . | [
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] | [
"public",
"health",
"and",
"epidemiology/epidemiology",
"infectious",
"diseases/viral",
"infections",
"public",
"health",
"and",
"epidemiology/global",
"health",
"public",
"health",
"and",
"epidemiology/infectious",
"diseases"
] | 2010 | High Dengue Case Capture Rate in Four Years of a Cohort Study in Nicaragua Compared to National Surveillance Data |
Dengue infection ranks as one of the most significant viral diseases of the globe . Currently , there is no specific vaccine or antiviral therapy for prevention or treatment . Monocytes/macrophages are the principal target cells for dengue virus and are responsible for disseminating the virus after its transmission . Dengue virus enters target cells via receptor-mediated endocytosis after the viral envelope protein E attaches to the cell surface receptor . This study aimed to investigate the effect of silencing the CD-14 associated molecule and clathrin-mediated endocytosis using siRNA on dengue virus entry into monocytes . Gene expression analysis showed a significant down-regulation of the target genes ( 82 . 7% , 84 . 9 and 76 . 3% for CD-14 associated molecule , CLTC and DNM2 respectively ) in transfected monocytes . The effect of silencing of target genes on dengue virus entry into monocytes was investigated by infecting silenced and non-silenced monocytes with DENV-2 . Results showed a significant reduction of infected cells ( 85 . 2% ) , intracellular viral RNA load ( 73 . 0% ) , and extracellular viral RNA load ( 63 . 0% ) in silenced monocytes as compared to non-silenced monocytes . Silencing the cell surface receptor and clathrin mediated endocytosis using RNA interference resulted in inhibition of the dengue virus entry and subsequently multiplication of the virus in the monocytes . This might serve as a novel promising therapeutic target to attenuate dengue infection and thus reduce transmission as well as progression to severe dengue hemorrhagic fever .
Dengue infection ranks as one of the most clinically significant and prevalent mosquito-borne viral diseases of the globe . It is an expanding public health problem particularly in the tropical and subtropical areas [1] . Following an incubation period of 3 to 14 days , fever and a variety of symptoms occur , coinciding with the appearance of dengue virus ( DENV ) in blood [2] . Immunopathological studies suggest that many tissues may be involved during dengue infection , as viral antigens are expressed in liver , lymph node , spleen and bone marrow [3] , [4] , [5] . DENV can infect and replicate in different mammalian cells , including monocytes , macrophages , dendritic cells , B and T leukocytes , endothelial cells , and bone marrow- , hepatoma- , neuroblastoma- and kidney-derived cells . Based on several observations and antibody dependent enhancement hypothesis , monocyte lineage cells are the major target for DENV [6] , [7] , [8] , [9] . These cells are responsible for replication and dissemination of the virus after the infection from mosquito bites . Since monocytes/macrophages are active phagocytic cells with cytoplasmic lysosomal components that can eliminate microorganisms [10] , the interaction of DENV with monocytes/macrophages may have detrimental effects on both virus and cells . DENV infected monocytes/macrophages release soluble mediators that strongly influence the biological characteristics of endothelial cells and the hematopoietic cell population . This indicates that the interactions between DENV and monocytes/macrophages are important in the pathogenesis of dengue hemorrhagic fever ( DHF ) and dengue shock syndrome ( DSS ) . Previous studies suggest that DENV enters target cells after the viral envelope protein E attaches to an uncharacterized cell receptor [11] . Current studies indicate that multiple cell surface molecules , including GRP78 [12] , heat shock proteins ( Hsp ) 70 and 90 [13] , [14] , lipopolysaccharide-binding CD 14-associated molecule [8] , [15] , [16] , laminin receptor [17] , mannose receptor [18] , and DC-SIGN [19] , were involved in DENV binding and subsequent virus infection in different target cells . CD-14 associated molecule has implicated as a surface receptors on monocytes for DENV-2 entry [8] , [15] , [16] . CD-14 associated molecule is a membrane protein expressed by monocytes , antigen presenting cells and neutrophils and plays a role in the innate immune system . CD-14 associated molecule is necessary for the cellular response in infections mediated by bacterial lipopolysaccharide , which activates monocytes for the expression of cytokines , growth factors , and procoagulatory factors [20] . Failure of interaction of lipopolysaccharide with CD-14 has been implicated in susceptibility to infectious diseases [21] . After binding to the surface receptor , DENV internalizes into the cell cytoplasm by membrane fusion and this process may take place within intracellular vesicles ( pH-dependent ) . This endocytic internalization offers the advantage of guiding the virion to an adequate site for replication and bypassing many cytoplasmatic barriers . Several endocytic pathways have recently been identified [22] , but the clathrin-mediated endocytosis has been demonstrated as the main entry pathway for DENV-2 [23] , [24] , [25] . Clathrin-mediated endocytosis plays a crucial role in the formation of coated vesicles , antigen presentation , nutrient acquisition , clearance of apoptotic cells , pathogen entry , receptor regulation , hypertension , and synaptic transmission . Despite the importance and increasing incidence of DENV as a human pathogen , there are no antiviral agents or vaccines available for treatment or prevention . It is necessary to develop new therapeutic strategies by targeting the DENV binding molecules on susceptible host cells . The blockade of DENV entry into the host cell is an interesting antiviral strategy because it represents a barrier to suppress the onset of infection [26] . RNA interference ( RNAi ) is a potent and specific post-transcriptional gene silencing event induced by double stranded RNAs , which degrade the target mRNA that has the homologous sequence [27] , [28] . RNAi has become the most useful and powerful research tool for studies of gene functions , regulation of gene transcriptions and gene therapies over the last few years [29] . Compared to other traditional gene silencing methods , it has the advantage of significantly enhanced potency , specificity and versatility [30] , [31] . The mediators of sequence specific mRNA degradation are small interfering RNAs ( siRNA ) . siRNAs are 21 to 29 nucleotide in length and carry 5′ phosphate and 3′ hydroxyl termini and 2-nt 3′ overhangs [32] , [33] . siRNAs can effectively suppress target genes in mammalian cells without triggering interferon production [30] , [34] . The transfection of siRNAs into mammalian cells resulted in a potent , long-lasting reaction typically several days and extraordinarily specific post-transcriptional silencing of target genes [33] , [35] . The use of siRNA is suitable for the design of novel gene-specific therapeutics by directly targeting the mRNAs of the related genes and holds great promise for the application of gene-specific therapies in treating acute diseases such as viral infection , cancer , and , perhaps , acute inflammation . Several preliminary studies show that , RNAi has been used effectively to inhibit the replication of different pathogenic viruses in culture , including respiratory syncytial virus , hepatitis viruses , influenza virus , poliovirus , rhinovirus and human immunodeficiency virus [30] , [35] . This study aimed to investigate the possible usage of siRNA in the reduction of DENV entry and replication in human monocytes by targeting the monocytes attachment receptor and clathrin-mediated endocytosis . CD-14 associated molecule has been targeted as an attachment receptor on the surface of monocytes . Besides that , two main components of the clathrin-mediated endocytosis pathway has also been selected; Clathrin heavy polypeptide ( CLTC ) which is required for clathrin-coated pits formation , and Dynamin 2 ( Human-DNM2 ) which is important for pinching off of endocytic vesicles from the plasma membrane . Three siRNAs were designed for each target gene ( CD-14 associated molecule , CLTC , and DNM2 ) to guarantee a greater potency than a single siRNA and eliminate off-targets effect .
Peripheral blood samples from different healthy donors who were not associated with prior dengue infection were collected in EDTA tubes . Peripheral blood mononuclear cells ( PBMC ) were isolated by density centrifugation with Ficoll-paque as described previously [36] . Human monocytes were purified by depletion of non-monocytes followed by negative selection using a MACS human monocytes isolation kit II ( cat #: 130-091-153 , Miltenyi Biotec GmbH , Gladbach , Germany ) in accordance to manufacturer's protocol . Briefly , non-monocytes were indirectly magnetically labeled with a cocktail of biotin-conjugated mouse monoclonal antibodies against CD3 , CD7 , CD16 , CD19 , CD56 , CD123 and Glycohorin A , as primary labeling reagents , and anti-biotin monoclonal antibodies that were conjugated to MicroBeads , as secondary labeling reagents . The magnetically labeled non-monocytes were depleted by retaining them on a MACS® column in the magnetic field of a MACS separator , while the unlabeled monocytes passed through the column . Written informed consent was obtained from blood donors . Ethical clearance ( REF No . : 794 . 5 ) was obtained for the main project from the Scientific and Ethical Committee at the University Malaya Medical Center prior to commencement of study . The nucleotide sequences for the CD-14 associated molecule ( NM_000591 ) , CLTC ( NM_004859 ) , and DNM2 ( NM_001005360 ) transcripts were obtained from GenBank . siRNA sequences were designed by using the web-based tool IDT SciTools RNAi Design available from Integrated DNA Technologies ( IDT ) , Inc . at www . idtdna . com . Three siRNAs were designed for each gene ( Table 1 ) . A pool consists of the three siRNAs for each gene was custom chemically synthesized by 1st BASE Pte Ltd , Singapore . The synthesized siRNAs were purified by HPLC , and a 2′-O-methyl modification at position 2 was introduced to deactivate the off-target activity of the siRNA without compromising the silencing effectiveness [37] . The siRNAs were transfected using siLentFect™ Lipid Reagent ( cat #: 170-3361 , Bio-Rad Laboratories , Hercules , CA , USA ) . Transfection was achieved at a cell density of 1 . 5×105 cells/well in a 24 wells plate by using 1 . 0 µl of lipid transfection reagent and optimized siRNA concentration ( 25 , 50 , and 50 nM for CD14 associate molecule , CLTC , and DNM2 ) , respectively . The positive control was transfected with pre-validated siRNA oligonucleotide targeting β-actin gene . The negative control received the siLentFect™ Lipid Reagent plus the scramble siRNA oligonucleotides . After 24 h , the cells were harvested for monitoring the gene silencing effect at the mRNA expression level while the infectious entry was evaluated after infecting the transfected monocytes with DENV-2 after 72 h . Cytotoxicity was tested by a fluorescent assay that measure LDH released from cells with compromised membrane using CytoTox-ONE™ Homogeneous Membrane Integrity Assay ( Promega , Madison , WI ) in accordance to manufacturer's protocol . Viable monocytes also were counted by staining with 0 . 4% trypan blue ( Invitrogen ) . The number of trypan blue positive and negative cells were counted on a hemocytometer under light microscope . Cell counts were used to verify the LDH results . Total RNA was extracted using RNeasy® Plus Micro Kit ( cat #: 74034 , QIAGEN , Hilden , Germany ) for gene expression analysis and intracellular DENV quantification by reverse transcription quantitative real-time PCR ( RT-qPCR ) . RNA extraction was done according to the manufacturer's instructions . Total RNA was purified using gDNA eliminator columns and was checked by including no-reverse transcription control in the RT-qPCR amplification procedure . For quantification of DENV in the cellular supernatant , viral RNA was extracted separately from 200 µl of tissue culture supernatant using the QIAmp viral RNA mini kit ( cat #: 52904 , QIAGEN , Hilden , Germany ) according to the manufacturer's specifications . All RNA samples were examined for their purity and concentration using Implen NanoPhotometer™ ( Implen GmbH , Munich , Germany ) . All extracted samples have absorbance ratios of 1 . 8–2 . 0 at 260/280 nm , indicating all the samples were free from potentially accumulating proteins during the RNA extraction procedure . In addition , the absence of RNA degradation was assessed by agarose gel electrophoresis . Isolated RNA was stored at −80°C until used . For gene expression analysis , 500 ng of total RNA were reverse-transcribed into cDNA using the iScript™ cDNA synthesis kit ( cat #: 170-8891 , Bio-Rad Laboratories , Hercules , CA , USA ) . All sequences of target gene [CD-14 ( CD-14 associated molecule ) ; CLTC ( Clathrin heavy polypeptide ) ; and DNM2 ( Dynamin 2 ) ] , experimental control gene [ACTB ( β-actin ) ] and reference genes [RPL22 ( ribosomal protein L22 ) ; and RPS29 ( ribosomal protein S29 ) ] were retrieved from GenBank . qPCR primer pairs for CD-14 , CLTC , DNM2 , and ACTB were designed by using Primer Express software V3 . 0 . Primers for RPL22 and RPS29 were obtained from published sequences [38] . Various concentrations of primer sets and a range of annealing temperatures were analyzed to achieve optimal qPCR specificity and efficiency ( data not shown ) . A dissociation analysis was performed after each qPCR runs with cDNA as a template , in order to show that each primer set amplified the expected single product . A pair of primer was considered valid when the efficacy of amplification is between 90–110% with a minimum r2 of 0 . 980 . Table 2 shows the primers used in this study . We used the geNorm algorithm v3 . 5 [39] to determine the number and the most stable reference genes for this study ( data not shown ) . qPCR was performed on CFX96™ Real-Time PCR Detection System ( Bio-Rad , Hercules , CA , USA ) . The reaction was carried out in a final volume of 20 µl , including 10 µl 2× iQ™ SYBR® Green Supermix ( cat #: 170-8882 , Bio-Rad Laboratories , Hercules , CA , USA ) , 2 µl cDNA template , 6 µl RNase/DNase-free sterile water and 1 . 0 µl ( 250 nM final concentration ) of each primer . qPCR was carried out with an initial 5 min hot start activation of the polymerase at 95°C then 40 cycles of 10 sec denaturation at 95°C , 20 sec annealing at 57 . 7°C with a single fluorescence emission measurement and 30 sec extension at 72°C , followed by 5 min at 72°C for final extension . The specificity of amplicons was verified by melting curve analysis ( 74 to 95°C ) with a heating rate 0 . 5°C per 5 sec to verify the identity and purity of the amplified products . qPCR experiments were performed in triplicate . Also , no template and no-reverse transcription controls were included . Baseline and quantification cycle ( Cq ) values and gene expression analysis were done using the Bio-Rad CFX Manager Software 1 . 6 . DENV-2 New Guinea Clone ( NGC ) stock was propagated in C6/36 cells and stored at −80°C until used as described previously [40] . Virus was titrated using the plaque formation assay on PscloneD cells as described previously [41] . Infection of silenced and non-silenced monocytes was performed in a 24-wells tissue culture plate . 1 . 5×105 cells per well were seeded in each well . At the time of infection , medium was removed , DENV-2 ( MOI of 2 ) was added , and followed by incubation at 37°C for 2 h to allow viral adsorption . The culture plate was gently agitated every 15 min for optimal virus to cell contact . Thereafter , the viral supernatant was removed , and the cells were washed three times with serum free media to remove residual virus . Fresh complete growth RPMI 1640 medium was added and then incubated at 37°C and 5% CO2 for 72 h prior to harvest . Cellular supernatant was collected and stored in aliquot at −80°C until use for viral RNA copies quantification by RT-qPCR , while monocytes were harvested for monitoring intracellular infectious DENV by flow cytometry and quantification of viral RNA copies by RT-qPCR . Quantification of dengue viral RNA by RT-qPCR requires a reliable standard curve . 10-fold serial diluted of a known copies number of DENV RNA was diluted and used to generate the standard curve . One-step RT-qPCR was carried out in CFX96™ Real-Time PCR Detection System using the iScript™ One-Step RT-PCR Kit with SYBR® Green ( cat #: 170-8893 , Bio-Rad Laboratories , Hercules , CA , USA ) . After optimization , the reaction was performed in a final volume of 25 µl , including 12 . 5 µl 2× SYBR® Green RT-PCR reaction mix , 6 . 25 µl RNase/DNase-free sterile water and 0 . 5 µl ( 200 nM final concentration ) of each primer ( Table 2 ) , 0 . 25 µl iScript Reverse Transcriptase for One-Step RT-PCR and 5 µl RNA template . The thermal cycling profile of this assay consisted of a 30 min reverse transcription step at 50°C , 15 min of Taq polymerase activation at 95°C , followed by 35 cycles of PCR at 95°C of denaturing for 30 sec , 40 sec annealing at 58 . 0°C and 50 sec extension at 72 . 0°C with a step of a single fluorescence emission data collection followed by 10 min at 72°C for final extension . The specificity of amplicon was verified by melting curve analysis ( 72 to 95°C ) with a heating rate 0 . 5°C per 5 sec to check the identity and purity of the amplified products . Triplicate reactions were carried out for each sample , and no template control was included . Cq values and number of copies for DENV-2 RNA per sample were calculated using the Bio-Rad CFX Manager Software 1 . 6 . Silenced and non-silenced DENV-2 infected and non-infected monocytes ( control ) were harvested at 72 h after DENV-2 infection for flow cytometric determination of DENV infection . Cells were washed twice in PBS by centrifugation , fixed and permeabilized in 200 µl Cytofix/Cytoperm solution ( BD Biosciences , San Diego , CA ) for 15 min at room temperature followed by washing twice in Cytoperm/Cytowash solution ( BD Biosciences , San Diego , CA ) . Cells were stained using indirect staining method . First , cells were incubated in 100 µl of MAb ( anti-DENV-2 for detection of positive samples and anti-DENV-3 as a negative control ) as a primary antibody for 1 h on ice . The cells were then washed twice with Cytoperm/Cytowash solution and incubated with a FITC-labeled goat anti-mouse IgG ( 50 µl ) as a secondary antibody at a final concentration of 3 . 5 µg/mL for 30 min on ice in the dark . Cells were subsequently washed twice with an excess amount of Cytoperm/Cytowash solution and resuspended in 0 . 5 ml of stain buffer ( BD Biosciences , San Diego , CA ) . As a final point , 50 , 000 events were acquired on a FACScaliber using Cell Quest software ( Becton Dickinson Immunocytometry System , San Diego , CA ) . The percentage of dengue positive cells was determined from FITC fluorescence histogram using a region that was defined based on analysis of the infected control monocytes . All assays ( Cytotoxicity assay , gene expression analysis , monocytes infection experiment , flow cytometry detection of infected cells , and intracellular and extracellular quantification of viral RAN by RT-qPCR ) were done in triplicate . All statistical analyses were performed using GraphPad Prism version 5 . 01 ( GraphPad Software , San Diego , CA ) . P values<0 . 05 were considered significant . Error bars are expressed as ± SD .
Our aim was to perform an efficient siRNA transfection to achieve a maximum silencing effect of the target mRNAs . We have optimized the transfection experiment to produce a maximum silencing with minimal or no toxicity to the host cells . We observed that the highest efficiency of siRNA transfection could be achieved at a cell density of 1 . 5×105 cells/well in a 24 wells plate by using 1 . 0 µl of siLentFect Lipid transfection reagent . The optimal siRNA concentration was 25 , 50 , and 50 nM for CD14 associate molecule , CLTC and DNM2 respectively ( data not shown ) . RT-qPCR was performed with mRNA extracted from the transfected monocytes to examine the knockdown level of the target mRNAs induced by siRNA transfection . Gene expression was calculated using the algorithm provided by Bio-Rad CFX Manager Software 1 . 6 , which is based on the algorithm outlined previously [39] . Results were expressed as normalized fold expression as compared to the non-transfected control and normalized to reference genes . First , we screened the silencing efficiency of pooled siRNAs , which comprised of three different siRNAs targeting the same gene as shown in Figure 1 . The knockdown levels were 82 . 4%±1 . 9 , 86 . 0%±0 . 6 and 78 . 5%±3 . 0 for pooled CD-14 associated molecule siRNA , pooled CLTC siRNA and pooled DNM2 siRNA respectively ( One-way ANOVA with Dunnett's post-test , P<0 . 0001 ) . The monocytes were then transfected with a combination of the three different siRNA pools to target on all the three genes simultaneously . We found that there are no significant differences in the knockdown level for all the genes ( 82 . 7%±2 . 4 , 84 . 9%±1 . 2 and 76 . 3%±1 . 7 for CD-14 associated molecule , CLTC and DNM2 respectively ) when compared to separate transfection ( Two-way ANOVA with Bonferroni post-test , P>0 . 05 ) . Furthermore , scramble siRNA had no inhibitory effect on any gene expression and observation was similar to the non-transfected control ( Figure 1 ) . Cytotoxicity from the siRNA delivery was monitored by comparing viable cell numbers in cultures that were transfected with siRNA to that of non-transfected control . In this study , cytotoxicity was determined by measuring LDH released from cells with compromised membrane . Transfection experiment was optimized by treating the monocytes with increasing concentration of siRNA and transfection reagent . Results showed no evidence of toxicity regardless of the concentration of siRNA used and a mild toxicity with the increase of transfection reagent ( data not shown ) . Furthermore , no cytotoxicity effect was observed at 72 h post-transfection of combined siRNAs in monocytes ( >90% viable cells ) ( One-way ANOVA with Dunnett's post-test , P>0 . 05 ) . This observation was further validated by counting the number of viable trypan blue stained cells under the microscope as shown in Figure 2 . Silenced and non-silenced monocytes were independently infected with DENV-2 at a MOI of 2 in order to determine whether silencing CD-14 associated molecule and/or clathrin endocytosis pathway could inhibit DENV entry into monocytes and , therefore , reduce its multiplication . We observed a marked reduction in the percentage of infected cells , CD-14 associated molecule , CLTC , and DNM2 silenced monocytes , by using flow cytometry . The percentage of infected cells was reduced from 34 . 9%±3 . 5 in non-transfected monocytes to 14 . 3%±0 . 4 ( 59 . 0% ) , 12 . 7%±0 . 3 ( 63 . 8% ) , and 18 . 5%±3 . 6 ( 47 . 1% ) in CD-14 associated molecule , CLTC , and DNM2 silenced monocytes , respectively . Interestingly , combined silencing ( CD-14 associated molecule , CLTC , and DNM2 ) of monocytes showed a higher inhibitory effect on DENV entry and replication ( 5 . 1%±0 . 6 ) which represented 85 . 2% reduction in the percentage of infected cells compared with non-silenced monocytes as shown in Figure 3 ( One-way ANOVA with Dunnett's post-test , P<0 . 0001 ) . This observation was further confirmed by RT-qPCR analysis . The viral RNA load in silenced monocytes was compared to the non-silenced monocytes and was normalized to the reference gene ( RPL22 ) . Data is expressed as relative fold expression to non-silenced monocytes , which was defined as 1 . 0 fold ( viral RNA copy number is 3 . 35×103/µl ) . Results showed a significant reduction of DENV RNA level in silenced monocytes: 0 . 57 fold±0 . 10 , 0 . 59 fold±0 . 07 , 0 . 39 fold±0 . 18 , and 0 . 73 fold±0 . 05 in CD-14 associated molecule , CLTC , DNM2 , and combined silenced monocytes , respectively as shown in Figure 4 ( One-way ANOVA with Dunnett's post-test , P<0 . 0001 ) . DENV excreted from the silenced monocytes into the culture supernatant was titrated by RT-qPCR and compared to non-silenced monocytes . Results showed a similar observation when compared to the intracellular quantification results . Figure 5 described the DENV RNA load of the culture supernatant of silenced monocytes compared with non-silenced monocytes , which is defined as 100% ( viral RNA copy number is 1 . 06×104/µl ) . The reduction in viral RNA was 51 . 4%±3 . 4 , 63 . 7%±2 . 9 , 52 . 2%±3 . 0 , and 63 . 0%±5 . 5 for CD-14 associated molecule , CLTC , DNM2 , and combined silenced monocytes , respectively . This result is statistically significant ( One-way ANOVA with Dunnett's post-test , P<0 . 0001 ) .
We successfully inhibit the DENV entry and subsequent virus multiplication in monocytes by silencing the surface receptor and clathrin mediated endocytosis using RNA interference . This tool might serve as a novel promising therapeutic agent for the attenuation of dengue infection , which could subsequently leads to the reduction of disease transmission as well as progression to severe form of the disease DHF . NM_000591 , NM_004859 , NM_001005360 , NM_001101 , NM_000983 , NM_001030001 , NC001477 . | Prevention and treatment of dengue infection remain a serious global public health priority . Extensive efforts are required toward the development of vaccines and discovery of potential therapeutic compounds against the dengue viruses . Dengue virus entry is a critical step for virus reproduction and establishes the infection . Hence , the blockade of dengue virus entry into the host cell is an interesting antiviral strategy as it represents a barrier to suppress the onset of infection . This study was achieved by using RNA interference to silence the cellular receptor , and the clathrin mediated endocytosis that enhances the entry of dengue virus in monocytes . Results showed a marked reduction of infected monocytes by flow cytometry . In addition , both intracellular and extracellular viral RNA load was shown to be reduced in treated monocytes when compared to untreated monocytes . Based on these findings , this study concludes that this therapeutic strategy of blocking the virus replication at the first stage of multiplication might serve as a hopeful drug to mitigate the dengue symptoms , and reduction the disease severity . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [
"biotechnology",
"medicine",
"public",
"health",
"and",
"epidemiology",
"small",
"molecules",
"microbiology",
"host-pathogen",
"interaction",
"developmental",
"biology",
"emerging",
"infectious",
"diseases",
"microbial",
"growth",
"and",
"development",
"applied",
"microbiology",
"infectious",
"diseases",
"medical",
"microbiology",
"gene",
"expression",
"biology",
"public",
"health",
"molecular",
"biology",
"virology",
"microbial",
"control",
"molecular",
"cell",
"biology"
] | 2011 | Inhibition of Dengue Virus Entry and Multiplication into Monocytes Using RNA Interference |
Transcription of the Neurospora crassa circadian clock gene frequency ( frq ) is an essential process in the negative feedback loop that controls circadian rhythms . WHITE COLLAR 1 ( WC-1 ) and WHITE COLLAR 2 ( WC-2 ) forms the WC complex ( WCC ) that is the main activator of frq transcription by binding to its promoter . Here , we show that Centromere Binding Factor 1 ( CBF-1 ) is a critical component of the N . crassa circadian clock by regulating frq transcription . Deletion of cbf-1 resulted in long period and low amplitude rhythms , whereas overexpression of CBF-1 abolished the circadian rhythms . Loss of CBF-1 resulted in WC-independent FRQ expression and suppression of WCC activity . As WCC , CBF-1 also binds to the C-box at the frq promoter . Overexpression of CBF-1 impaired WCC binding to the C-box to suppress frq transcription . Together , our results suggest that the proper level of CBF-1 is critical for circadian clock function by suppressing WC-independent FRQ expression and by regulating WCC binding to the frq promoter .
Circadian rhythms exist in almost all kingdoms of life [1–3] . The circadian clock is a highly conserved timekeeping system that allows organisms to anticipate and adjust to daily environmental changes [4] . These rhythms are generated by self-sustained molecular oscillators , based on transcription-translation negative feedback loops in eukaryotic model organisms [1] . In Neurospora crassa , the loops involve in a number of core clock proteins including WHITE COLLAR proteins WC-1 and WC-2 , FREQUENCY ( FRQ ) , FRQ-interacting RNA helicase ( FRH ) , Casein Kinase I ( CKI ) , Casein Kinase II ( CKII ) , and several other accessory factors [5–8] . In the N . crassa circadian system , WC-1 and WC-2 , two PER-ARNT-SIM ( PAS ) domain-containing transcription factors , form a heterodimeric complex ( WCC ) that initiates transcription of the clock gene frq through binding to PLRE and C-box regulatory sequences [9 , 10] . FRQ protein dimerizes and forms a complex with FRH , functioning as the negative element in the circadian negative-feedback loop [11] . Once FRQ protein is made , it becomes progressively phosphorylated . Then , the FRQ-FRH complex in conjunction with kinases , such as CKI and CKII , promotes phosphorylation of WCC , leading to its inactivation [11–18] . The phosphorylation kinetics determines circadian period length [19] . When FRQ protein becomes extensively phosphorylated , it is ubiquitinated by an E3 ligase complex and degraded through the proteasome , allowing the cycle to restart [19–22] . Rhythmic activation and repression of frq transcription generates rhythmic frq mRNA , which is the basis of circadian gene expression . Activation of the frq transcription by WCC has been characterized in detail [9 , 10 , 23] . Two WCC-binding elements within the promoter of frq are necessary to maximally induce frq transcription in response to light [10] . A distal element , the C-box , is necessary and sufficient to promote rhythmic frq expression in constant darkness ( DD ) [24 , 25] . WC complex was long thought to be the only transcriptional activator of frq transcription [9] . However , we recently discovered that WC-independent frq transcription can be regulated by the RCO-1/RCM-1 complex , the SET-2 pathway , and the IEC-1-INO80 complex [26–29] , suggesting that regulation of frq transcription is complex . Centromere Binding Factor 1 ( CBF-1 ) is initially identified by its ability to bind to centromeric DNA element I ( CDEI ) in Saccharomyces cerevisiae to ensure correct separation of chromosome [30] . CBF-1 , which is a basic helix-loop-helix-leucine zipper ( bHLH-LZ ) factor , recognizes the consensus sequence 5’-CACGTG-3’ [31] and plays a key role in the regulation of methionine metabolism involving in the formation of the CBF-l-Met4-Met28 complex [32–34] . In phosphate metabolism , CBF-1 and the transcription factor Pho4 regulate the sensitivity of promoters to phosphate-concentration levels [35] . CBF-1 functions to modulate chromatin structure at centromeres and regulates transcription from several CDEI-carrying promoters ( i . e . , MET25 , TRP1 , GAL2 ) [36 , 37] . Thus , CBF-1 is a multifunctional protein that influences a number of biological processes . In this study , we discovered that the CBF-1 homolog in N . crassa is necessary for the normal function of circadian clock . Both overt and molecular rhythmicities were severely dampened in cbf-1 deletion or overexpression strains . We found that CBF-1 rhythmically binds to the C-box of the frq promoter . Furthermore , loss of CBF-1 led to WC-independent frq transcription by decreasing RCM-1 recruitment to the frq locus . However , high levels of CBF-1 suppressed frq transcription by impairing WCC binding to the C-box , indicating that CBF-1 plays a critical role in regulating rhythmic WC-dependent frq transcription .
To identify new components that regulate the transcription of clock gene frq in N . crassa , we generated viable knockout mutants of transcription factors and performed race tube assays to screen for mutants with defects in circadian conidiation rhythms . We found that the deletion of cbf-1 gene ( NCU08999 ) resulted in 2-hour longer period and much lower amplitude of circadian conidiation rhythm than those of the wild-type strain ( Fig 1A ) . To confirm the period of the cbf-1 mutant at the molecular level , we introduced a plasmid that carries a luciferase reporter construct ( frq-luc ) into the cbf-1KO strain at the his-3 locus . As shown in Fig 1B , the bioluminescence rhythm of the cbf-1KO , frq-luc strains was of very low amplitude and long period comparing to the wt , frq-luc strains . Sequence alignment revealed that the helix-loop-helix ( HLH ) region of CBF-1 protein is highly conserved from yeast to mammals ( Fig 1C ) . To determine how CBF-1 influences the circadian clock , we examined the FRQ expression profile at different time points in constant darkness ( DD ) . FRQ protein levels were lower in the cbf-1KO strain than in the wild-type strain in DD . Moreover , the peak of FRQ protein was delayed in the cbf-1KO strain relative to the wild-type strain ( Fig 1D ) , which is consistent with its long period phenotype . The levels of frq mRNA in the cbf-1KO strain were increased in DD ( Fig 1E ) , suggesting that CBF-1 suppresses frq transcription . We then evaluated FRQ stability after the addition of the protein synthesis inhibitor cycloheximide ( CHX ) . FRQ stability was similar in the cbf-1KO strain and the wild-type strain ( S1 Fig ) . The mechanism that results in low FRQ protein levels but high frq mRNA levels in the cbf-1KO strain is unknown , but these molecular data indicate that the circadian clock is dampened in the cbf-1KO strain due to impaired regulation of frq expression . To further confirm the function of CBF-1 in the circadian clock , a construct in which the expression of Myc-tagged CBF-1 is driven by quinic acid ( QA ) -inducible qa-2 promoter was introduced to the cbf-1KO strain . We quantified Myc-CBF-1 expression in growth medium containing different QA concentrations ( 0 to 10−2 M ) . The amount of Myc-CBF-1 with 10−7 M QA and without QA ( S2A Fig ) partially restore the conidiation rhythm of cbf-1 mutant ( Fig 2A ) , indicating that the low levels of Myc-CBF-1 partially rescue the circadian conidiation defects of cbf-1KO strain . The conidiation rhythm of the cbf-1KO , qa-Myc-CBF-1 strain was completed rescued when the QA concentration was between 10−6 and 10−4 M in the medium . QA at 10−3 M , however , resulted in a slow growth rate and arrhythmic conidiation rhythm in the cbf-1KO , qa-Myc-CBF-1 strain ( Fig 2A ) , suggesting that the proper level of CBF-1 is critical for circadian clock function . QA-inducible gene expression was known to be suppressed by catabolites [38] . To determine whether the leaky expression of Myc-CBF-1 is subjected to catabolite repression , we evaluated the conidiation rhythm of the cbf-1KO , qa-Myc-CBF-1 strain in different concentrations of glucose in the race tube medium without QA . The race tube phenotype of the cbf-1KO , qa-Myc-CBF-1 strain was not affected by glucose concentration ( S3 Fig ) , suggesting that the leaky expression of Myc-CBF-1 in the absence of QA is probably unrelated to a glucose effect . To verify whether the levels of CBF-1 protein are important for conidiation rhythm , the pqa-Myc-CBF-1 plasmid was introduced into the wild-type stain . Addition of QA ( 0 to 10−2 M ) resulted in increased levels of Myc-CBF-1 in a dose-dependent manner ( S2B Fig ) . As expected , the race tube assay showed that QA ( 10−3 M ) induced expression of high levels of Myc-CBF-1 in the wt , qa-Myc-CBF-1 strain and that the overexpression of Myc-CBF-1 resulted in growth and circadian conidiation defects ( Fig 2B ) . To further confirm these results , we introduced a luciferase reporter construct ( frq-luc ) into the wt , qa-Myc-CBF-1 strain . Overexpression of CBF-1 in the wild-type strain resulted in very low amplitude circadian bioluminescence rhythm ( Fig 2C ) . Thus , CBF-1 is critical for the clock function and high levels of CBF-1 protein interfere with growth and circadian clock function in N . crassa . We then examined the rhythmic expression of FRQ protein and frq mRNA in the wild-type and wt , qa-Myc-CBF-1 strains . High levels of Myc-CBF-1 were induced in the presence of 10−3 M QA ( S2C Fig ) at different time points in DD . Overexpression of Myc-CBF-1 caused a marked decrease in FRQ expression , and FRQ cycling amplitude was also affected ( Fig 2D ) . Northern blot showed that the level of frq mRNA was also reduced in the wt , qa-Myc-CBF-1 strain compared to that in wild-type strain in the first 24 hours ( Fig 2E ) , indicating that CBF-1 suppresses frq transcription . After 24 hours in DD , however , the level of frq mRNA was comparable in mutant and wild-type strains , suggesting that overexpressed CBF-1 also affected FRQ expression at a post-transcriptional level . The low levels of FRQ protein in the wt , qa-Myc-CBF-1 strain promoted us to examine the expression of WC-1 and WC-2 in this strain . The levels of WC-1 and WC-2 proteins in the wt , qa-Myc-CBF-1 strain were lower than those in the wild-type strain ( S4A Fig ) . However , the levels of wc-1 and wc-2 mRNA were similar ( S4B Fig ) . These results suggest that overexpressed CBF-1 negatively regulates the levels of WC-1 and WC-2 proteins in a post-transcriptional manner . To further investigate whether CBF-1 directly regulates frq transcription by binding to the frq promoter , we performed electrophoretic mobility shift ( EMSA ) and chromatin immunoprecipitation ( ChIP ) assays . For the EMSA assay , purified GST-CBF-1 or GST-CBF-1Δ ( 187–214 ) fusion proteins or GST only ( S5A Fig ) were incubated with a radioactively labeled C-box oligonucleotide probe . A migrating complex was observed when the GST-CBF-1 fusion protein was present , but not when GST or the GST-CBF-1Δ ( 187–214 ) fusion proteins were used ( Fig 3A ) . The interaction was specific for the C-box sequence as an unlabeled C-box oligonucleotide disrupted the complex , but an oligonucleotide with a different sequence did not ( Fig 3A ) . These results suggest that CBF-1 binds directly to the C-box of frq promoter sequence . To test whether CBF-1 binds to the frq promoter in vivo , we generated a CBF-1-specific antibody , which recognized a specific band at predicted molecular weight in the wild-type strain but not in the cbf-1KO strain ( Fig 3B ) . A ChIP assay using this antibody showed that the enrichment of CBF-1 at the C-box of frq promoter was specific and rhythmic , peaking at DD14 when frq transcription and WCC binding are high ( Fig 3C ) . The levels of CBF-1 protein , however , were constant ( Fig 3D ) . Because both CBF-1 and WCC are transcription factors and rhythmically bind to the C-box , we evaluated the relationship between the two factors in frq promoter binding . We examined the binding of CBF-1 to the frq promoter in wc-1KO , wc-2KO , and frq9 mutant strains . In the latter , a frame-shift mutation of frq results in truncated FRQ protein and defective negative feedback loop [39] . The binding of CBF-1 to the C-box was constantly low in the wc-1KO and wc-2KO strains but was constantly high in the frq9 mutant strain with high levels WCC activity ( Fig 3E and 3F ) . Moreover , the CBF-1 protein levels were not altered in these mutants ( Fig 3G ) . These data suggested that the rhythmic CBF-1 association with the frq promoter is regulated by WCC activity . To determine how CBF-1 regulates frq transcription , we performed ChIP assays using WC-1 and WC-2 antibodies [27] . Our results showed that WCC rhythmically bound to the C-box of the frq promoter in DD with a peak at DD14 in wild-type strain ( Fig 4A and 4B ) . However , the robust rhythmic binding of WCC to the C-box was dramatically decreased with a low amplitude and delayed peak in the cbf-1KO strain ( Fig 4A and 4B ) . Previous studies showed that hypophosphorylated WC-1 and WC-2 efficiently bound to the C-box activating frq transcription [16 , 17] and that hyperphosphorylated WC-1 and WC-2 had lower affinity for the C-box of frq promoter in DD [40] . Western blots showed that WC-1 and WC-2 were hyperphosphorylated in the cbf-1KO strain compared with the wild-type strain throughout DD ( Fig 4C ) . However , we observed no significant differences in levels of WC-1 or WC-2 in the cbf-1KO strain compared to those in the wild-type strain ( S5B Fig ) . These results suggest that the loss of CBF-1 decreases WCC activity by promoting phosphorylation of WC-1 and WC-2 . A previous study showed that hyperphosphorylation of WC-1 and WC-2 , which was mediated by FRQ , resulted in less binding to the C-box of frq [41] . Thus , the low WCC activity in the cbf-1KO strain may be mediated by FRQ . The decreased binding of WC-1 and WC-2 at C-box of frq promoter and the high levels of frq mRNA in the cbf-1KO strain prompted us to examine whether there is WC-independent frq transcription in the cbf-1KO mutant . Thus , we generated the cbf-1KO wc-2KO double mutant and compared FRQ levels with those in the wc-2KO single mutant . As expected , constant levels of FRQ expression were observed in cbf-1KO wc-2KO double mutant but not in wc-2KO single mutant ( Fig 4D ) . Further , constant high levels of frq mRNA were also observed in the cbf-1KO wc-2KO double mutant ( Fig 4E ) , indicating that CBF-1 is required for suppression of WC-independent transcription of frq . These results indicate that the constant expression of FRQ in the cbf-1KO strain mediates the hyperphosphorylation of WC proteins , resulting in decreased binding of WCC to the C-box of the frq promoter . WC-independent frq expression leads to the increased frq mRNA level in the cbf-1 mutant . However , low luc expression levels in the cbf-1KO , frq-luc strains suggest the lack of WC-independent transcription . The entire ORF and 3’UTR of frq are replaced by luc in frq-luc transgene and the frq-luc transgene is targeted to the his-3 locus . So , the lack of WC-independent transcription of the frq-luc reporter may be caused by the chromatin state of the frq-luc locus being different from the chromatin state of the frq locus [27] . Our previous study showed that the transcriptional co-repressor RCM-1 suppresses WC-independent frq transcription by binding to the frq locus and that hyperphosphorylation of RCM-1 impairs its suppressor activity [28] . The level of RCM-1 enrichment at the C-box of frq promoter was dramatically reduced and RCM-1 became hyperphosphorylated in the cbf-1KO strain compared to the wild-type strain ( Fig 4F and 4G ) , but the levels of RCM-1 were similar in the two strains ( Fig 4H ) . Therefore , the WC-independent frq transcription in the cbf-1KO strain may be caused at least partially by hyperphosphorylation of RCM-1 . Our results suggest that the decreased binding of WCC to the C-box in the cbf-1KO strains is due to FRQ-mediated hyperphosphorylation of WCC . To further test whether CBF-1 directly regulates WCC binding of the C-box independent of FRQ , we generated the cbf-1KO frq9 double mutant ( S6A Fig ) . A ChIP assay showed that WC-2 was significantly enriched at the C-box of the cbf-1KO frq9 double mutant compared to the frq9 single mutant ( Fig 5A ) even though the WC levels were similar in the two strains ( S6B Fig ) . The results indicate that CBF-1 can suppress WC-2 binding at the C-box independently of FRQ . The levels of frq mRNA were higher in the cbf-1KO frq9 double mutant than in the frq9 single mutant ( Fig 5B ) . WC-1 and WC-2 were both hypophosphorylated in the cbf-1KO frq9 double mutant and in the frq9 single mutant ( Fig 5C ) , suggesting that the increased binding of WC-2 to the frq C-box was due to the absence of CBF-1 protein but was not affected by WC phosphorylation status . Taken together , these results suggest that CBF-1 suppresses WCC binding to the C-box of frq promoter through a FRQ-independent mechanism . The HLH DNA binding domains are conserved in all CBF-1 homologues from yeast to human ( Fig 1C ) . To test the role of CBF-1 binding to the frq promoter , we generated FLAG-tagged CBF-1 constructs that contain a E195A point mutation or deletion of the 187–214 amino acid region in the HLH domain [42] . Unlike the wild-type FLAG-CBF-1 protein , the mutant FLAG-CBF-1 proteins failed to rescue the long period conidiation phenotype of the cbf-1KO strain ( Fig 6A and S7A Fig ) , suggesting that the DNA binding activity of CBF-1 is required for its circadian clock function . A ChIP assay showed that CBF-1 enrichment at the C-box in cbf-1KO , pcbf-1-FLAG-CBF-1E195A , and cbf-1KO , pcbf-1-FLAG-CBF-1Δ ( 187–214 ) strains was abolished ( Fig 6B ) . These results demonstrate that CBF-1 regulates the circadian clock via its HLH DNA binding domain . Similarly , neither the defect of WCC binding to the C-box nor the hyperphosphorylation of WC-1 and WC-2 was rescued by mutant FLAG-CBF-1 proteins ( Fig 6C–6F ) . The levels of WC proteins were not affected in the cbf-1KO mutants ( S7B and S7C Fig ) . These results suggest that CBF-1 binding at the C-box of frq promoter is required for its role in the circadian clock . To determine whether the decreased levels of frq mRNA in the wt , qa-Myc-CBF-1 strain is caused by impaired WCC recruitment to the C-box of frq promoter , we performed ChIP assays with WC-1 and WC-2 antibodies . ChIP data showed that robust rhythmic binding of WCC to the C-box was markedly decreased by the overexpression of Myc-CBF-1 in the wt , qa-Myc-CBF-1 strain compared to the wild-type strain ( Fig 7A and 7B ) . In contrast , ChIP assays with CBF-1 antibody showed that the recruitment of CBF-1 to the C-box of frq promoter was increased in the wt , qa-Myc-CBF-1 strain ( Fig 7C ) . These results suggest that overexpression of CBF-1 interferes with WC recruitment to the C-box . To confirm this conclusion , we created a wt , qa-Myc-CBF-1E195A strain and a wt , qa-Myc-CBF-1Δ187–214 strain . As expected , the circadian conidiation phenotype was not affected by overexpression of these mutant Myc-CBF-1 proteins that cannot bind to frq C-box ( Fig 7D ) . Furthermore , ChIP assays with WC-1 and WC-2 antibodies showed that binding of WCC to the C-box was similar in the wt , qa-Myc-CBF-1E195A and wt , qa-Myc-CBF-1Δ ( 187–214 ) strains to that in the wild-type strain ( Fig 7E and 7F ) . In addition , the levels of WCC and FRQ proteins were not affected in the wt , qa-Myc-CBF-1E195A or wt , qa-Myc-CBF-1Δ187–214 strains ( S8 Fig ) . Together , these results are consistent with a model in which CBF-1 binding to the C-box region impairs WCC binding . As WCC and CBF-1 appear to be mutual regulators in frq promoter binding , we tested whether they interact . Co-immunoprecipitation assays were performed using pre-immune serum as the negative control . We found that CBF-1 co-immunoprecipitated with WC-1 and WC-2 , suggesting that these proteins interact to regulate frq transcription ( Fig 7G and 7H ) . Taken together , our results suggest that the proper level of CBF-1 is critical for modulating the binding of WCC at the C-box of the frq promoter to allow rhythmic WC-dependent frq transcription .
Transcriptional control of circadian clock genes is an essential step in negative feedback loops of all eukaryotic clock systems . Previous studies have demonstrated PAS domain-containing transcription factors , such as WC-1 and WC-2 in N . crassa and CLOCK and BMAL1 in mammals , are responsible for rhythmically activating clock gene transcription [43–45] . In this study , we found that CBF-1 , a helix-loop-helix domain-containing transcription factor , is also involved in regulating frq transcription in N . crassa . In the cbf-1KO strain , the circadian conidiation rhythm was severely affected , and the FRQ protein oscillation was delayed in DD . Overexpression of CBF-1 resulted in low amplitude rhythms , decreased levels of frq mRNA , and reduced WCC binding to the C-box of the frq promoter . Finally , rhythmic binding of WCC to the C-box of frq promoter required functional CBF-1 . Taken together , our results suggest that CBF-1 is critical for robust rhythmic frq transcription . The rhythmic association of CBF-1 with the C-box in the frq promoter is regulated by WCC activity . Rhythmic binding of CBF-1 to the C-box was disrupted in wc-1 , wc-2 , and frq9 mutants ( Fig 3F ) , but high levels of WCC activity and high CBF-1 recruitment were observed in the frq9 strain ( Fig 3E and 3F ) . Our data suggest that CBF-1 has dual functions in the circadian clock both by influencing on WCC binding at the frq promoter and by suppressing WC-independent FRQ expression . In the wild-type strain , both WCC and CBF-1 rhythmically bind to the C-box of frq promoter to activate frq transcription ( Fig 8 ) . FRQ protein then promotes phosphorylation of the WCC by CKI and CKII kinases , leading to its inactivation and inhibition of frq transcription . FRQ is progressively phosphorylated by CKI and CKII kinases and degraded . After FRQ degrades to a certain level , WCC is reactivated and the cycle restarts . In the cbf-1KO strains , WC-independent frq transcription is activated , which promotes WCC phosphorylation and inhibits WCC binding to the C-box , resulting in low amplitude and long period phenotype . As shown in Fig 4D and 4E , constant intermediate levels of frq mRNA and FRQ protein were detected in the wc-2KO cbf-1KO double mutant , indicating the activation of WC-independent frq transcription in cbf-1KO mutants . Consistent with this notion , constant high levels of frq mRNA were observed in DD in the cbf-1KO mutant ( Fig 1E ) . In the CBF-1 overexpression strain , we observed elevated CBF-1 binding to C-box region but reduced WCC recruitment , resulting in low level of frq mRNA and low amplitude rhythm . Because both CBF-1 and WCC bind to C-box in the frq promoter and CBF-1 can also bind to C-box independent of WCC , it is possible that high CBF-1 level inhibits WCC C-box binding through competitive binding to C-box . In addition , the reduced WC levels in the CBF-1 overexpression strain can also contribute to the reduced WCC binding to C-box ( S4A Fig ) . Therefore , CBF-1 protein levels must be tightly regulated to allow robust rhythmic WC-dependent frq transcription . A role for CBF-1 in regulating frq transcription by modulating WCC binding at the frq promoter is supported by several lines of evidence . First , EMSA and ChIP assays showed that CBF-1 rhythmically binds to the C-box region in vitro and in vivo ( Fig 3A and 3C ) . Second , the binding of WCC was significantly higher in the frq9 cbf-1KO double mutant than that in frq9 single mutant , suggesting that CBF-1 can suppress WCC binding to the C-box independently of FRQ ( Fig 5A ) . Third , even though CBF-1 and WCC interacts and WCC promotes the C-box binding of CBF-1 , CBF-1 can also bind to C-box independent of WCC ( Fig 3A ) . As a result , overexpression of CBF-1 led to increased CBF-1 binding to the C-box region but decreased WCC binding ( Fig 7C , 7E and 7F ) . Together , these results suggest that CBF-1 binding to the C-box region impairs WCC binding to the C-box region of the frq promoter . Our results suggest that CBF-1 acts as a repressor for WC-independent frq transcription by promoting RCM-1 recruitment ( Fig 4F–4H ) . WC-independent FRQ expression was previously observed in the rco-1KO and rcm-1RIP strains [27 , 28 , 46] . As shown here ( Fig 1A and 1B and S3 Fig ) and in a previous study [46] , roles of CBF-1 and RCO-1/RCM-1 in control of the clock depend on conditions ( i . e . , glucose concentration and liquid/solid media ) . The absence of CBF-1 leads to long period or arrhythmic conidiation when glucose concentration is high ( S3 Fig ) . The inconsistency in period of luciferase assay and race tube assay may also be caused by different media . Therefore , the absence of CBF-1 or RCO-1/RCM-1 leads to a more severe circadian phenotype in high glucose media . In addition to the two functions of CBF-1 discussed above , the relatively low levels of FRQ protein but high frq mRNA level in the cbf-1KO mutant suggest that CBF-1 has additional function in the clock . The inconsistency between frq RNA and FRQ protein was not unique for the cbf-1 mutants and was previously also observed in mcb mutant [17] . In the cbf-1 mutant , we showed that the WC-independent frq expression in the cbf-1mutant is sufficient to promote hyperphosphorylation of WC proteins and inhibit WCC DNA binding ( Fig 4 ) . Therefore , the relatively low FRQ level in the cbf-1 mutant is sufficient to repress WCC activity . Comparison of the FRQ phosphorylation profiles showed that FRQ stayed constant hypophosphorylated in DD in different cbf-1 mutants ( Figs 1D and 4D ) , suggesting that CBF-1 can impact on FRQ phosphorylation due to an unknown mechanism . It is possible that these hypophosphorylated species of FRQ in the mutant is more potent for WCC inhibition than those in the wild-type strain . Consistent with this interpretation , a role for FRQ phopshorylation in the negative feedback loop was previously suggested by several studies [47 , 48] . Here , we showed that the conserved transcription factor CBF-1 , which contains HLH domain , plays an important role in the circadian negative feedback loop . CBF-1 is a member of the evolutionarily conserved bHLH-LZ transcription factor family [42] . Mammalian USF1 , a homolog of N . crassa CBF-1 , is a dominant suppressor of the ClockΔ19 mutation and competes with the CLOCK:BMAL1 complex for binding to E-box sites in target clock genes to regulate circadian gene expression [49] . The protein levels of CBF-1 in N . crassa or USF1 in mammals are very important for circadian clock . Therefore , our study here suggests that the regulation of positive element occupancy at the promoter of the negative element by CBF-1 homologues might be a conserved feature in eukaryotic circadian clock mechanisms . However , there are some differences in how CBF-1 and USF1 act in each clock system . CBF-1 suppresses frq transcription under normal conditions , whereas USF1 activates per/cry transcription when mutant CLOCKΔ19:BMAL1 is not transcriptional competent . In addition , E-box binding pattern of USF1 is antiphase to that of CLOCK . These differences might be evolutionary results from adaption to different clock systems .
The 87–3 ( bd , a ) strain was used as the wild-type strain in this study [50] . The ku70RIP ( bd , a ) strain , generated previously [16] , was used as the host strain for creating the cbf-1 knockout mutant . The cbf-1KO strain was created by deleting the entire cbf-1 ORF through homologous recombination using a protocol described previously [51] . The wc-1KO , wc-2KO , rcm-1RIP and frq9 strains , generated previously [28 , 52–55] , were also used in this study . The newly created cbf-1KO wc-2KO and cbf-1KO frq9 double mutants were obtained by crossing . The 301–6 ( bd , his-3 , A ) and cbf-1KO ( bd , his-3 , A ) strains were the host strains for his-3 targeting construct transformation . The wt , pqa-Myc-CBF-1 , wt , pqa-Myc-CBF-1E195A , and wt , pqa-Myc-CBF-1Δ ( 187–214 ) strains were created by transferring pqa-Myc-CBF-1 , pqa-Myc-CBF-1E195A , and pqa-Myc-CBF-1Δ ( 187–214 ) constructs into the his-3 locus of 301–6 host strain . Using the same method , the cbf-1KO , pqa-Myc-CBF-1 , cbf-1KO , pqa-Myc-CBF-1E195A , cbf-1KO , pqa-Myc-CBF-1Δ ( 187–214 ) , cbf-1KO , pcbf-1-FLAG-CBF-1 , cbf-1KO , pcbf-1-FLAG-CBF-1E195A , and cbf-1KO , pcbf-1-FLAG-CBF-1Δ ( 187–214 ) strains were generated . For each transformation , the transformants were first analyzed by western blot for the expression of tagged CBF-1 proteins , and the positive transformants were examined by race tube assays . Escherichia coli BL21 cells and pGEX-4T-1 plasmid were used for expression of GST-CBF-1 and GST-CBF-1Δ ( 187–214 ) fusion proteins . The medium for race tube assays contained 1x Vogel’s salts , 0 . 1% glucose , 0 . 17% arginine , 50 ng/mL biotin , and 1 . 5% agar . In the race tube medium containing QA , 0 . 1% glucose was replaced with the desired concentration of QA ( 0-10-2 M ) . Strains were grown in constant light at 25°C for 24 hours before being transferred to DD at 25°C . Densitometric analyses of race tubes and calculations of period length were performed as described [56] . Growth conditions were as described previously [57] . Liquid cultures were grown in minimal medium ( 1x Vogel’s , 2% glucose ) . When QA was used , liquid cultures were grown in low-glucose medium ( 1x Vogel’s , 0 . 1% glucose , 0 . 17% arginine ) with different concentration of QA ( 0-10-2 M ) . The luciferase reporter assays were performed as described previously [58 , 59] . The 301–6 ( bd , A ) , frq-luc strain was used as control strain in this study . The cbf-1KO strains were crossed with the 301–6 ( bd , A ) , frq-luc strain to obtain the cbf-1KO , frq-luc strain . The luciferase reporter construct was co-transformed with a pBT6 plasmid into CBF-1 overexpression strain to obtain wt , qa-Myc-CBF-1 , frq-luc strain . LumiCycle ( ACTIMETRICS ) and the autoclaved fructose-glucose-sucrose ( FGS ) -Vogel’s medium ( 1x FGS , 1x Vogel’s medium , 50 μg/L biotin , and 1 . 8% agar ) containing 50 μM firefly D-luciferin were used for the luciferase assay . Conidia suspensions in water were placed on autoclaved FGS-Vogel’s medium and grown in constant light overnight . The cultures were then transferred to constant darkness , and luminescence was recorded in real time using a LumiCycle after one day in DD . The data were normalized with LumiCycle Analysis Software by subtracting the baseline luciferase signal which increases as cell grows . Under our experimental conditions , luciferase signals are highly variable during the first day in the LumiCycle and become stabilized afterwards , which is likely due to an artifact caused by the light-dark transfer of the cultures . Thus , the results presented were recorded after one day in DD . The GST-CBF-1 ( amino acids M1-N136 ) , GST-WC-1 ( amino acids G291-E708 ) , GST-WC-2 ( amino acids M8-C423 ) , GST-RCM-1 ( amino acids D480-P883 ) [28] , and GST-FRQ ( amino acids S249-K315 and E359-G766 ) [60] fusion proteins were expressed in E . coli BL21 cells , and the recombinant proteins were purified and used as the antigens to generate rabbit polyclonal antiserum as described previously [61] . Protein extraction , quantification , western blot analyses , and co-immunoprecipitation assays were performed as previously described [62] . Equal amounts of total protein ( 40 μg ) were loaded in each protein lane . After electrophoresis , proteins were transferred onto PVDF membrane , and western blot analysis was performed . To analyze the phosphorylation profiles of WC-1 , WC-2 , and RCM-1 , phosphatase inhibitors were added to protein extraction buffer and 7 . 5% SDS-PAGE gels containing a ratio of 149:1 acrylamide/bis-acrylamide were used . Otherwise , 7 . 5% SDS-PAGE gels contained a ratio of 37 . 5:1 acrylamide/bis-acrylamide were employed . Total RNA was extracted using Trizol , and then further purified with 2 . 5 M LiCl as described previously [63] . For northern blot analysis , equal amounts of total RNA ( 20 μg ) were loaded onto agarose gels . After electrophoresis , the RNA was transferred onto Amersham Hybond-N+ membrane . The membrane was probed with 32P-UTP-labeled RNA probes specific for frq , wc-1 , or wc-2 . RNA probes were transcribed in vitro from PCR products by T7 RNA polymerase . The northern primer sequences used for the template amplification were frq-N term F ( 5’-GGGTAGTCGTGTACTTTGTCAGGCATAGATCTC-3’ ) , frq-N-term T7 + R ( 5’-TAATACGACTCACTATAGGGGGCAGGGTTACGATTGGATT-3’ ) , wc1 F ( 5’-GTTATACCTGGTTTGAAAGC-3’ ) , wc1 T7 + R ( 5’-TAATACGACTCACTATAGGGACAACTGTTGCATAGATCTC-3’ ) , wc2 F ( 5’-CTGCAGATGACTTCCGACCC-3’ ) , and wc2 T7 + R ( 5’-TAATACGACTCACTATAGGGATCTCGGTCTAGGGGAATC-3’ ) . The T7 promoter regions are underlined . EMSA assays were performed in a manner similar to that described previously [64] . To make the probe , oligonucleotides were dissolved in double distilled H2O ( ddH2O ) to a final concentration of 10 μM . We then mixed 10 μL each of complementary oligonucleotides with 30 μL ddH2O and heated at 95°C for 10 minutes . After overnight at room temperature , oligonucleotide labeling was performed at 37°C for 30 minutes in 50 μL reaction with 5 μL of double-stranded oligonucleotide , 5 μL of T4 polynucleotide kinase buffer ( NEB ) , 2 . 5 μL of T4 polynucleotide kinase ( NEB ) , 7 . 5 μCi γ32P-ATP and 30 μL ddH2O . After the kinase reaction , the sample was purified using Bio-Gel P-30 chromatography columns ( Bio-Rad ) . The oligonucleotides annealed for use as probe were ACF57 ( 5’-CGTCCTGATGCCGCTGCAAGACCGATGACGCTGCAAAATTGAGATCTA-3’ ) ; and ACF58 ( 5’-TAGATCTCAATTTTGCAGCGTCATCGGTCTTGCAGCGGCATCAGGACG-3’ ) . Binding reactions using BL21 cells expressing fusion protein contained 1x binding buffer [20 mM HEPES , pH 7 . 9 , 1 mM EDTA , 2 mM MgCl2 , 10% ( v/v ) glycerol , 20 μM ZnCl2] , 0 . 1 μg poly ( dI-dC ) , 1 μL probe , and 3 μg of GST , GST-CBF-1 , or GST-CBF-1Δ ( 187–214 ) fusion protein ( which was added last to the binding reactions ) in a total volume of 20 μL . Binding reactions were incubated for 30 minutes on ice prior to electrophoresis at 4°C on nondenaturing 4% polyacrylamide gels containing 0 . 5x TBE and 2 . 5% ( v/v ) glycerol . Gels were dried at 80°C for 15 minutes and were exposed it to X-ray film for 2 to 10 hours . ChIP assays were performed as described previously [62] . Briefly , N . crassa tissues were fixed with 1% formaldehyde for 15 minutes at 25°C with shaking . Glycine was added at a final concentration of 125 mM , and samples were incubated for another 5 minutes . The crosslinked tissues are ground and resuspended at 0 . 5 g in 6 mL lysis buffer containing protease inhibitors ( 1 mM PMSF , 1 μg/mL leupeptin and 1 μg/mL pepstatin A ) . Chromatin was sheared by sonication to approximately 200–500 base pair fragments . A 1mL aliquot of protein solution ( 2 mg/mL ) was used for each immunoprecipitation reaction , and 10 μL was kept as the input DNA . The chromatin immunoprecipitations were carried out with 2 . 5 μL WC-2 , 3 . 5 μL WC-1 , 2 . 5 μL RCM-1 , or 3 μL CBF-1 antibodies . The corresponding knock-out strains were used as the negative controls . Immunoprecipitated DNA was quantified using real-time PCR . The primer sets used are frq C-box F ( 5’-GTCAAGCTCGTACCCACATC-3’ ) and frq C-box R ( 5’-CCGAAAGTATCTTGAGCCTCC-3’ ) were described in a previous study [29] . Occupancies were normalized by the ratio of ChIP to Input . The relative values of protein occupancy were calculated using the 2-ΔΔCT method by comparing the cycle number for ChIP sample with that for the Input control [65] . Quantification of western blot and northern blot data were performed using Quantity One software . All experiments were performed at least three independent times . For blots , representative images are shown . Error bars are standard deviations of triplicate data . Statistical significance was determined by Student’s t test for ChIP analyses . | Circadian clocks , which measure time on a scale of approximately 24 hours , are generated by a cell-autonomous circadian oscillator comprised of autoregulatory feedback loops . In the Neurospora crassa circadian oscillator , WHITE COLLAR complex ( WCC ) actives transcription of the frequency ( frq ) gene . FRQ inhibits the activity of WCC to close the negative feedback loop . Here , we showed that the transcription factor CBF-1 functions as a repressor to modulate WCC recruitment to the C-box of frq promoter . Our data showed that deletion or overexpression of CBF-1 dampened circadian rhythm due to impaired WCC binding at the frq promoter . As CBF-1 is conserved in eukaryotes , our data provide novel insights into the negative feedback mechanism that controls the biological clocks in other organisms . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] | [
"phosphorylation",
"gene",
"regulation",
"regulatory",
"proteins",
"messenger",
"rna",
"dna-binding",
"proteins",
"circadian",
"oscillators",
"transcription",
"factors",
"chronobiology",
"molecular",
"biology",
"techniques",
"research",
"and",
"analysis",
"methods",
"transcriptional",
"control",
"proteins",
"hyperexpression",
"techniques",
"gene",
"expression",
"molecular",
"biology",
"molecular",
"biology",
"assays",
"and",
"analysis",
"techniques",
"gene",
"expression",
"and",
"vector",
"techniques",
"circadian",
"rhythms",
"biochemistry",
"rna",
"post-translational",
"modification",
"nucleic",
"acids",
"genetics",
"biology",
"and",
"life",
"sciences"
] | 2018 | Transcription factor CBF-1 is critical for circadian gene expression by modulating WHITE COLLAR complex recruitment to the frq locus |
Visceral leishmaniasis ( VL ) is a disease caused by two known vector-borne parasite species ( Leishmania donovani , L . infantum ) , transmitted to man by phlebotomine sand flies ( species: Phlebotomus and Lutzomyia ) , resulting in ≈50 , 000 human fatalities annually , ≈67% occurring on the Indian subcontinent . Indoor residual spraying is the current method of sand fly control in India , but alternative means of vector control , such as the treatment of livestock with systemic insecticide-based drugs , are being evaluated . We describe an individual-based , stochastic , life-stage-structured model that represents a sand fly vector population within a village in India and simulates the effects of vector control via fipronil-based drugs orally administered to cattle , which target both blood-feeding adults and larvae that feed on host feces . Simulation results indicated efficacy of fipronil-based control schemes in reducing sand fly abundance depended on timing of drug applications relative to seasonality of the sand fly life cycle . Taking into account cost-effectiveness and logistical feasibility , two of the most efficacious treatment schemes reduced population peaks occurring from April through August by ≈90% ( applications 3 times per year at 2-month intervals initiated in March ) and >95% ( applications 6 times per year at 2-month intervals initiated in January ) relative to no control , with the cumulative number of sand fly days occurring April-August reduced by ≈83% and ≈97% , respectively , and more specifically during the summer months of peak human exposure ( June-August ) by ≈85% and ≈97% , respectively . Our model should prove useful in a priori evaluation of the efficacy of fipronil-based drugs in controlling leishmaniasis on the Indian subcontinent and beyond .
The known VL vector on the Indian subcontinent is the sand fly species Phlebotomus argentipes [6] . Phlebotomine sand flies are small Diptera , rarely exceeding a length of 3 mm , in the family Psychodidae and subfamily Phlebotominae [7] and are holometabolous consisting of four life stages: eggs , larvae , pupae , and adults . Sand flies are active primarily at night and are regarded as silent feeders [8] . P . argentipes females host blood feed primarily on cattle and humans within rural villages [9–12] . The blood meal is required in order to complete the oviposition process . Immature sand flies in Bihar have been found largely in areas within and surrounding cattle sheds [13–15] , suggesting cattle feces may serve as a food source for larvae which feed on organic matter . Results of several laboratory experiments have found sand fly processes such as development , mortality and reproduction to be temperature-dependent with many of these processes occurring more rapidly at higher temperatures [16–20] . Nightly air temperatures in Bihar will exceed 20°C between March-October and are highest during the summer ( June-August ) and the observed sand fly population is small in January and February when minimum temperatures are lowest [21] . It has been suggested that further research regarding alternative or integrated vector control approaches should be examined to supplement the current practice [22] . Vector control in India comes in the form of indoor residual spraying ( IRS ) performed historically with DDT and more recently with synthetic pyrethroids . IRS controls endophilic sand flies , but blood-fed sand flies have been collected outdoors and indoors [9 , 21 , 23] . A survey concluded that roughly 95% of Bihari villager households have family members that sleep outdoors at least part of the year [24] . Logically , these villagers are therefore not protected by IRS and are potentially exposed to exophilic sand flies . Fipronil-based drugs , orally administered to cattle and rodents , have been successful in killing laboratory-reared sand flies under controlled conditions , targeting blood-feeding adults and larvae that feed on host feces [25–27] . Orally applied fipronil can remain in the system of animals for several weeks to several months , dependent on the concentration administered ( mg/kg body weight ) and fipronil has a lengthy half-life of approximately 128 days [28] , meaning that sand fly control can potentially be maintained for several months following a single treatment . With this form of treatment , the success of vector control could be independent of exophilic or endophilic feeding behavior and be dependent on host and oviposition site preferences . Hence , this form of treatment could potentially supplement the current practice of IRS by targeting exophilic , cattle-feeding adult sand flies and larval sand flies feeding on organic matter in the form of cattle feces . A reduction in vector density should lead to a reduction in the transmission rate of VL as suggested by a recent VL model which predicted that either reducing vector density >67% through application of adulticides or >79% through breeding site destruction could eliminate the ability of the VL pathogen to persist [29] . Vector and pathogen seasonality in addition to social practice should be taken into consideration when developing a control plan . Not only should overall vector density be considered , but one also should consider vector density during spring/summer months ( April-August ) when villagers could potentially be at greatest risk of exposure to infected sand flies . Clinical VL in Bihar is commonly reported between the months of April and August [30] . Proper bed net usage during the warmer months of the year has been found to be strongly protective against VL [31] . However , several publications suggest that bed net usage in Asia and Africa declines in response to increased temperature [32–37] . Susceptible-Infected-Recovered ( SIR ) compartment models , and variants of this , have been developed in the past to represent VL epidemiology within human populations on the Indian subcontinent . The first such model examined three historical VL epidemic peaks in Assam , India which occurred between 1875 and 1950 and concluded that intrinsic processes related to host and vector dynamics , rather than extrinsic factors such as earthquakes or influenza outbreaks , provided the simplest explanation of the timing of the peaks [38] . More recent models representing VL epidemiology within human populations in Bihar have examined VL underreporting [39] , antimony resistant VL [40] , VL treatment , prevention , and control [41] , and more specific vector control strategies , namely the application of adulticides and destruction of sand fly breeding sites [29] . The latter model represented the application of adulticides and the destruction of sand fly breeding sites via variables that reduced sand fly life expectancy and breeding site capacity , respectively , and predicted the impact of reducing vector density on the ability of the pathogen to persist ( as indicated by the basic reproduction number Ro [42] ) . The model originally published by [40] and subsequently used in [39] and [29] is a deterministic SIR-type model that focuses heavily on the natural history of VL infection within human populations , represented by 11 distinct stages . However , the vector ( sand fly ) population is represented by only three stages: the susceptible , latent , and infectious , with abundance of the latter used to calculate the VL transmission rate to humans . Emergence rate of susceptibles and mortality rates of each stage are held constant . The egg , larval , and pupal stages of the sand fly life cycle are not represented in this model , or in any other SIR-type VL model to the best of our knowledge . These limitations to current SIR-type models have been recognized and the exploration of individual-based , stage-structured , stochastic modelling approaches has been recommended [29] , which could allow explicit evaluation of stage-specific impacts of vector control strategies on sand fly populations in Bihar . In this paper we describe an individual-based , stochastic , stage-structured model that represents a temperature-driven sand fly vector population within a village in Bihar , India and simulates the effects of vector control through the use of fipronil-based drugs orally administered to cattle . The model does not include a human population or VL pathogen , but rather focuses on the effects of fipronil-induced mortality of larval and adult life stages on sand fly population dynamics . We first describe the model and evaluate its performance . We then use the model to simulate several fipronil-based control schemes in which we vary treatment frequency and timing of treatment application , focusing on resulting reductions in sand fly populations during spring/summer and especially during the period of peak human exposure ( June-August ) . We also examine sensitivity of model predictions of treatment efficacy to parametric uncertainty .
The model represents the lifecycle of sand flies as they develop from eggs to larvae to pupae to pre-reproductive adults to pre-oviposition adults to reproductive adults to post-reproductive adults , as well as fipronil-induced larval and adult mortality ( Fig 1 ) . Rates of development , natural mortality , and reproduction depend on the environmental temperatures to which the sand flies are exposed . Eggs , larvae , and pupae are exposed to temperatures of the organic matter in which they develop , whereas adults are exposed to ambient temperatures . Natural mortality of larvae also depends on the density of larvae in the organic matter in which they are feeding . Fipronil-induced mortality occurs in adult flies that obtain a blood meal from fipronil-treated cattle , and in larvae that feed on feces from fipronil-treated cattle . Simulations are run on a daily time step , thus all rates and probabilities described below are calculated on a daily basis . Eggs , larvae , and pupae are represented as daily cohorts whereas adults are represented as individuals . That is , the size of each daily cohort of eggs that enter the system is monitored as these eggs develop into larvae and then into pupae . When a cohort of pupae develops to the adult stage , the resulting adults are represented as individual organisms and are followed through pre-reproductive , pre-oviposition , reproductive , and post-reproductive stages ( only adult females are represented in the model ) . Below we present the equations used in the model to represent the development , reproduction , natural mortality , and fipronil-induced mortality of sandflies . To calculate rates of development of immature stages ( eggs , larvae , pupae ) , we drew upon results of laboratory experiments conducted under constant temperatures [16 , 18 , 19] and then estimated temperature-dependent development under variable temperature regimes using the general equation described by [43]: 100/nl = K/[1+ exp ( a − bx ) ] . This is a bisymmetrical , sigmoid curve with the distance between the lower and upper developmental temperature thresholds ( K ) estimated as K=[2C1C2C3−C22 ( C1+C3 ) ]/ ( C1C3−C22 ) , where C1 , C2 , and C3 are values for 100/nl on the curve at three temperatures on the abscissa . We represented the temperature-dependent development of eggs , larvae , and pupae as: Ci , Eggs=0 . 5/[1+exp ( −0 . 1601⋅Ti , O+5 . 6067 ) ] ( 1 ) Ci , Larvae=0 . 0688052/[1+exp ( −0 . 4754⋅Ti , O+11 . 298 ) ] ( 2 ) Ci , Pupae=0 . 25/[1+exp ( −0 . 2736⋅Ti , O+7 . 7067 ) ] ( 3 ) where Ci , Eggs , Ci , Larvae , and Ci , Pupae represent the contributions of the current daily temperature on day i toward the development of eggs , larvae , and pupae , respectively , and Ti , O represents current temperature ( °C ) within the organic matter on day i ( Fig 2a–2c ) . The model accumulates Ci over time separately for each cohort , and when ΣiCi = 1 . 0 for a given cohort , the organisms in that cohort advance to the next developmental stage ( Fig 1 ) . After pupation , pre-reproductive adults must obtain a blood meal to advance to the pre-oviposition stage ( Fig 1 ) . We estimated the daily probability of obtaining a blood meal based on laboratory experiments in which 0 , 3 , 60 , 85 , 94 , and 96% of flies obtained their first blood meal by the end of their first , second , third , fourth , fifth , and sixth day , respectively , as an adult [44] , and used these results to develop the following curve: Pi , Blood Meal=0 . 940321952 /[1+exp ( −3 . 7061⋅Di , PE+10 . 551 ) ] ( 4 ) where Pi , Blood Meal is the probability of a pre-reproductive adult obtaining a blood meal on day i and Di , PE is the number of days-post-emergence from pupation ( Fig 3a ) . We estimated the temperature-dependent development of adults from the pre-oviposition stage to the reproductive stage based on laboratory data collected by [19] in the same manner as described above for eggs , larvae , and pupae: Ci , POAdults=0 . 363755/[1+exp ( −0 . 1503⋅Ti , A+4 . 0206 ) ] ( 5 ) where Ci , POAdults is defined and calculated in the same manner as the analogous terms in Eqs 1 through 3 , except that Ti , A represents current air temperature ( °C ) on day i rather than temperature within organic matter ( Fig 2d ) . When ΣiCi = 1 . 0 , flies advance from the pre-oviposition to the reproductive stage ( Fig 1 ) . Females lay eggs the day they advance from the pre-oviposition to the reproductive stage . We represented the number of eggs laid per reproductive female ( Eqs 6 and 7 ) as a function of temperature ( Fig 3b ) based on laboratory observations [19]: If Ti , A ≤ 28 . 5 then Pi , OElAdults=25 . 1464684/[1+exp ( −0 . 5238⋅Ti , A+12 . 441 ) ] ( 6 ) If Ti , A > 28 . 5 then Pi , OElAdults=25 . 1464684−25 . 1464684/[1+exp ( −0 . 5238⋅Ti , A+12 . 441 ) ] ( 7 ) where Pi , OElAdusts represents the number of eggs laid by a female on day i and Ti , A represents the current air temperature ( °C ) on day i . However , no eggs are laid if Ti , A ≤ 15C [19] . After oviposition , reproductive females have a 90% chance of becoming post-reproductive and a 10% chance of returning to the pre-reproductive stage [17] . If they return to the pre-reproductive stage , the daily probability of obtaining another blood meal is calculated using Eq 4 , except Di , PE is redefined as the number of days since returning to the pre-reproductive stage . Natural mortality of cohorts of eggs , larvae , and pupae depend on the temperature ( Ti , O ) of the organic matter ( Fig 4a–4c ) in which they are located , whereas natural mortality of adults depends on air temperatures ( TiA ) ( Fig 4d ) . We represented the temperature-dependent natural mortality of eggs , larvae , pupae , and adults based on laboratory experiments conducted by [16 , 45]: Pi , MEggs=0 . 00052737⋅Ti , O2−0 . 02872971⋅Ti , O+0 . 39946900 ( 8 ) If Ti , A ≤ 28 . 5 then Pi , MlLarvae=0 . 3898*exp ( −0 . 156*Ti , O ) ( 9 ) If Ti , A > 28 . 5 then Pi , MuLarvae=0 . 0000000000144*exp ( 0 . 68195*Ti , O ) ) ( 10 ) Pi , MPupae=0 . 00004973⋅Ti , O2−0 . 00261400⋅Ti , O+0 . 03635092 ( 11 ) If Ti , A ≤ 10 then Pi , Ml Adults=0 . 5556*exp ( −0 . 239*TiA ) ( 12 ) If Ti , A > 10 then Pi , Mu Adults=0 . 0005*exp ( 0 . 1918*TiA ) ( 13 ) where Pi , MEggs , Pi , MlLarvae , Pi , MuLarvae , Pi , MPupae represent the proportion of eggs , larvae , and pupae dying on day i and Pi , Ml Adults and Pi , Mu Adults represent the daily probability of dying for adults on day i . Independent of temperature , we estimated the maximum longevity of adults in the wild to be 30 days based on findings presented by [46] . We also represented density-dependent natural mortality of larvae based on rates of cannibalism observed in laboratory experiments conducted with different larval densities [47]: Pi , MLarvae_C= ( ri , Can ) *Ni , Larvae ( 14 ) where Pi , MLarvae_c represents the proportion of larvae dying on day i due to cannibalism , Ni , Larvae represents the number of larvae in the system on day i , and ri , Can represents a proportional increase in cannibalism as the number of larvae increases . The data presented by [47] suggest an approximately linear relationship between larval density and rate of cannibalism , the slope of which ( ri , Can ) we calibrated , as described in the Model evaluation and Model calibration sections below . In addition to depending on the frequency of treatment application and the proportion of the cattle treated , which we represented as management variables , rates of fipronil-induced mortality depend on ( 1 ) the proportion of adult sand flies that feed on cattle , ( 2 ) the proportion of larvae that feed in organic matter containing cattle feces , ( 3 ) the efficacy of fipronil contained in the blood of cattle , and ( 4 ) the efficacy of fipronil contained within cattle feces . We assumed that 50% of adult flies obtain their blood meal from cattle [9] and that 90% of eggs are laid on , and hence larvae develop in , organic matter containing cattle feces [13] . We represented the efficacy of fipronil within the blood of cattle as decreasing exponentially as a function of the number of days after fipronil application: P′i , MAdults=0 . 515 ⋅exp ( −0 . 094⋅Di , PT ) ( 15 ) where P’i , MAdults represents the daily probability of dying for an adult fly that obtained a blood meal from treated cattle Di , PT days post-treatment ( days after application of fipronil ) [25] ( Fig 5a ) . Once an adult obtains a blood meal from a treated cow , we assumed that its daily probability of dying due to fipronil did not change , that is , efficacy of the fipronil within the fly remained constant . We represented the proportion of larvae dying due to fipronil within cattle feces as decreasing exponentially as a function of the number of days post-defecation [28 , 48]: P′i , Larvae=Ej⋅exp ( −0 . 00545⋅Di , PD ) ( 16 ) where P’i , MLarvae represents the proportion of larvae dying on day i that are feeding on feces of treated cattle Di , PD days post-defecation ( Di , PD days after the feces were deposited ) , assuming the feces were deposited j days after application of fipronil . The initial ( maximum ) efficacy of fipronil in cattle feces ( Ej ) ( Eq 17 ) itself decreases exponentially over time [25]: Ej=0 . 567⋅exp[−0 . 073 ( Di , PT−1 ) ] ( 17 ) For example , fresh feces deposited 1 day after cattle are treated have a higher efficacy than fresh feces deposited 2 days after cattle are treated ( Fig 5b ) . We assumed that fipronil-induced mortality and natural mortality were completely additive . To evaluate the potential usefulness of the model in simulating the population-level response of sand flies to fipronil-induced mortality , we first verified that the model simulated adequately the rates of development , reproduction , natural mortality , and fipronil-induced mortality observed under laboratory conditions . That is , that the model code produced simulated data that mimicked the laboratory data from which it was parameterized when we simulated the laboratory experiments . We next calibrated the model to represent environmental conditions typical of Bihar , India such that the simulated population established a seasonally-varying , dynamic equilibrium under baseline conditions ( without fipronil-induced mortality ) . We then evaluated performance of the baseline model by ( 1 ) assessing the ecological interpretability of seasonal trends in the simulated sand fly life cycle and ( 2 ) comparing simulated fluctuations in abundance of adult sand flies to fluctuations observed in each of three villages in Bihar over a 12-month period . We ran 10 replicate stochastic ( Monte Carlo ) simulations for each portion of the model evaluation procedure , except for the simulations required for verification of the temperature-dependent development and mortality rates of eggs , larvae , and pupae , which were deterministic .
We calibrated the model to represent environmental conditions typical of Bihar , India by representing annual fluctuations in ( 1 ) simulated air temperatures ( Ti , A ) with a time series of 365 minimum daily air temperatures recorded at a village in Bihar [21] and ( 2 ) simulated temperatures within organic matter ( Ti , O ) by fitting a cosine curve to a graphical representation of annual fluctuations in soil temperatures in West Bengal , India presented by [49] ( Fig 10 ) . We further calibrated the model by adjusting the parameter controlling the density-dependent mortality of larvae due to cannibalism ( 9 . 5 x 10−7 ) such that the simulated population established a seasonally-varying , dynamic equilibrium under baseline conditions ( without fipronil-induced mortality ) in which the mean annual abundance of adults was approximately equal to the breeding site capacity , or number of vectors ( 7 , 344 ) estimated by [29] . We evaluated the baseline model by ( 1 ) assessing the ecological interpretability of seasonal trends in the simulated sand fly life cycle and ( 2 ) comparing simulated fluctuations in relative abundance of adult sand flies to fluctuations in relative abundance of adults caught in light traps in each of three villages in Bihar over a 12-month period using a Sign Test . Simulated seasonal trends were representative of the general temperature-dependent trends characteristic of the sand fly life cycle in Bihar ( Fig 11 ) . Simulated oviposition did not occur until mid-February ( day-of-year 42 ) , when temperatures first exceeded the 15 C threshold suggested by [19] , with the first mass emergence of adults occurring 85 days later during May ( day-of-year 127 ) , and the largest peak in adult abundance occurring during the latter portion of July ( day-of-year 205 ) , as observed by [21] . Simulated egg , larval , and pupal fluctuations were impossible to validate due to lack of field data , but showed similarity to simulated adult fluctuations ( Fig 12a–12d ) . The fluctuations of egg and pupal population densities were more distinct because the developmental periods are considerably shorter than that of larvae and adults . Simulated fluctuations in relative abundance of adults were not significantly different from the general trends in relative abundance of adults caught in the three villages in Bihar ( sign test: p < 0 . 1263 , p < 0 . 5000 , and p < 0 . 0704 , respectively ) , although , not surprisingly , trends in the field samples were less distinct [21] ( Fig 13 ) . Interestingly , trends in relative abundance at one village ( Mohammadpur ) were markedly different from both the simulated trends and the trends observed at the other two villages ( p < 0 . 10 ) , most likely due to markedly lower abundances during September , October , and November .
Simulation results suggest that the success of fipronil treatments in controlling sand flies depends not only on the frequency of applications but also on the timing of applications relative to the sand fly lifecycle . Synchronizing applications to maintain high efficacy of the drug in cattle feces during the period of high larval abundance seems particularly important . While more frequent applications obviously are more efficacious , they also are more expensive and more difficult logistically . Thus , the ability to assess not only efficacy of treatment schemes per se but also their cost-effectiveness and their logistical feasibility is of paramount importance . Adequate a priori assessment of novel control schemes targeted at specific aspects of a vector’s life cycle requires novel approaches , including models that explicitly represent key aspects of the processes by which the control method intervenes in the life cycle of the target species . Several previous studies of VL epidemiology have focused on villages in Bihar and have included models with detailed representations of disease dynamics within human populations [29 , 39–41] . One study modeled the effect of specifically-targeted sand fly control strategies including the application of adulticides and the destruction of breeding sites , which were represented by reducing sand fly life expectancy and breeding site capacity , respectively [29] . Their model predicted that either reducing vector density >67% through application of adulticides or >79% through breeding site destruction could eliminate the ability of the pathogen to persist , as indicated by the value of the basic reproduction number ( Ro < 1 . 0; [42] ) . Although providing a wealth of details concerning disease dynamics within humans and valuable information regarding the general magnitude of vector reduction required to control transmission to humans , environmental factors affecting the vector life cycle , and hence the transmission process , were not modeled explicitly [29] . These authors recognized the limitations this imposed on use of their model and provided appropriate caveats [29] . Other studies also have modeled the effect on VL control via direct manipulation of model parameters controlling mortality rate [50 , 51] or biting rate [52] of adult sand flies , again providing valuable information pertinent to objectives of their studies , but without explicit representation of environmental factors affecting the seasonality of such rates . By explicitly representing the effects of seasonally-varying temperatures on development and survival of the various sand fly life stages , our model has allowed initial assessment of a novel control scheme targeted specifically at both larvae and adults . By specifically examining the relationship among the timing and frequency of treatment applications , the duration of drug efficacy , and the seasonality of the sand fly lifecycle , we can make initial assessments not only in terms of reducing average sand fly abundance , but also in terms of cost-effective reduction of human exposure to sand flies given local social practice and availability of alternative hosts . As a common social practice , some family members of the vast majority of Bihari villager households sleep outdoors , particularly during the months with the hottest evening temperatures ( June , July , August ) [21 , 24] . Although indoor residual spraying , when properly applied [22] , and bed net usage [31] offer protection against infected sand flies feeding indoors , outdoor sleeping places a sizeable portion of the population at risk of exposure to infected exophilic sand flies during periods of peak sand fly abundance . Thus reduction of sand fly abundance during the period when villagers are most likely to be exposed to outdoor-feeding sand flies is of particular importance when assessing the efficacy of control schemes . Considering the results from our model in terms of the efficacy of the various treatment schemes to reduce sand fly abundance during this critical period . Economically , Bihar is the poorest state in India , with roughly $100 million gross domestic product compared to the national average of $274 million [53] , and people within the VL-endemic zones are among the most impoverished people in the world [54] . Thus considerations of cost effectiveness become paramount in terms of the commercial feasibility of drug application by villagers in VL-endemic regions . The cost of treating with this form of drug is estimated at approximately $1 per cow per treatment , but milk production per cow is estimated to increase by $0 . 50 per day [55] , offering incentive to villagers to pay to treat their animals . Our simulations assumed 100% of cattle were treated , as would occur in a field trial . However , likely <100% of cattle would be treated if this form of treatment became commercially available , since individual livestock owners would be responsible for its application . In this regard , the influence of regional differences in socio-economic conditions on the efficacy of alternative sand fly control schemes should be evaluated closely . Although beyond the scope of the present study , a future use of our model could include a sensitivity analysis aimed at assessing the impact of regional socio-economic differences on the efficacy of different treatment schemes under different hypotheses regarding local economic conditions and social practice . Potential environmental and human health impacts , as well as effects on non-target species , always are a concern when evaluating new vector control methods . In this regard , treating cattle orally with fipronil-based drugs may have benefits over conventional IRS . IRS often involves the application of insecticides to the walls of homes and cattle sheds , thus exposing human inhabitants as well as non-target species coming into contact with the walls . DDT has known environmental consequences , but chronic exposure also could potentially be linked to human health concerns such as pancreatic cancer [56–58] making the choice to switch to synthetic pyrethroids logical . Fipronil-based drugs can be safely administered to cattle , given the acute oral LD₅₀ for fipronil-fed rats is ≈97mg/kg of body weight [28] , and exposure of cattle to orally applied drugs with fipronil at a concentration of 0 . 5 mg/kg of body weight can provide effective control of adult and larval sand flies . We are not suggesting treatment with fipronil-based drugs as a replacement for IRS , but rather as a complimentary component . The impact of current practices of IRS and bed net administration on the vector population and human VL transmission has been inconclusive and thus there is much interest in alternative control methods and integrated control schemes [22] . By targeting sand flies feeding on cattle outdoors and larvae developing in cattle feces ( areas not targeted by IRS ) , fipronil-based drug treatment may prove to be a potentially important component of an integrated pest management program . Further evaluation of the effects of sand fly control through the use of fipronil-based drugs orally administered to cattle ideally would involve a field trial in Bihar . Among the most critical data obtained from such an experiment in terms of increasing confidence in our model predictions would be those shedding light on the proportion of adult sand flies that obtain their blood meal from cattle and the proportion of eggs oviposited in organic matter containing cattle feces . By far the most restrictive assumptions we have made in our model are that 50% of adult sand flies obtain their blood meal from cattle [9] and that 90% of sand fly eggs are laid on , and hence larvae develop in , organic matter containing cattle feces [13] . Although it would have been desirable to use sand fly field collections [25] and blood meal analysis of field caught sand flies [9] to parameterize equations representing adult and larval sand fly developmental processes , currently this is infeasible . Locations of immature sand flies in the field are largely unknown , making field observations difficult . It is also difficult to ascertain the age at which blood fed sand flies collected from the field acquire a blood meal . In contrast , daily immature P . argentipes and P . papatasi processes and adult blood feeding probability and oviposition can be observed under controlled conditions in the laboratory [16–20 , 44] . The interaction of vector feeding tendencies and host availability on the success of vector control based on systemic insecticides is a topic of current investigation . A recent study modeled the effect of assuming different hypothetical functional relationships between biting behavior of mosquitos ( e . g . , indiscriminate , anthropophilic , zoophilic ) and human host availability on subsequent predictions of malarial infections [59] . This author’s model results indicated that control efficacy was dependent on both intrinsic host preferences and variation in encounter rates with alternative hosts . Our model results assumed that a fixed 50% of sand flies blood fed on cattle treated with fipronil and that the remaining 50% fed on alternative hosts that did not contain fipronil in their blood . Previous blood meal analysis suggests opportunistic feeding behavior of P . argentipes . Research conducted in eight districts in the West Bengal , a state neighboring Bihar , by [10] and in a single district by [12] suggested that P . argentipes blood feeding was not driven by particular host preference but rather by host availability , as they fed primarily on humans in human dwellings and cattle in cattle sheds . Research conducted by [60] suggests P . argentipes feed on cattle primarily , but feed readily on humans in human dwellings , with the researchers referring to P . argentipes as “chance feeders . ” In our model , the probability of a sand fly acquiring a blood meal from cattle implicitly represents relative host availability . The results of our sensitivity analysis in which the mean number of SFD decreased in response to an increased proportion of adults feeding on cattle are indicative of the predicted tendency of indiscriminately feeding mosquitoes presented by [59] . It goes without saying that this treatment is dependent on the presence of cattle . Although P . argentipes feed indiscriminately , they have been reported to be zoophilic in the past as well [11] . Therefore , explicit representation of sand fly feeding tendencies , not unlike that presented by [59] , may be a beneficial addition to our model in the future . Empirical evidence regarding the substrates in which oviposition occurs is sparse . Sensitivity analysis indicated the obvious importance of assumptions which directly affect the exposure of simulated sand flies to the drug . Immature sand flies are typically collected from the floors of cattle sheds and human dwellings in India , but often in small numbers [13–15] . Studies have uncovered sand fly larvae in larger numbers from diverse locations such as forest floors [61] and abandoned sheds [62–64] , and adult sand flies have been collected from palm tree canopies in Bihar [23] , suggesting oviposition may be occurring in a variety of microhabitats . Other data limitations affecting parameterization of the present model included the need to use developmental data from laboratory studies of another species of Phlebotomus rather than field data on our target species . The use of P . papatasi , in addition to P . argentipes , data for model parameterization was necessary because laboratory data for P . papatasi are more abundant and the lifetables developed for them [18–19] are from studies conducted under a wider range of temperatures than those of P . argentipes [16] . The P . papatasi data provide upper and lower thermal limits [45] , and thus provide at least a general basis for predicting how phlebotomine sand flies function at extreme temperatures . Although , similarities have been observed between these two species in terms of development [17] , we suggest that laboratory studies be conducted to further study the temperature-dependence of the incriminated vector of VL on the Indian subcontinent , P . argentipes . Notwithstanding the inevitable parametric uncertainties associated with the current model , we would suggest that our model structure might be adapted for initial evaluation of fipronil-based sand fly control under a range of different environmental conditions involving a variety of potential hosts . For instance , the VL vector in East Africa , Phlebotomus orientalis , is highly zoophilic in Ethiopia but feeds heavily on donkeys in addition to cattle [65 , 66] , suggesting the need to include multiple host species . The hare is a potential reservoir for VL in Spain [67] , suggesting a different route of administration may be required , such as a fipronil-based pour-on or a grain bait , rather than a bolus . Dogs are the primary reservoir for zoonotic VL in the Americas [68] and insecticide impregnated dog collars have shown promise in reducing the VL infection rate in dogs [69 , 70] , suggesting that fipronil impregnated dog collars could provide a means of targeting sand flies in Latin America . Although the host species treated and route of drug administration would need to be modified depending on the situation , the current structure of our model should accommodate the necessary re-parameterizations . While 90% of reported VL cases occur in six countries on the Indian Subcontinent: India , Bangladesh , South Sudan , Sudan , Ethiopia , and Brazil [71] , both VL and cutaneous leishmaniasis ( CL ) also are present in Europe and climatic projections suggest that the Central European climate will become increasingly suitable for sand flies capable of vectoring VL and CL [72] . The large number of refugees fleeing to Europe from countries such as Afghanistan and Iraq , where CL is known to occur and where clinical VL cases have been documented [73] , creates the potential for widespread leishmaniasis outbreaks . Our hope is that our model will prove useful in the a priori evaluation of the potential role of treatment schemes involving the use of fipronil-based drugs in the control of leishmaniasis on the Indian Subcontinent and beyond . | Visceral leishmaniasis is a disease caused by a virulent vector-borne parasite transmitted to man by phlebotomine sand flies . Fipronil-based drugs , administered to cattle orally , provide a potential means of sand fly control by permeating in cattle blood and being excreted in cattle feces , targeting adult females feeding on cattle blood and larvae feeding on cattle feces , respectively . An agent-based , stochastic simulation model was developed to represent sand fly population dynamics in a village in Bihar , India , at all developmental stages , with the goal of predicting the impact of various vector control strategies , utilizing drug treated cattle , on vector population numbers . Results indicate that success of treatment is dependent on the number of treatments applied annually and the seasonality of the sand fly lifecycle . Results further suggest that treatment schemes are most effective in reducing vector populations when high drug efficacy is maintained in cattle feces during periods of high larval density . Our approach incorporates detailed representation of the vector population and provides an explicit representation of the effects of insecticide application on adult and larval sand flies . Hence , this model predicts treatment schemes that may have the greatest potential to reduce sand fly numbers . | [
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] | [
"death",
"rates",
"livestock",
"medicine",
"and",
"health",
"sciences",
"body",
"fluids",
"insect",
"metamorphosis",
"ruminants",
"demography",
"vertebrates",
"sand",
"flies",
"animals",
"mammals",
"developmental",
"biology",
"pupae",
"population",
"biology",
"infectious",
"disease",
"control",
"insect",
"vectors",
"zoology",
"veterinary",
"science",
"veterinary",
"medicine",
"infectious",
"diseases",
"epidemiology",
"livestock",
"care",
"disease",
"vectors",
"hematology",
"agriculture",
"people",
"and",
"places",
"population",
"metrics",
"larvae",
"blood",
"anatomy",
"entomology",
"physiology",
"biology",
"and",
"life",
"sciences",
"metamorphosis",
"cattle",
"amniotes",
"bovines",
"organisms"
] | 2016 | Visceral Leishmaniasis on the Indian Subcontinent: Modelling the Dynamic Relationship between Vector Control Schemes and Vector Life Cycles |
The circadian clock controls many physiological processes in higher plants and causes a large fraction of the genome to be expressed with a 24h rhythm . The transcripts encoding the RNA-binding proteins AtGRP7 ( Arabidopsis thaliana Glycine Rich Protein 7 ) and AtGRP8 oscillate with evening peaks . The circadian clock components CCA1 and LHY negatively affect AtGRP7 expression at the level of transcription . AtGRP7 and AtGRP8 , in turn , negatively auto-regulate and reciprocally cross-regulate post-transcriptionally: high protein levels promote the generation of an alternative splice form that is rapidly degraded . This clock-regulated feedback loop has been proposed to act as a molecular slave oscillator in clock output . While mathematical models describing the circadian core oscillator in Arabidopsis thaliana were introduced recently , we propose here the first model of a circadian slave oscillator . We define the slave oscillator in terms of ordinary differential equations and identify the model's parameters by an optimization procedure based on experimental results . The model successfully reproduces the pertinent experimental findings such as waveforms , phases , and half-lives of the time-dependent concentrations . Furthermore , we obtain insights into possible mechanisms underlying the observed experimental dynamics: the negative auto-regulation and reciprocal cross-regulation via alternative splicing could be responsible for the sharply peaking waveforms of the AtGRP7 and AtGRP8 mRNA . Moreover , our results suggest that the AtGRP8 transcript oscillations are subordinated to those of AtGRP7 due to a higher impact of AtGRP7 protein on alternative splicing of its own and of the AtGRP8 pre-mRNA compared to the impact of AtGRP8 protein . Importantly , a bifurcation analysis provides theoretical evidence that the slave oscillator could be a toggle switch , arising from the reciprocal cross-regulation at the post-transcriptional level . In view of this , transcriptional repression of AtGRP7 and AtGRP8 by LHY and CCA1 induces oscillations of the toggle switch , leading to the observed high-amplitude oscillations of AtGRP7 mRNA .
Circadian clocks are endogenous timekeepers that can be found among all taxa of life [1]–[3] . They are able to generate stable oscillations with a period of approximately 24h that persist even under constant ( free-running ) conditions , i . e . in the absence of any rhythmic environmental influences that impact the clock . Entrainment by environmental signals such as light and temperature can synchronize the clock to the period of the Earth's rotation . Such a clockwork may confer a higher fitness to an organism as it allows to anticipate daily cycles of light and temperature in a spinning world [4] , [5] . Circadian clocks are usually described as molecular networks including ( interlocked ) transcriptional - translational feedback loops [6] . In the higher plants model organism Arabidopsis thaliana an interplay of experiments and mathematical modeling shaped the current view on the circadian clock's network [7]–[13] . Locke et al . first modeled the structure of the circadian clock as a “simple” two-gene negative feedback loop [7] , where the two partially redundant MYB transcription factors LATE ELONGATED HYPOCOTYL ( LHY ) and CIRCADIAN CLOCK ASSOCIATED 1 ( CCA1 ) ( combined to one variable LHY/CCA1 ) inhibit the transcription of their activator TIMING OF CAB EXPRESSION 1 ( TOC1 ) . However , in silico and experimental mutant analysis revealed inconsistencies between the model and data [7] , [8] . The assumed circadian clock architecture was therefore extended in successive steps [8]–[11] from this simple design to the idea of a clockwork that has a repressilator-like architecture at its core [13] . In this recent picture a “morning loop” consists of the morning-expressed genes LHY/CCA1 that activate the transcription of the PSEUDO RESPONSE REGULATORS 9 , 7 and 5 ( PRR9 , PRR7 and PRR5 ) which in turn inhibit the transcription of LHY/CCA1 . Furthermore , LHY/CCA1 is assumed to repress the transcription of the “evening loop” genes EARLY FLOWERING 3 ( ELF3 ) and 4 ( ELF4 ) , LUX ARRHYTHMO ( LUX ) , GIGANTEA ( GI ) , and TOC1 , respectively . ELF3 , ELF4 and LUX form a protein complex ( evening complex , EC ) that inhibits the transcription of PRR9 , thereby connecting the evening loop with the morning loop , which closes the feedback loop circuitry [14] . The circadian clock affects many physiological processes in Arabidopsis thaliana , including the oscillation of free cytosolic calcium [15] , stomatal opening , cotyledon and leaf movement [16] , and even enables the plant to measure day-length , track seasons and thereby triggers the onset of flowering [17] . Underlying these physiological rhythms is a widespread control of gene expression by the circadian clock [18] . However , it is still not completely understood how the rhythmicity of the circadian clock is transmitted to its output genes . This may occur either directly by binding of clock proteins to their target genes or indirectly via signal transduction chains . One possibility to maintain the rhythmicity along such a signal transduction chain could be via slave oscillators that are driven by the circadian core oscillator and shape their oscillatory profile due to negative auto-regulation . Colin Pittendrigh already proposed in 1981 that “ any feedback loop in the organism is a potential slave oscillator , and if the circadian pacemaker can make input to the loop , the slave will assume a circadian period and become part of the temporal program that the pacemaker drives” [19] . Genetic variation in such a slave oscillator can change its properties , e . g . the phase relation to the core oscillator , and thus the organisms' “ temporal program is open to evolutionary adjustment” [19] without the need for change in the core oscillator itself . Since driven by the core oscillator , the slave oscillator does not have to share all of the core oscillator's properties: It is not necessary that the slave oscillator exhibits independent self-sustaining oscillations , shows temperature compensation , or gains direct input from light [19] , [20] . On the other hand , an indispensable pre-requisite of a slave oscillator is that it must not to act in any way back onto the core oscillator . The two RNA binding proteins Arabidopsis thaliana Glycine Rich Protein 7 and 8 ( AtGRP7 and AtGRP8 ) , also known as Cold and Circadian-Regulated 2 and 1 ( CCR2 and CCR1 ) , respectively , have been proposed to represent such a molecular slave oscillator [21]–[23] . These proteins share 77 percent of sequence identity and contain an approximately 80 amino acid long RNA-recognition motif at the amino-terminus and a carboxy terminus mainly consisting of glycine repeats [21] , [24] . The transcripts of both genes undergo circadian oscillations with evening peaks . The maximum of AtGRP8 slightly precedes that of AtGRP7 by 1–2 hours [25] . The AtGRP7 protein oscillates with a four hour delay compared to its transcript [22] . In plants constitutively over-expressing CCA1 [26] or LHY [27] , AtGRP7 mRNA oscillations are dampened under constant light conditions , approaching the trough value of their corresponding oscillations in wild type plants , and thus suggesting that the transcription of AtGRP7 is rhythmically repressed rather than activated by these partially redundant core oscillator genes . Apart from this transcriptional regulation AtGRP7 also negatively auto-regulates the steady-state abundance of its own mRNA via a post-transcriptional mechanism [28] . When AtGRP7 protein levels are high , an alternatively spliced transcript is produced at the expense of the fully spliced mRNA [22] . This alternative splice form is generated through the use of an alternative 5′ splice site and retains part of the intron . Due to a premature termination codon this alternatively spliced transcript cannot be translated into functional protein and is rapidly degraded via the nonsense-mediated decay ( NMD ) pathway [28] , [29] . Since AtGRP7 binds to its own transcript in vitro and in vivo , this alternative splicing likely is promoted by direct binding of AtGRP7 to its own pre-mRNA [30] , [31] . AtGRP8 also auto-regulates itself and both proteins cross-regulate each other by the same mechanism . Our regulatory network is therefore composed of two auto-regulatory negative feedback loops , interlocked with each other and driven by the circadian core oscillator , as depicted in Figure 1 . Apart from the negative auto-regulation , AtGRP7 affects the accumulation of a suite of circadian clock regulated genes in a time-of-day dependent manner , supporting the hypothesis that it acts as a slave oscillator between the core oscillator and the clock output: Rhythmic transcripts , whose steady state abundance is reduced upon AtGRP7 overexpression , peak in the evening like AtGRP7 itself , whereas rhythmic transcripts with an elevated steady state abundance peak out of phase towards the morning [32] . Furthermore , it has been shown that AtGRP7 has an impact on various other physiological processes: It promotes the floral transition [33] , plays a role in the plants innate immune system [34] , [35] , and is known to mediate responses to stresses such as oxidative stress , high salt , mannitol , or cold [21] , [36] , [37] . Recently , various mathematical models for the circadian core oscillator in Arabidopsis thaliana have been developed [7]–[13] . In this paper we model the AtGRP7 and AtGRP8 feedback loops in terms of ordinary differential equations and thus propose the first mathematical model of a molecular slave oscillator in Arabidopsis thaliana . We note that a related model of a clock-controlled system has been put forward by Salazar et al . [38] . The molecular components of this system do not incorporate any feedback mechanism and are therefore unable to reshape their own oscillatory profile . Thus , they do not adopt all of the above mentioned specifications of a slave oscillator .
In order to model the essential layers of AtGRP7 and AtGRP8 regulation we need six dynamical variables , namely the concentrations of the pre-mRNA ( ) , mRNA ( ) , and protein ( ) of AtGRP7 and AtGRP8 . In the absence of any measured data that distinguish between cytoplasmic and nuclear protein concentrations , we , in particular , do not take into account that AtGRP7 and AtGRP8 localize to both the nucleus and the cytoplasm [39] , [40] , as it was done e . g . , in [11] . The driving force of the AtGRP7 oscillations is the periodic change in protein concentration of the core oscillator components LHY/CCA1 , combined into one variable . Throughout the first part of the paper we adopt the previously established mathematical model of Pokhilko et al . [11] . In principle , one could also use any other time periodic function or generic oscillator model that properly imitates the observed protein concentration for a given experimental situation . Two examples of this type are a modified Poincaré oscillator and the refined model of Pokhilko et al . [13] as considered towards the end of our paper ( see section Robustness Against Variations in the LHY/CCA1 Protein Oscillations ) . The original model provided by Pokhilko et al . [11] involves dynamical variables and parameters whose quantitative values are taken over from that paper . Likewise , we utilize the same specific initial conditions for the core oscillator as in [11] . The externally imposed light input consists of either constant light ( LL ) or diurnal conditions such as 12 hours of light and 12 hours of darkness ( abbreviated as ) or 8 hours of light and 16 hours of darkness ( ) , also denoted as short day conditions . These light conditions enter our core oscillator dynamics as detailed in [11] ( especially continuous transitions instead of binary , i . e . on–off , light-dark transitions are used ) . Typical examples of the protein concentrations obtained in this way are depicted as dashed lines in Figure 2 . In view of the fact that the AtGRP7 mRNA steady state abundance seems not to be light-induced ( unpublished data ) we assume no direct light effect on the slave oscillator . This assumption is also coherent with Pittendrigh's definition , proposing that the slave oscillator could receive the light input only indirectly via the core oscillator [19] . Given the input of the core oscillator to the AtGRP7 and AtGRP8 feedback loops , we model the temporal evolution of the slave oscillator's dynamical variables , , , , , and as follows ( 1 ) ( 2 ) ( 3 ) ( 4 ) ( 5 ) ( 6 ) Consistent with other circadian clock models [7]–[11] , in the first term on the right-hand-side of equation ( 1 ) we use a sigmoidal Hill repressor function , describing the negative regulation of AtGRP7 transcription by LHY/CCA1 . The pertinent transcription rate of AtGRP7 is then given in terms of the maximal transcription rate , the Hill coefficient , the activation coefficient , and the LHY/CCA1 protein concentration . The loss term in equation ( 1 ) describes the normal and alternative splicing of AtGRP7 pre-mRNA . It is assumed that the AtGRP7 pre-mRNA is either spliced into its mature mRNA or into its alternative splice form , without considering any further degradation pathway . The kinetics for the splicing of the AtGRP7 pre-mRNA into its alternative splice form , promoted by the binding of AtGRP7 protein to its own pre-mRNA , is assumed to depend on the splicing coefficient and the concentrations of the AtGRP7 pre-mRNA ( ) and protein ( ) . Equivalent kinetics are used for the alternative splicing of AtGRP7 pre-mRNA promoted by the binding of AtGRP8 protein . Note that is the coupling parameter between AtGRP8 and AtGRP7 , i . e . the impact of AtGRP8 on alternative splicing of the AtGRP7 pre-mRNA . The normal splicing of AtGRP7 pre-mRNA into its mature mRNA is supposed to depend on a splicing coefficient as well as the pre-mRNA concentration and appears as the gain term in the first part of equation ( 2 ) . The second part of equation ( 2 ) describes the mRNA degradation as Michaelis-Menten kinetics that account for saturation by means of the Michaelis constant and the maximal degradation rate . A similar Michalis-Menten degradation appears in equation ( 3 ) , while describes the translation of mRNA into protein . Analogous considerations apply to equations ( 4 ) – ( 6 ) , modeling AtGRP8 . As usual , all the kinetic parameters in ( 1 ) – ( 6 ) are tacitly restricted to positive real values . Collecting all kinetic parameters into a vector and the six dynamical variables into a vector with components , , equations ( 1 ) – ( 6 ) can be written in the form of a parameterized non-autonomous dynamical system ( 7 ) where the explicit dependence on time is a consequence of the external driving term in ( 1 ) and ( 4 ) . In analogy to [7] , we use the value as initial conditions for all six dynamical variables in ( 1 ) – ( 6 ) . Then , we numerically solve equations ( 1 ) – ( 6 ) for 14 days under 12 h∶12 h LD ( entrainment ) conditions followed by 13 days under constant light ( free-running ) conditions ( see Methods for further details ) . In general the solutions are different for every parameter set . As it is often the case in biological modeling , none of these parameters is known from experiments [7]–[10] . So , the remaining challenge is now to identify the specific parameter set for which the solution reproduces as well as possible the following known ( sparse and often noisy ) experimental findings: 1 . Both transcripts perform periodic oscillations with the same period as the core oscillator , both under LD and LL conditions [25] . 2 . The transcript oscillations exhibit evening peaks with the peak of AtGRP8 preceding that of AtGRP7 by approximately 1–2 hours [22] , [25] , [29] . The corresponding AtGRP7 protein concentrations oscillate with an approximately four hour delay compared to the transcript [22] . 3 . The amplitudes of their oscillations are roughly comparable to those of the core oscillator [25] . 4 . The waveform of the mRNA and protein oscillations have been characterized by means of experimental time series [22] , [23] , [25] . 5 . AtGRP7 mRNA is reduced to 50% within hours after experimentally suppressing its transcription [28] . In order to find an optimal parameter set , we defined a cost function ( described in detail in Text S1 A ) which quantifies the deviation of the corresponding solution from these experimental findings 1–5 for every given parameter set . In a next step we minimized this cost function with respect to . The detailed optimization procedure is described in Methods . Here we only summarize the main steps: To take into account the similarity of the two paralogous proteins AtGRP7 and AtGRP8 we first sampled the parameters for a reduced system , only consisting of AtGRP7 , using two million Antonov-Saleev quasi-random parameter sets . The network motif was then extended to the complete interlocked feedback loop structure , including also AtGRP8 . The parameters were chosen in order to generate two identical oscillatory profiles for AtGRP7 and AtGRP8 . The best one hundred solutions were then further optimized in the local neighborhood of a given parameter set using a Nelder Mead downhill simplex algorithm [41] . This modified sampling and optimization method led to better results than the full parameter space sampling and optimization , i . e . the best solutions have a lower cost function value and thus better fit the experimental data ( compare Figure 3 A ( discussed in the next section ) and Figure S1 ) . It might also reflect a possible evolutionary origin of that network motif since the high sequence similarity of AtGRP7 and AtGRP8 suggests that these genes are paralogues , originating from a gene duplication event [42] , [43] . Two genes that mutually repress each other by transcriptional inhibition are known to constitute a genetic toggle switch – a prototypical example of a biological system showing bistability [49] . Gardner et al . reconstructed such a toggle switch in Escherichia coli and proposed a two variable model ( Gardner model ) in order to explain the necessary conditions for bistability [50] . In the system studied here , both genes , AtGRP7 and AtGRP8 , also cross-regulate each other . However , the reciprocal regulation of AtGRP7 and AtGRP8 occurs at the post-transcriptional level via alternative splicing followed by nonsense-mediated decay of the alternative splice forms instead of mutual inhibition of transcription . This led us to the question whether the slave oscillator could act as a toggle switch . Therefore , we decoupled the slave oscillator from the core oscillator by setting for all times , thus neglecting the transcriptional repression of AtGRP7 and AtGRP8 by LHY/CCA1 . In other words AtGRP7 and AtGRP8 are now transcribed at constant rates and , respectively , see also text below equation ( 6 ) . Note that both AtGRP7 and AtGRP8 show negative auto-regulation as an additional feature not described for the toggle switch as proposed in [50] . As a first step , we investigated whether our decoupled slave oscillator system ( i . e . equations ( 1 ) – ( 6 ) with ) can exhibit bistability . While we show in Text S1 B that the simplified model with a single AtGRP7 feedback loop can only have one fixed point ( either stable or unstable ) , the interlocked AtGRP7 and AtGRP8 feedback loop may give rise to bistability , i . e . a scenario where two stable steady states can coexist: In order to test the system's ability to show a bistable behavior , we randomly sampled parameter sets in the same range as before . A linear stability analysis applied to every fixed point of a given parameter set ( see Methods ) revealed a monostable longterm behavior in of all cases , bistability in , and oscillatory behavior in . Such oscillations were not possible in the two-variable model by Gardner et al . [50] . Moreover , we found that a tiny rest of about exhibited still other phase space structures , such as the coexistence of a stable fixed point and a limit cycle attractor . Figure 4 A illustrates the situation when only the two parameters and are varied , while all other parameters are kept at their values from Table 1 . Such variations of and are of particular interest since they effectively correspond to variations of at fixed and in equations ( 1 ) and ( 4 ) : The transcription of AtGRP7 and AtGRP8 is repressed whenever and the corresponding transcription rates and ( see text below equation ( 6 ) ) adopt values smaller than their maximal transcription rates and . For our optimal parameter set from Table 1 , the system shows bistability ( see intersection of the dashed lines in Figure 4 A ) . Similar to the Gardner model [50] a bistable region separates two monostable regimes in Figure 4 A . In those two monostable regions either high AtGRP7 fixed point protein concentrations dominate over AtGRP8 fixed point protein concentrations or vice versa ( Figures 4 C and D ) . The one parameter bifurcation diagrams , following the dashed lines in Figures 4 A , show the typical hysteretic behavior of a toggle switch ( Figures S6 ) . Intuitively understandable , the and protein fixed point concentrations increase with increasing maximal transcription rates and , respectively . In the Gardner model [50] the degree of cooperativity of the reciprocal transcriptional inhibition determined the slope of the bifurcation lines and therefore the size of the bistable region . In our case , the strength of the reciprocal control of alternative splicing ( ) has an analogous effect , as one can see in Figure 4 B . An increase of the splicing coefficient nearly exclusively alters the slope of the bifurcation line bordering the monostable region where AtGRP8 protein dominates , and similarly for . Gonze already showed in 2010 that periodically forcing the transcription of one of the two genes in the Gardner model can induce limit cycle oscillations [51] . Specifically , high forcing amplitudes can drive the system from one monostable region to the other by crossing the bistable regime . In our system , the LHY/CCA1 protein in equations ( 1 ) and ( 4 ) was assumed to affect both AtGRP7 and AtGRP8 transcription . We therefore have to investigate this phenomenon in a two parameter bifurcation diagram . Indeed , if we pursue the trajectory of the transcriptional rates and ( see text below equation ( 6 ) ) of AtGRP7 and AtGRP8 ( black curved line in Figure 4 A ) during one cycle under 12h∶12h LD conditions , one observes that the rhythmic transcriptional repression via LHY/CCA1 drives the system from one monostable region to the other by crossing a narrow bistable branch . This is only possible due to different kinetics of AtGRP7 and AtGRP8 transcription ( see Table 1 ) . Completely identical transcription kinetics for AtGRP7 and AtGRP8 ( i . e . , , , and therefore ) would lead to a straight line of unit slope in the – bifurcation diagram instead of the curved shape , not allowing the system to reach one monostable region from the other . In order to examine the effect of variations in the core oscillator input on the AtGRP7-AtGRP8 slave oscillator we substituted the core oscillator protein concentrations obtained from the model of Pokhilko et al . [11] by obtained from a modified Poincaré oscillator , similar to the model used in [55] . This generic oscillator , described in detail in the section Methods , is tunable in its period and amplitude . A third parameter determines the shape of the oscillations , ranging from sinusoidal ( ) to increasingly spiky oscillations with increasing , and a fourth parameter determines the trough value . In particular , for , , , and , the resulting oscillations are very similar to under 12h∶12h LD conditions ( see black lines in Figure 6 A ) . Likewise , the corresponding slave oscillator dynamics differ only little from those obtained by a coupling of the AtGRP7-AtGRP8 feedback loops to the more complex core oscillator model [11] , as one can see in Figure 6 A . In other words , we can replace the complex core oscillator model , being composed of many differential equations and parameters , by any other model which faithfully imitates the actual protein oscillations of LHY/CCA1 . In particular , we verified that almost identical solutions for the slave oscillator dynamics are recovered ( exemplified for by Figure S8 ) , when we replace our original model from [11] by the recently published refined core oscillator model from [13] . While shape and phase of the LHY/CCA1 protein oscillations are fairly similar in both core oscillator models , the amplitude of approximately doubles for the refined model from [13] . As expected from equations ( 1 ) and ( 4 ) , adapting the activation coefficients according to and then results in almost identical results for the slave oscillator , see Figure S8 . It is known that oscillations , governed by a hysteretic switch mechanism , exhibit oscillations with a robust amplitude , mainly determined by the height of the hysteretic loop , while being easily tunable in their period [56] , [57] . In order to investigate the effect of changes in the LHY/CCA1 protein concentrations , and whether our driven AtGRP7-AtGRP8 slave oscillator shows robust amplitudes for varying as well , we examined the behavior of the system for different amplitudes and waveforms of the core oscillator while keeping and constant . Figure 6 B shows the color-coded values of amplitude obtained from simulations with different and . For a given shape parameter , the amplitude of oscillations nearly stays constant after reaching a certain driving amplitude even if we further increase , i . e . the values of are strong enough to overcome the bistable region and to repress the system to a trough value of the oscillations close to zero . This threshold amplitude increases for more spiky oscillations with increasing since the timespan of the transcriptional repression becomes shorter and the systems dynamics needs time to react to the corresponding “movement” in the – bifurcation diagram in Figure 4 A ( similar diagrams can be obtained for the other concentration species , , , , and ) . Nevertheless , oscillations of robust amplitudes can be induced for a wide range of combinations of and ( see red area in Figure 6 B ) . The lhy cca1 double mutant does not express LHY and CCA1 , hence the protein concentration of the core oscillator must vanish . We have shown in the previous section that the resulting autonomous dynamical system ( in equations ( 1 ) – ( 6 ) ) approaches a steady state in the bistable region , see Figures 4 A and 7 A , i . e . oscillatory solutions are ruled out . This theoretical result is in contradiction to the experimental finding that the AtGRP7 transcript shows diurnal oscillations with a phase shift to dawn in the lhy cca1 double mutant [58]–[60] . As a first possible resolution of this contradiction we considered the possibility of modifying the kinetic parameters of Table 1 without changing our model ( 1 ) – ( 6 ) itself in order to generate oscillatory solutions of the autonomous dynamics ( ) . As demonstrated by Figure 7 B and detailed in Text S1 C this is indeed possible but the obtained periods of oscillation are prohibitively small . Moreover , tiny parameter variations in an ensemble of autonomous oscillators will lead to deviating oscillation periods and hence the oscillations average out in the longterm . Next we considered the possibility to explain the experimental facts by means of noise effects . Indeed , noise is omnipresent in biological systems due to the probabilistic nature of molecular reactions or fluctuating environmental influences [61] , [62] and noise induced oscillations have been reported in numerous other models [63]–[65] . Again , as shown in Figures 7 C/D and detailed in Text S1 D , we were able to generate noise-induced self-sustained oscillations on the single cell level , but not in the ensemble . An obvious remedy in both our attempts discussed above is to introduce coupling between the individual oscillators . However , in the experimentally relevant case of many cells the details of their mutual interaction are still not fully clarified , but a global synchronization mechanism seems unlikely [66]–[68] . Moreover , we note that both our attempts are also unable to explain one more experimental fact , namely the entrainment of AtGRP7 mRNA oscillations to 24h-periodic light-dark cycles in the lhy cca1 double mutant [59] , [60] . In conclusion , the only remaining possibility to explain the observed rhythmicity of AtGRP7 mRNA in lhy cca1 double mutants seems to include to the model ( 1 ) – ( 6 ) additional influences of the core oscillator variables ( as already stated , a direct influence of light seems negligible ( unpublished data ) ) , e . g . additional transcriptional activators or inhibitors . We introduced and analyzed a mathematical model for the molecular regulatory network of the AtGRP7 and AtGRP8 slave oscillator in Arabidopsis thaliana . Based on experimental results , we assumed that the slave oscillator gains input from the circadian core oscillator via transcriptional repression by the LHY/CCA1 proteins . Furthermore , we assumed that it shapes its oscillatory profile due to a negative auto-regulation and reciprocal cross-regulation between AtGRP7 and AtGRP8 via alternative splicing followed by nonsense-mediated decay of the alternative splice form . Although alternative splicing is abundant among circadian clock genes [69] , [70] , this is as far as we know the first mathematical model of a circadian clock-related molecular network that includes alternative splicing as a regulatory mechanism . We determined the model's kinetic parameters by a two-step optimization process including random sampling and an evolutionary algorithm . With the resulting optimal parameter set we could successfully reproduce most of the pertinent experimental findings such as waveforms , phases , and half-lives of the time-dependent concentrations . Furthermore , the model can account for experimentally observed mutant behavior in LHY-ox , ztl , and toc1 mutant plants . The observed AtGRP7 mRNA oscillations can be sufficiently explained through the altered behavior of the LHY/CCA1 protein oscillations in these mutants . We note again that the slave oscillator , since it is driven by the core oscillator , does not have to share all the properties of the core oscillator such as self-sustaining oscillations , temperature compensation , or direct light input [19] , [20] . Indeed , we find dampened dynamics rather than independent self-sustained oscillations for the optimal parameter set from Table 1 ( see e . g . Figure 7 A ) . The model can also be used to predict properties not considered by our optimization procedure or properties not measured so far . It suggests a shorter half-life of AtGRP8 compared to AtGRP7 mRNA and a fast and highly saturated protein degradation of both AtGRP7 and AtGRP8 . The latter finding is consistent with recent experimental results showing that AtGRP7 and AtGRP8 proteins are among those with the highest degradation rates in Arabidopsis thaliana [71] . Furthermore , the model revealed that AtGRP7 may have a stronger impact on the alternative splicing of the AtGRP7 and AtGRP8 pre-mRNAs than AtGRP8 . This may be the mechanism underlying the observed earlier peak of AtGRP8 mRNA compared to AtGRP7 mRNA . As highlighted in [72] it might also be interesting to investigate the persistence of the above general predictions for parameters which differ from the optimal parameter set considered so far . Figure S9 indicates that the subordination of AtGRP8 to AtGRP7 seems to be a robust feature of the optimization procedure , while the other two features ( shorter AtGRP8 mRNA half-life and saturated protein degradation ) seem to be less robust . Our modeling process also provided theoretical insight into possible mechanisms underlying the experimentally observed AtGRP7 and AtGRP8 oscillations: The slave oscillator model from equations ( 1 ) – ( 6 ) is potentially able to show bistability and indeed does so for the parameter set found by our optimization scheme , suggesting that the core oscillator basically triggers periodic switching of the slave oscillator between two monostable branches by crossing a bistable regime . Our AtGRP7-AtGRP8 slave oscillator could therefore be the first in vivo manifestation of the purely theoretical proposal of a genetic toggle switch driven by an autonomous self-sustained oscillator [51] . What evolutionary benefit could such a mechanism have ? It is known that oscillations based on a hysteretic switch can show robust amplitudes . Indeed , our present slave oscillator also shows oscillations which are robust in amplitude for a considerable variety of different driving oscillations . The formation of a driven interlocked auto-regulatory feedback loop that originated from a gene duplication event in the case of AtGRP7 and AtGRP8 , can thus lead to a system showing a hysteretic behavior and resulting , if forced with an appropriate amplitude , in oscillations with a robust amplitude . Finally , we proposed two possible changes in the current view of the regulatory network of AtGRP7 and AtGRP8: First , we can still reproduce the experimental findings even without the common assumption of transcriptional repression of AtGRP8 by LHY/CCA1 . Up to now , the latter assumption has been justified by reasons of similarity with AtGRP7 but not by direct experimental measurements [26] , [27] . Second , we have discussed modifications of the model ( 1 ) – ( 6 ) in order to reproduce the experimental behavior in the lhy cca1 double mutant . In contrast to the simulation of the lhy cca1 double mutant , the AtGRP7 transcript shows oscillatory behavior with a phase shift to dawn under entrainment conditions [58]–[60] . We therefore tested natural possibilities how to cure this shortcoming of the model: Two of them , namely the autonomous oscillations due to noise effects and a change of the kinetic parameters from Table 1 could be readily excluded since they cannot explain the phase locking of the AtGRP7 mRNA in the lhy cca1 double mutant to 24h-periodic light-dark cycles . We therefore concluded that additional influences of the core on the slave oscillator , on top of the transcriptional repression by LHY/CCA1 , have to be incorporated to consistently explain both the wild-type and the lhy cca1 double mutant behavior . Furthermore , it has to be taken into account that AtGRP7 influences many physiological processes: It promotes the floral transition at least partly by down-regulating the floral repressor FLC [33] . Furthermore , it plays a role in the plants innate immune system since grp7-1 plants that do not produce AtGRP7 mRNA are more susceptible to Pseudomonas syringae [34] , [35] . AtGRP7 is also known to mediate responses to stresses such as oxidative stress , high salt , mannitol , or cold [21] , [36] , [37] . Our modeling results could be used in future work to integrate the AtGRP7 and AtGRP8 feedback loops with these other regulatory cues .
The numerical solutions of equations ( 1 ) – ( 6 ) , or equivalently of equation ( 7 ) , have been obtained by using the odeint function of SCIentificPYthon which uses LSODA from the Fortran library ODEPACK . In particular , we remark that LSODA is able to identify and solve initial value problems for both stiff and non-stiff problems . In this section we provide the details of the optimization procedure as referred to in the section Parameter Estimation . Similarly as in [7] , we started our search for an optimal fit by generating Antonov-Saleev quasi-random parameter sequences ( adopting the gsl_qrng_sobol routine from the GNU Scientific Library ) that were subsequently tested for their fitness ( for the explicit definition of , see Text S1 A ) . To take into account the similarity of AtGRP7 and AtGRP8 we first sampled the parameters for a reduced system consisting only of AtGRP7 , see also Text S1 B . After this random sampling step , the network motif was extended to the complete system while choosing the parameters in order to generate two identical oscillatory profiles for AtGRP7 and AtGRP8 ( upon comparison of equation ( 1 ) – ( 6 ) in the main text and those in Text S1 B , all parameters have to be duplicated except for the rate constant which has to be set to the half of its previous value and then has to be identified with and ) . In the next step , we took the best one hundred values and further minimized the cost function in their local neighborhood . In order to solve this dimensional minimization problem we used the gradient-free Nelder Mead Downhill Simplex method , where an initial simplex with vertices , including the starting parameter set , “crawls” amoeba-like via shape transformations ( reflection , contraction and expansion ) through parameter space in the direction of lower cost [41] . We modified the original algorithm in a way that negative and therefore biologically not meaningful parameter values were penalized by setting the cost-function value of such vertices to infinity . The starting simplex was defined by the initial parameter set and the set of vertices defined by where the 's are the unit vectors in each parameter space's direction and is a constant chosen to be in our simulations . The reflection , expansion and contraction coefficients were chosen as throughout the simulations and after the algorithm claimed to be finished it was restarted four times from the best point found in the previous run . We also tried out a Monte-Carlo Hillclimbing method instead of the simplex optimization , which however led to worse results . As detailed in the main text , the lhy cca1 double mutant can be modeled by setting in equations ( 1 ) and ( 4 ) to zero for all times . In the slave oscillator model proposed here , this is equivalent to the deletion of all links to the core oscillator . Equations ( 1 ) – ( 6 ) , or equivalently ( 7 ) , then define an autonomous dynamical system which is easy enough to calculate the fixed points analytically . More precisely , for one component of the fixed point , namely , one obtains the following closed quartic equation ( 9 ) with coefficients ( 10 ) ( 11 ) ( 12 ) ( 13 ) ( 14 ) and abbreviations ( 15 ) ( 16 ) ( 17 ) ( 18 ) for . In principle the quartic equation ( 9 ) can be solved analytically by means of the formula of Cadano & Ferrari . We used the root finding package root of SCIentificPYthon instead . In general , we thus obtained four different solutions of the quartic equation ( 9 ) . Once these four solutions are determined , the remaining components of the four fixed points , , can be readily obtained from the equations ( 19 ) ( 20 ) ( 21 ) For the optimal parameter set from Table 1 we thus obtained the following four fixed points ( 22 ) ( 23 ) ( 24 ) ( 25 ) where the last one is not biologically meaningful due to its negative concentration values . A standard linear stability analysis based on the eigenvalues of the Jacobian matrix ( 26 ) reveals that two of the remaining fixed points ( 22 ) , ( 23 ) , and ( 24 ) are ( locally ) stable ( namely and ) and one is ( locally ) unstable ( namely ) . A similar algorithm was used to generate Figure 4 and Figures S6 and S7: For each parameter set we first calculated the four fixed points as described above . In a next step , those with negative or complex components were sorted out . Finally , we performed a linear stability analysis as described above . In order to better highlight the dependence of our slave oscillator on properties like the amplitude or peak broadness of , we replaced the differential equations for the molecular core oscillator model provided by Pokhilko et al . [11] by an easily tunable generic oscillator in the form of a modified nonuniform Poincaré oscillator as proposed in [55] . Its radial evolution is given by ( 27 ) therefore converging for any initial condition to the stable fixed point , amounting to the amplitude of the resulting oscillations . The phase dynamics are given by ( 28 ) where determines the shape of the oscillations , ranging from a sinusoidal ( ) to a more and more spiky oscillator ( ) with period ( 29 ) ( 30 ) Note that the period depends on the choice of both parameters and . In [55] , the model parameter in ( 28 ) was originally chosen as a small non-zero positive constant in order to make sure that never becomes zero , since for the solution of equation ( 28 ) would evolve to its fixed point in phase . For our purpose , we set ( 31 ) so that , for any given , the oscillator exhibits oscillations with a fixed period . Finally , we define ( 32 ) as the input substituting the LHY/CCA1 oscillations in ( 1 ) and ( 4 ) . The extra parameter in ( 32 ) denotes the trough value of the oscillations and is set to the trough-value of the oscillations . | The circadian clock organizes the day in the life of a plant by causing 24h rhythms in gene expression . For example , the core clockwork of the model plant Arabidopsis thaliana causes the transcripts encoding the RNA-binding proteins AtGRP7 and AtGRP8 to undergo high amplitude oscillations with a peak at the end of the day . AtGRP7 and AtGRP8 , in turn , negatively auto-regulate and reciprocally cross-regulate their own expression by causing alternative splicing of their pre-mRNAs , followed by rapid degradation of the alternatively spliced transcripts . This has led to the suggestion that they represent molecular slave oscillators downstream of the core clock . Using a mathematical model we obtain insights into possible mechanisms underlying the experimentally observed dynamics , e . g . a higher impact of AtGRP7 protein compared to the impact of AtGRP8 protein on the alternative splicing explains the experimentally observed phases of their transcript . Previously , components that reciprocally repress their own transcription ( double negative loops ) have been shown to potentially act as a toggle switch between two states . We provide theoretical evidence that the slave oscillator could be a bistable toggle switch as well , operating at the post-transcriptional level . | [
"Abstract",
"Introduction",
"Results/Discussion",
"Methods"
] | [
"systems",
"biology",
"model",
"organisms",
"plant",
"and",
"algal",
"models",
"mathematics",
"theoretical",
"biology",
"plant",
"biology",
"regulatory",
"networks",
"biology",
"computational",
"biology",
"signaling",
"networks",
"nonlinear",
"dynamics",
"genetics",
"and",
"genomics",
"arabidopsis",
"thaliana"
] | 2013 | A Circadian Clock-Regulated Toggle Switch Explains AtGRP7 and AtGRP8 Oscillations in Arabidopsis thaliana |
Enhanced root hair production , which increases the root surface area for nutrient uptake , is a typical adaptive response of plants to phosphate ( Pi ) starvation . Although previous studies have shown that ethylene plays an important role in root hair development induced by Pi starvation , the underlying molecular mechanism is not understood . In this work , we characterized an Arabidopsis mutant , hps5 , that displays constitutive ethylene responses and increased sensitivity to Pi starvation due to a mutation in the ethylene receptor ERS1 . hps5 accumulates high levels of EIN3 protein , a key transcription factor involved in the ethylene signaling pathway , under both Pi sufficiency and deficiency . Pi starvation also increases the accumulation of EIN3 protein . Combined molecular , genetic , and genomic analyses identified a group of genes that affect root hair development by regulating cell wall modifications . The expression of these genes is induced by Pi starvation and is enhanced in the EIN3-overexpressing line . In contrast , the induction of these genes by Pi starvation is suppressed in ein3 and ein3eil1 mutants . EIN3 protein can directly bind to the promoter of these genes , some of which are also the immediate targets of RSL4 , a key transcription factor that regulates root hair development . Based on these results , we propose that under normal growth conditions , the level of ethylene is low in root cells; a group of key transcription factors , including RSL4 and its homologs , trigger the transcription of their target genes to promote root hair development; Pi starvation increases the levels of the protein EIN3 , which directly binds to the promoters of the genes targeted by RSL4 and its homologs and further increase their transcription , resulting in the enhanced production of root hairs . This model not only explains how ethylene mediates root hair responses to Pi starvation , but may provide a general mechanism for how ethylene regulates root hair development under both stress and non-stress conditions .
As an essential macronutrient , phosphorus ( P ) plays vital roles in plant growth , development , and metabolism . P not only serves as structural elements of nucleic acids and phospholipids , but is also involved in many important biological processes , including photosynthesis , oxidative phosphorylation , regulation of enzymatic activities , and cell signaling . Using the Pi transporters localized on the root surface , plants take up P from the soil in the form of inorganic phosphate ( Pi ) [1] . When plants are exposed to Pi deficiency , they activate an array of adaptive responses to cope with this nutritional stress . These responses involve developmental , biochemical , and physiological changes , including the reprogramming of root development; increased activities of high affinity Pi transporters; the induction and secretion of acid phosphatases , RNases , and organic acids; and the accumulation of anthocyanin and starch [2 , 3] . The Pi starvation-induced changes in root development include the inhibition of primary root growth and the increased production of lateral roots and root hairs [4] . Root hairs , which are tubular outgrowths of root epidermal cells , account for a large portion of the root surface area involved in water and nutrient uptake [5] . For rye plants grown under Pi starvation , root hairs are responsible for nearly 60% of the Pi absorbed [6] . In Pi deficient-Arabidopsis plants , root hairs represent 91% of the total root surface area [7] . Under low Pi conditions , wild-type ( WT ) Arabidopsis plants acquire more Pi than mutants that are defective in root hair formation [8] . Root hairs also modify the rhizosphere by exuding large amounts of organic acids , enzymes , mucilage , and secondary metabolites [9] . The enhanced growth of root hairs has been thought to be the earliest root morphological response to Pi starvation [8] . Low Pi availability increases root hair length by increasing root hair growth rate and growth duration [10] . The increase in root hair density in Pi deficient-plants is due to the increase in trichoblast file number , the reduction in trichoblast length , and/or the increase in the percentage of trichoblast cells that form root hairs [11 , 12] . Pi starvation alters ethylene biosynthesis and sensitivity in plants ( for a review , see [13] ) . Ethylene is a gaseous phytohormone that has many functions in plant growth and development and in plant responses to stress conditions . In Arabidopsis , ethylene is perceived by five receptors located in endoplasmic reticulum ( ER ) membranes [14] . In the absence of ethylene , ethylene receptors activate CTR1 ( CONSTITUTIVE TRIPL1 RESPONSE1 , a Raf1-like kinase ) , which subsequently suppresses the function of its downstream target EIN2 ( ETHYLENE INSENSTIVE2 ) . When ethylene binds to its receptors , it inactivates CTR1 , which causes the translocation of a C-terminal fragment of EIN2 ( EIN2-C’ ) into the nucleus . In the nucleus , EIN2-C’ increases the protein levels of EIN3 ( ETHYLENE INSENSITIVE3 ) and its closest homolog EIL1 ( ETHYLENE INSENSITIVE3-LIKE1 ) , two key transcription factors of the ethylene signaling pathway . EIN3 and EIL1 then turn on the transcription of a battery of downstream target genes , which initiate a diverse array of plant responses to internal and external cues . Ethylene has long been known to be a positive regulator of root hair development [15 , 16] . It acts downstream of the key transcription factors GL2 and RHD6 to promote root hair formation [17] . Previous reports have also indicated that ethylene plays an important role in Pi starvation-induced root hair development [7 , 12] . Application of inhibitors of ethylene biosynthesis or action reduces Pi starvation-induced root hair development in Arabidopsis . Similarly , the root hair responses to Pi starvation are significantly attenuated in ethylene-insensitive mutants of Arabidopsis [12] . To date , however , the molecular mechanism of how ethylene mediates Pi starvation-induced root hair development remains unknown . In this report , we characterize an Arabidopsis mutant , hps5 , that carries a mutation in the ethylene receptor ERS1 . hps5 accumulates higher levels of EIN3 protein than the WT under both Pi sufficiency and deficiency . Pi starvation also increases the accumulation of EIN3 . By combining genetic , molecular , and transcriptomic approaches , we first identify a group of ethylene-regulated genes involved in Pi starvation-induced root hair development . We show that EIN3 directly binds to the promoters of genes that are the targets of a key transcription factor that regulates root hair development . Finally , we propose a working model of how ethylene mediates root hair responses to Pi starvation . This model may also provide a general mechanism for how ethylene regulates root hair development under both stress and non-stress conditions .
To identify the molecular components involved in Pi starvation-induced root hair production , we performed a large-scale screen for the Arabidopsis mutants with enhanced root hair production under Pi deficiency . One such mutant , hps5 ( hypersensitive to Pi starvation5 ) , was obtained through this screen . Both root hair length and root hair density were increased in hps5 compared to the WT under Pi deficiency ( P- ) ( Fig 1A and 1B ) . Root hair production was also greater in hps5 than in the WT under Pi sufficiency ( P+ ) . In addition to producing more root hairs , hps5 also exhibited altered primary and lateral root growth . Pi starvation inhibited primary root growth and promoted lateral root formation in the WT ( Fig 1 ) . At 7 days after germination ( DAG ) , the primary root was about 30% shorter for hps5 than for the WT on P+ medium but was about 50% shorter on P- medium ( S1A Fig ) . At 9 DAG , lateral root density was less for hps5 than for the WT on P+ medium but was similar on P- medium ( S1B Fig ) . Next , we examined other Pi responses of hps5 . Induction and secretion of acid phosphatases ( APases ) is a hallmark response of plants to Pi deficiency [18] . The Pi starvation-induced APases function in Pi remobilization and recycling in old tissues , and in the utilization of organophosphate compounds in the rhizosphere . The APase activity on the root surface can be detected by histochemical staining using the APase substrate BCIP ( 5-bromo-4-chloro-3-indolyl-phosphate ) [19] . Cleavage of BCIP by APases produces a blue precipitate . Using this method , we found that hps5 had a stronger blue staining on its root surface than the WT under Pi deficiency ( S2A Fig ) . Under Pi sufficiency ( P+ ) , hps5 roots also exhibited a light blue staining while WT roots were white ( S2A Fig ) . The enhanced root surface-associated APase activity in hps5 was confirmed by quantitative analysis ( S2B Fig ) . The expression of eight Pi starvation-induced marker genes was analyzed . These genes encode two non-coding transcripts , At4 and IPS1; an RNase , RNS1; a microRNA , miRNA399D; two APases , ACP5 ( AtPAP17 ) and AtPAP10; and two high-affinity Pi transporters , AtPT1 ( Pht1;1 ) and AtPT2 ( Pht1;4 ) [20] . At 7 DAG , the expression of all of these marker genes , except AtPT1 and AtPT2 , was significantly higher in hps5 seedlings than in the WT under Pi deficiency ( S3 Fig ) . Accumulation of anthocyanin in leaves is another characteristic response of plants to Pi starvation . Induction of anthocyanin is thought to protect the chloroplast membranes under Pi deficiency [3] . At 10 DAG on P- medium and P- medium with 10–100 μM Pi , the leaves of the seedlings had turned purple but the intensity of the purple was greater in the WT than in hps5 ( S4 Fig ) . These results indicated that hps5 was less sensitive to Pi starvation-induced anthocyanin accumulation in leaves . The total P and cellular Pi contents in shoots and roots of 9-day-old seedlings did not significantly differ between hps5 and the WT under either P+ or P- conditions ( S5 Fig ) . When hps5 was backcrossed to the WT , all of the F1 plants showed WT phenotypes . The F2 progeny derived from the selfed F1 plants segregated into WT and mutant phenotypes in a ratio of about 3:1 ( 203:62 ) , indicating that the mutant phenotypes were caused by a single recessive mutation . Using a map-based cloning approach ( S6 Fig ) , we found a C to T mutation in the second exon of the ERS1 gene ( At2g40940 ) in hps5 ( Fig 2A ) . This mutation converted a proline to a leucine at the position of 110 of the ERS1 protein . Depending on the program used for prediction , the mutated proline residue was predicted to be located at the third or fourth amino acid position after the transmembrane domain III [21 , 22] . This amino acid residue is highly conserved in the ERS1 orthologs in all major crops ( S7 Fig ) , suggesting that it is important for the receptor function of ERS1 . To confirm that the mutated ERS1 gene was responsible for the hps5 phenotypes , we introduced the genomic sequence of WT ERS1 back into hps5 plants under the control of the cauliflower mosaic virus ( CaMV ) 35S promoter or ERS1’s own promoter . Both gene constructs ( CaMV 35S::ERS1 and ERS1::ERS1 ) could fully complement the hps5 phenotypes on P+ medium , in terms of reduced primary root growth , increased root hair production , and enhanced root-associated APase activity ( Fig 2B and S8 Fig ) . On P- medium , however , the two gene constructs could only partially rescue these mutant phenotypes ( Fig 2B and 2C , S8 Fig ) . Partial complementation under P- condition may result from interference between the mutated ERS1 protein and the other ethylene receptors that is not fully reversed by the WT ERS1 transgenic protein . When treated with ethylene in the dark , Arabidopsis seedlings display a characteristic “triple response” , i . e . , an inhibition of root and hypocotyl growth , a radial swelling of the root and hypocotyl , and an increase in the curvature of the apical hook [23] . When grown in soil , WT Arabidopsis plants treated with ethylene or the mutant with constitutive ethylene response are small in stature . We compared these growth characteristics among the WT and the mutants hps5 , ein2 , and ctr1 . ein2 is completely insensitive to ethylene [24] , and ctr1 displays constitutive ethylene responses [25] . When grown in the dark in the absence of ACC ( 1-aminocyclopropane-1-carboxylic acid ) , a precursor of ethylene biosynthesis , the WT and ein2 had a long hypocotyl and a mild apical hook whereas ctr1 had a much shorter hypocotyl and an exaggerated apical hook ( Fig 3A and 3B ) . For hps5 , the length of the hypocotyl and the curvature of the apical hook were intermediate relative to the WT and ctr1 . When treated with 1 μM ACC , the WT , hps5 , and ctr1 produced shorter hypocotyls and apical hooks with increased curvature , while ein2 did not respond to ACC . Under normal growth conditions , hps5 also showed higher expression of the ethylene-responsive genes ERF1 , EBF2 , and CHIB [26] than the WT ( Fig 3C ) . Together , these results indicate that hps5 seedlings display a moderate constitutive ethylene response . The size and morphology of adult plants grown in soil did not significantly differ between hps5 and the WT ( S9 Fig ) . Arabidopsis has five ethylene receptors: ETR1 , ERS1 , ETR2 , EIN4 , and ERS2 [14] . Because of functional redundancy , the loss-of-function mutation in a single receptor gene does not cause any mutant phenotypes . In contrast , a point mutation that abolishes the binding of ethylene to any of these five receptors makes plants insensitive to ethylene [27 , 28] . Because the point mutation in ERS1 caused constitutive ethylene responses , we suspected that this mutation had a “dominant-negative” effect , an effect that interferes with the functions of other ethylene receptors . To test this hypothesis , we isolated the mutated ERS1 gene ( mERS1 ) from hps5 and introduced it into WT plants under the control of the CaMV 35S promoter . When grown in soil , most of 25 primary transformants were smaller than the WT . Overexpression of the mERS1 gene was confirmed by qPCR in six selected transgenic lines ( S10 Fig ) . We then generated homozygote plants for the transgene ( 35S::mERS1 ) for these lines . When grown on P+ and P- media , the seedlings of these lines showed strong inhibition of primary root growth ( Fig 4A and S11 Fig ) , increased root hair production ( Fig 4B ) , and enhanced root-associated APase activity ( S12A and S12B Fig ) , mimicking the phenotypes of hps5 . When grown in soil , the size of these six transgenic lines varied , but they were all smaller than the WT and hps5 ( Fig 5 ) . These results suggested that the mutated ERS1 protein might interfere with the functions of WT ERS1 and other ethylene receptors in the transgenic plants . To investigate how ethylene signaling is enhanced in hps5 , we compared the mRNA and protein levels of EIN3 in roots of the WT and hps5 . qPCR analysis indicated that the mRNA level of EIN3 in the WT and hps5 was similar under both P+ and P- conditions ( S13A Fig ) ; as revealed by Western blot , however , the level of EIN3 protein was higher in hps5 than in the WT ( Fig 6A ) . We then examined the effects of Pi starvation on the transcription and protein accumulation of EIN3 . Like hps5 mutation , Pi starvation did not alter the level of EIN3 mRNA ( S13B Fig ) but increased the accumulation of EIN3 protein ( Fig 6B ) . Ethylene or ACC treatment is known to induce the accumulation of EIN3 protein by blocking its degradation as mediated by the F-box proteins EBF1 and EBF2 [26] . By Western blot analysis , we found that the ACC-induced accumulation of EIN3 protein in roots was further enhanced by Pi starvation ( Fig 6B ) . This result was corroborated by the confocal analysis of the expression of EIN3-GFP fusion proteins in the transgenic line expressing the EIN3-GFP fusion gene under the control of the 35S promoter ( Fig 6C ) . Because Pi starvation increases the accumulation of EIN3 protein , we next asked whether EIN3 and its closest homolog , EIL1 , are directly involved in the Pi starvation-induced root hair development . To answer this question , we analyzed the root hair phenotypes of ein3 and ein3eil1 mutants and of an EIN3-overexpressing line ( EIN3 OX ) . On P+ medium , the root hair density and root hair length of ein3 and ein3eil1 did not significantly differ from those of the WT ( Fig 7 ) . On P- medium , root hair density was lower for both ein3 and ein3eil1 than for the WT but root hair length was reduced only for ein3eil1 . In contrast , both root hair length and root hair density were greatly increased in EIN3 OX relative to the WT . These results indicate that EIN3 and EIL1 are involved in Pi starvation-induced root hair production . To identify the downstream targets of the ethylene signaling involved in Pi starvation-induced root hair development , we compared the transcriptomes of the WT , hps5 , ein3 , and ein3eil1 . The cDNA libraries were constructed using the total RNAs isolated from 7-day-old seedlings grown on P+ or P- medium . For each combination of genotype and Pi treatment , two biological replicates were used . Log2≥1 or ≤-1 ( 2-fold change in expression levels ) and FDR ( false discovery rate ) ≤ 0 . 01 were used as the cut off for selection of the genes whose expression levels significantly differed from that of P+ WT . The transcriptomic data indicated that the expression of Pi-responsive marker genes , such as Pht1;1 , Pht1;4 , ACP5 , At4 , and SPX1 , was upregulated in P- WT compared to P+ WT ( S1 Table ) , indicating that the experimental conditions for RNA-seq analyses were proper . We first compared the expression profiles of hps5 and the WT grown on P+ medium . Surprisingly , based on our selection criteria , there were only 61 genes whose expression was upregulated and 43 genes whose expression was downregulated in hps5 relative to the WT ( S2 Table ) . This indicated that the effect of hps5 mutation on gene expression at the genomic level was relatively mild , which was consistent with the absence of developmental abnormality for adult hps5 plants grown in soil . Among the 61 upregulated genes , there were seven RHS ( Root Hair-Specific ) genes [29] , including RHS2 ( a MATE [multidrug and toxin efflux] family protein ) , RHS12 ( a pectinesterase ) , RHS13 ( an extensin-like protein ) , RHS14 ( a pectate lyase ) , RHS15 ( a glycerol-3-phosphate permease , G3PP2 ) , RHS18 ( a peroxidase ) , and RHS19 ( a peroxidase ) . The induction level of RHS16 ( a leucine-rich repeat protein kinase ) was also close to 2-fold ( Log2 = 0 . 97 ) . RHS2 , RHS16 , and RHS18 have previously been shown to be involved in root hair elongation or branching [29] . In addition , the upregulated genes included two xyloglucan endotransglucosylases ( XTH12 and XTH14 ) , two other peroxidases ( PER13a and At3g49960 ) , and one pectin methyltransferase inhibitor . XTH12 and XTH14 are the two most closely related members of the XTH family in Arabidopsis and are mainly expressed in roots [30] . The enzymatic activity of the recombinant XTH14 protein has been demonstrated in vitro and this protein can affect cell elongation and root hair morphology [31] . In addition , there were 16 cell wall proteins , including 10 proline-rich extensins ( EXT2 , EXT7 , EXT10 , EXT11 , EXT12 , EXT13 , EXT14 , EXT15 , EXT16 , and EXT18 ) , two proline-rich proteins ( PRP1 and PRP3 ) , two leucine-rich repeat extensins ( LRX1 and LRX2 ) , one FASCICLIN-like arabinogalactan protein ( FLA6 ) , and one arabinogalactan protein ( AGP3 ) . Among them , the knockout mutants for EXT7 , EXT11 , EXT14 , EXT16 , and EXT18 [32] , and LRX1 and LRX2 [33] have abnormal root hair initiation , elongation , and morphologies . The expression of PRP3 and AGP3 is regulated by root hair-specific developmental pathways [34 , 35] . Besides , COW1 ( CAN OF WORMS1 ) , AHA7 ( Arabidopsis H+-ATPase 7 ) , and ARGOS-like ( ARL ) genes were upregulated in hps5 . COW1 encodes a Sec14p-like phosphatidylinositol transfer protein [36] . The loss-of-function mutants of COW1 and its homolog in rice , OsSNDP1 , produced short root hairs with irregular shapes [36 , 37] . The aha7 T-DNA insertional mutant has reduced root hair density under both normal and Fe-deficient conditions [38] . Because the initiation and elongation of root hairs are associated with a localized decrease in apoplastic pH [39] , AHA7 may regulate root hair density by modifying local pH levels . ARL functions in cell expansion-dependent organ growth [40] . Reduced expression or overexpression of ARL results in decreased or increased cell size in Arabidopsis cotyledons , leaves , and other lateral organs . Notably , five of the genes that were upregulated in hps5 are involved in jasmonate ( JA ) biosynthesis or signalling . These genes encode phospholipase A 2A ( PLA2A ) [41] , sulfotransferase 2A ( ST2A ) [42] , lipoxygenase 2 ( LOX2 ) [43] , salt tolerance zinc finger ( ZAT10 ) [44] , and octadecanoid-responsive AP2/ERF59 ( ORA59 ) [45] . This suggests that enhanced ethylene signaling might upregulate JA biosynthesis and signalling . None of the 43 genes that were downregulated in P+ hps5 relative to P+ WT was significantly enriched in any functional category , except that five of the genes encoding bifunctional inhibitor/lipid-transfer protein/seed storage 2S albumin superfamily proteins ( S2 Table ) . The biological relevance of the downregulation of these genes is unknown . We then analysed the transcriptomes of the Pi deficient-WT ( P- WT ) seedlings . Under our experimental conditions , 1 , 594 genes were upregulated ( Fig 8A ) and 673 genes were downregulated in the WT in response to Pi deficiency ( S1 Table ) . Among the 61 upregulated genes in P+ hps5 ( S2 Table ) , 43 were also upregulated by Pi starvation in the WT , and these included the seven RHSs and the genes encoding most cell wall proteins and cell wall modifying enzymes ( Fig 8A , Table 1 ) . The accuracy of the RNA-seq experiments was verified by qPCR analysis of six selected genes ( S14 Fig ) . We then examined the expression of these 43 genes in P- ein3 and P- ein3eil1 ( S3 and S4 Tables ) . The expression levels of 37 genes were lower in P- ein3 than in P- WT ( Fig 8B , Table 1 ) . And , the expression of most of these 37 genes was further downregulated in P- ein3eil1 ( Fig 8B , Table 1 ) . These RNA-seq results were also validated by qPCR analysis of six selected genes ( S15 Fig ) . Furthermore , we randomly selected eight genes for qPCR analysis and found that their expression in the EIN3 OX line was significantly elevated under normal growth conditions ( Fig 8C ) . Together , the results suggest that the transcription of these 37 starvation-induced genes is regulated by ethylene signaling via the modulation of levels of EIN3/EIL1 proteins . To confirm the function of some upregulated genes in hps5 in root hair development , we examined the phenotypes of the T-DNA insertion lines for 14 genes; we found that the T-DNA lines for six genes ( EXT13 , At2g16586 , UCC3 , ORA59 , RHS15 , and RHS19 ) showed some defects in root hair development . Because RHS15 and RHS19 are specifically expressed in root hairs [29] , we conducted a detailed analysis of the root hair phenotypes for these two genes using their T-DNA insertion lines and overexpressing lines , which were generated using a root hair-specific EXP7 promoter [47] . Three homozygote SALK lines containing T-DNA insertions in the promoter of RHS15 ( SALK_015573 , SALK_015669 , and SALK_117952 ) , and three homozygote SALK lines containing T-DNA insertions in the third exon ( SALK_093852 ) , the promoter ( SALK_010873 ) , and the 3’UTR ( SALK_020724 ) of RHS19 , respectively , were identified ( S16A Fig ) by genotyping using the primers listed in S5 Table . qPCR analyses showed that the three SALK lines of RHS15 were knockdown lines and that the three RHS19 T-DNA lines were nearly null mutants ( S16B and S16C Fig ) . Under Pi sufficiency , root hair density of all three T-DNA lines for RHS15 and RHS19 did not significantly differ from that of the WT; however , root hairs were shorter for two rhs15 lines and two rhs19 lines than for the WT ( Fig 9 ) . Under Pi deficiency , both root hair density and length of all three T-DNA lines for RHS15 and RHS19 were reduced compared to the WT . In the overexpressing lines for these two genes ( S16B and S16C Fig ) , in contrast , root hair density and length tended to be greater than for the WT under both P+ and P- conditions ( Fig 9 ) . EIN3/EIL1 bind to DNA with a consensus sequence of A ( C/T ) G ( A/T ) A ( C/T ) CT [48] . We then analyzed the promoter regions of 12 upregulated genes in P+ hps5 and found that all of them contained 1 to 8 putative EIN3-binding sites ( S17 Fig ) . To determine whether EIN3 can directly bind to these promoters , we performed chromatin immunoprecipitation ( ChIP ) assays using the transgenic line expressing an 35S::EIN3-GFP fusion gene [26] . The chromatins of 10-day-old seedlings of the WT and 35S::EIN3-GFP line grown under P+ or P- conditions were isolated , cross-linked , and precipitated with anti-GFP antibodies . The DNA fragments that precipitated with EIN3-GFP proteins were analyzed by qPCR . The results showed that PCR-amplified EIN3-binding sites in 10 of the 12 selected genes were enriched in the immuno-precipitated chromatins from the EIN3-GFP plants but not from the WT plants ( Fig 10 ) . These results indicated that EIN3 could directly bind to one or more EIN3-binding sites on the promoters of these 10 genes , including RHS15 and RHS19 , in planta . The binding affinity to EIN3 , however , varied among sites . The binding of EIN3 to most of these promoters was not further enhanced by Pi starvation , probably because of the overexpression of EIN3 as directed by the strong 35S promoter . To provide more evidence that EIN3 can directly bind to the promoter of these genes , we carried out electrophoretic mobility shift assays ( EMSA ) using recombinant EIN3 proteins produced in E . coli cells . The recombinant proteins contain the DNA-binding domain of EIN3 ( amino acids 141–352 ) . This truncated EIN3 protein was fused with a GST protein to facilitate protein purification . The DNA probes used in EMSA were tandemly repeated tetramers of the putative EIN3-binding elements with 5-bp flanking sequences on each side ( The sequences of the DNA probes are listed in S6 Table ) . The EIN3-binding elements tested in the EMSA assays were selected from promoters of four genes ( PRP1 , EXT14 , RHS15 , and RHS19 ) that showed high binding affinity to EIN3 in the ChIP assays . The EMSA results indicated that the recombinant GST-EIN3 proteins , but not GST protein alone , can bind to the biotin-labelled probes . The binding of EIN3-GST proteins to the biotin-labelled probes could be competed out by unlabelled probes , indicating that the binding is sequence-specific ( Fig 11 ) . RSL4 ( ROOT HAIR DEFECTIVE 6 [RHD6]-LIKE4 ) is a basic helix-loop-helix transcription factor that is important for root hair development . The loss-of-function mutant of RSL4 had reduced root hair density and root hair length [49] . In contrast , the RSL4 OX lines had increased root hair density and root hair length . Yi et al . [49] further showed that the transcription and protein accumulation of RSL4 are enhanced by Pi starvation and that the induction of root hair growth by Pi starvation is impaired in rsl4 . Eighty-three putative direct targets of RSL4 were identified by comparative RNA-seq analyses using the WT , rsl4 , and RSL4 OX lines [49] . We compared these 83 genes and 61 upregulated genes in P+ hps5 and found that 19 genes were in common ( Table 1 , the genes indicated by asterisks ) . Among them , the functions of seven genes in root hair development have been experimentally demonstrated . These results suggested that EIN3 and RSL4 might bind to the same promoters of a subset of downstream target genes involved in root hair development . To determine whether RSL4 itself is regulated by ethylene signaling , we examined RSL4 expression in 7-day-old seedlings grown on P+ and P- media in the absence or presence of ACC . The results indicated that the expression of RSL4 was induced by Pi starvation , which was consistent with [49] , but was not induced by ACC ( S18 Fig ) .
As an important adaptive response to Pi deficiency , plants enhance their root hair production , including the increase in both root hair density and root hair length . This response is thought to increase the surface area of roots for Pi uptake and root exudation . Root hair development induced by Pi starvation is mediated by the phytohormone ethylene [12] . Although the ethylene signaling pathway has been well elucidated [14] , the molecular mechanism of how ethylene mediates Pi starvation-induced root hair development remains unknown . In this work , we identified an Arabidopsis mutant , hps5 , that displays enhanced root hair production under both Pi sufficiency and deficiency ( Fig 1 ) . hps5 also has other altered responses to Pi deficiency , and these include increased root-associated APase activity , increased Pi starvation-induced gene expression , and reduced anthocyanin accumulation ( S2–S4 Figs ) . These altered Pi responses are caused by a mutation in the ethylene receptor ERS1 ( Fig 2 ) . By analyzing three Arabidopsis hps ( hypersensitive to Pi starvation ) mutants ( hps2 , hps3 , and hps4 ) , we previously showed that enhanced ethylene biosynthesis or ethylene signaling makes plant hypersensitive to Pi starvation [50 , 51 , 20] . The Pi responses of hps5 are similar to those of these three hps mutants and are consistent with an enhancement of ethylene signaling . This inference is further supported by the enhanced apical hook and increased expression of ethylene-responsive genes in hps5 seedlings , and by the retarded growth of plants overexpressing the mutated ERS1 gene ( Figs 3–5 ) . On the other hand , Pi deficiency has been shown to increase plant sensitivity to ethylene ( for a review , see [13] ) . For example , the effects of ethylene on the induction of the Pi transporter gene Pht1;4 [50] and root hair development [12] were much stronger under Pi deficiency than under Pi sufficiency , indicating a cross-talk between ethylene and phosphorus availability . Enhanced ethylene biosynthesis or signaling is known to induce the accumulation of EIN3 protein [26] . The results of the current study show that ethylene induces greater EIN3 accumulation when plants are exposed to Pi starvation ( Fig 6 ) , suggesting that Pi starvation increases the stability of EIN3 . Therefore , EIN3 may act as the convergent point for the cross-talk between ethylene and low-Pi signal . That Pi deficiency increases the accumulation of EIN3 protein may also be the molecular basis for how Pi deficiency enhances plant sensitivity to ethylene . Phenotypic and transcriptomic analyses of hps5 provided us with a good opportunity to identify the downstream targets of the ethylene signaling involved in Pi starvation-induced root hair development ( Fig 8 ) . Ethylene has multiple effects on plant growth and development , including the inhibition of root growth and the promotion of root hair production , leaf and flower senescence , and fruit ripening [52] . Because the morphological changes in hps5 were evident only for roots at the seedling stage , the root seems to be the most sensitive organ in response to the ethylene signal . Thus , a transcriptomic analysis of hps5 may help identify the target genes of ethylene signaling involved in root development , including root hair formation . Root hair formation involves three steps , i . e . , cell fate specification , hair initiation , and hair elongation . An elegant regulatory pathway regulating hair cell specification has been established in Arabidopsis [35 , 53 , 54] . GL2 is a negative transcriptional regulator of root hair formation . In the hair cells , an upstream complex consisting of TTG1 , GL3 , EGL3 , CPC , and other proteins suppresses the expression of GL2 and thus releases the expression of its target genes , including a basic helix-loop-helix transcription factor RHD6 and its several homolog RSLs ( RHD6-LIKEs ) [53] . RSL2 and RSL4 are also the immediate targets of RHD6 [49] . These transcription factors then turn on the expression of a battery of downstream genes that directly participate in the cellular processes involved in root hair initiation and elongation ( also see the diagram in Fig 12 ) . Root hair initiation and subsequent elongation involves polarized cell growth or tip growth that depends on a variety of highly coordinated cellular activities [55 , 56] . These activities involve the action of cell wall proteins , peroxidases , xyloglucan endotransglucosylases , H+-ATPases , and pectin modification enzymes . In Arabidopsis , these cell wall proteins and cell wall modification enzymes are encoded by multigene families . To provide a precise mechanism for the morphogenetic control of root hair development , plants must selectively activate a subset of these genes in root hair cells . However , which specific isoforms of the proteins encoded by these genes are involved in root hair development remain largely unknown . Among the 61 genes that are upregulated in hps5 relative to the WT under Pi-sufficient conditions , 13 genes had been demonstrated to be involved in root hair development before this work . The co-upregulation of other genes in hps5 , which have similar biochemical functions or root hair-specific expression patterns as those 13 genes , suggested that they might also participate in root hair development . These other genes include five RHS genes ( RHS12 , a pectinesterase; RHS13 , an extensin-like protein; RHS14 , a pectate lyase; RHS15 , a glycerol-3-phosphate permease; and RHS19 , a peroxidase ) , a xyloglucan endotransglucosylase ( XTH12 ) , two other peroxidases ( PER13a and At3g49960 ) , nine cell wall proteins ( PRP1 , PRP3 , FLA6 , AGP3 , EXT2 , EXT10 , EXT12 , EXT13 , and EXT15 ) , and one pectin methytransferase inhibitor . Another interesting group of upregulated genes in hps5 encode five proteins involved in JA biosynthesis and signaling . JA biosynthesis is essential for ethylene-induced root hair formation [57] . Thus , ethylene might regulate root hair development also by altering JA biosynthesis and signaling through the induction of these five genes . The similar co-expression patterns of these genes have also been reported in two other independent studies [32 , 35] . When we examined the root hair phenotypes of the T-DNA insertion lines for 14 putative root hair-related genes and overexpressing lines of RHS15 and RHS19 ( Fig 9 ) , we found that the T-DNA lines for six genes had defects in root hair development . This further supported our inference that these co-expressed genes might also be involved in root hair growth . The lack of root hair phenotypes of the T-DNA lines for the other eight genes might be due to the functional redundancy of genes belonging to the same family or to the involvement of those genes in other aspects of root development , such as primary and lateral root development . In summary , the analyses of the transcriptomic data of hps5 allowed us to identify the genes encoding the specific isoforms of the cell wall proteins and cell wall modification enzymes involved in the root hair development , including those that have been demonstrated before and new genes discovered in this research . Because both hps5 and Pi-deficient plants accumulate high levels of EIN3 protein ( Fig 6 ) , we wondered whether EIN3 and its closest homolog EIL1 are directly involved in Pi starvation-induced root hair development . Our RNA-seq analyses of ein3 and ein3eil1 mutants indicate that , among the 61 upregulated genes in hps5 relative to the WT , 43 are also induced by Pi starvation . The induction of 37 of these 43 genes by Pi starvation is partially or completely suppressed in ein3 and ein3eil1 mutants ( Fig 8 ) . In contrast , their expression is greatly elevated in the EIN3 OX lines ( based on the analysis of eight randomly selected genes ) . These promoter of these eight genes all contain putative EIN3-binding elements ( S17 Fig ) . Moreover , our ChIP and EMSA assays further demonstrated that EIN3 can bind to the promoter of these genes ( Figs 10 and 11 ) . Together , these results provide compelling evidence that these genes are the direct targets of EIN3 . RSL4 is a key positive regulator of root hair development [49] and acts downstream of GL2 and RHD6 . Eighty-three putative direct targets of RSL4 have been identified , including those that function in cell wall modifications . Interestingly , we found that among these 83 putative targets , 19 might also be the direct targets of EIN3 ( Table 1 ) . These results indicate that EIN3 and RSL4 bind to the promoters of a subset of the same target genes involved in cell wall modifications . RSL2 , a close relative of RSL4 , also acts downstream of RHD6 in regulating root hair development [49] . We speculate that EIN3 may also share some targets with RSL2 . In Fig 12 , we describe a working model for how ethylene mediates Pi starvation-induced root hair development . We propose that the level of ethylene in root cells is low under normal growth conditions . A group of key transcription factors , including RSL4 and its close relatives ( RSL1 , RSL2 , LRL1 , and LRL2 ) [53] , trigger basal level transcription of their target genes to promote root hair initiation and elongation by regulating a variety of cellular activities , such as cell wall modifications ( Fig 12 ) . Pi starvation increases the levels of EIN3 proteins in root cells through both enhanced ethylene biosynthesis and protein stability of EIN3 . These EIN3 proteins can bypass RHD6 and directly bind to the promoters of the genes targeted by RSL4 , RSL2 , and probably other transcription factors . At same time , the level of RSL4 proteins is also enhanced by Pi starvation . Thus , elevated levels of EIN3 and RSL4 increase the transcription of these genes , resulting in enhanced root hair production . Ethylene has also been reported to mediate root hair growth induced by other stresses , such as iron and boron deficiencies [58 , 59] , bacterial infection [60] , mechanical impedance [61] , local water stress [62] , and elevated levels of carbon monoxide [63] . Therefore , our working model may provide a general mechanism of how ethylene mediates root hair responses to other stress signals as well . Our working model is also consistent with previous observations about the effect of ethylene on root hair development under normal growth conditions . Under such conditions , the blocking of ethylene signaling by either an inhibitor of ethylene perception or by mutations in some key components of the ethylene signaling pathway has only a minor effect on root hair initiation and elongation ( [12 , 17 , 62] and this study ) . Root hair density and root hair length can be dramatically increased , however , by exogenously applied ACC , by mutations resulting in constitutive ethylene responses ( the ctr1 mutant ) or ethylene overproduction ( the eto1 mutant ) , or by overexpression of EIN3 [12 , 15 , 16 , 62] . In these plants , the increase of root hair density can be partly attributed to the ectopically formed root hairs ( the root hairs formed in the position of non-hair cells ) . In addition , ethylene has been shown to act downstream or in separate pathway of RHD6 [17] . According to our working model , this is because the contribution of EIN3 to the transcription of the genes involved in cellular processes related to root hair development is relatively low under normal growth conditions . In ACC-treated plants , in mutants with constitutive ethylene response or ethylene overproduction , and in the EIN3-overexpressing line , however , the levels of EIN3 proteins in both the hair cells and non-hair cells are dramatically increased . These EIN3 proteins bypass RHD6 to directly bind to the promoters of the RSL4- and RSL2-targeted genes , resulting in a great increase in root hair density and root hair length , including the formation of root hairs in non-hair cells . Finally , we note that hps5 will also be very useful for the study of ethylene receptor function . In Arabidopsis , there are five ethylene receptors that negatively regulate ethylene signaling . ETR1 and ERS1 belong to the same subgroup in this family . Because of the functional redundancy among these five receptors , a loss-of-function mutation in one receptor does not cause any phenotype [27] . A gain-of-function point mutation , in contrast , can make plants insensitive to ethylene because of a dominant-negative effect [22 , 28] . To date , no mutants with a single point mutation in any ethylene receptors have been reported to have constitutive ethylene responses . Previously , all reported point mutations that cause plants to be insensitive to ethylene occur in the receptor’s transmembrane domains . The mutated proline residue in hps5 , however , is located immediately after the transmembrane domain III ( Fig 2A ) . Therefore , we believe that the sequence of the cytoplasmic region immediately behind the transmembrane domain III is critical for the function of ERS1 . The phenotypic analyses of hps5 and the transgenic plants expressing the mutated ERS gene suggest that this region is involved in the protein-protein interactions between ERS1 and other receptors or in self interactions . Further study of the structural–functional relationship of this conserved proline residue in ERS1 may provide new insights into the function of ethylene receptors .
All plants used in this study were of the Columbia ecotype background . The SALK T-DNA insertional lines were obtained from the Arabidopsis Biological Resource Center ( ABRC ) . The primers used to confirm the insertion of the T-DNA lines are listed in S5 Table . The Pi-sufficient ( P+ ) medium was half-strength MS medium with 1% ( w/v ) sucrose and 1 . 2% ( w/v ) agar ( Sigma catalogue no . A1296 ) . In the Pi-deficient ( P- ) medium , the 1 . 25 mM KH2PO4 in the P+ medium was replaced with 0 . 65 mM K2SO4 . The Arabidopsis seeds were surface sterilized with 20% ( v/v ) bleach for 15 min . After three washes in sterile-distilled water , seeds were sown on Petri plates containing P+ or P- medium . After the seeds were stratified at 4°C for 2 days , the agar plates were placed vertically in a growth room with a photoperiod of 16 h of light and 8 h of darkness at 22–24°C . The light intensity was 100 μmol m -2s -1 . About 100 , 000 M2 seeds representing 6 , 000 EMS-mutagenized M1 plant lines were used for mutant screening . The EMS-mutagenized lines were generated according to [64] . Seeds were plated on P- medium and were grown in the growth room for 7 d . The root hair phenotypes were visually examined under a stereomicroscope ( Olympus SZ61 , Japan ) . The seedlings that had enhanced root hair production compared to the WT were identified as putative mutants and were transferred to soil . The plants were self-pollinated , and the mutant phenotypes were confirmed in the next generation . The mutants were back-crossed to the WT plants twice before they were characterized further . Root surface-associated and total APase activities were quantified as described by [19] . Total RNAs were extract from 7-day-old seedlings with the TIANGEN RNAprep pure plant kit . A 2-μg quantity of RNA was reverse transcribed in a 50-μl reaction using M-MLV reverse transcriptase ( TaKaRa , Kyoto , Japan ) according to the manufacturer’s manual . cDNA was amplified with SYBR Premix Ex Taq ( TaKaRa ) on the Bio-Rad CFX96 real-time PCR detection system . ACTIN2 gene expression was used as an internal control , and the relative expression level of each gene was calculated by the 2–△△Ct method . The genes and the primers used to detect their mRNA expression are listed in S7 Table . Cellular Pi and total P were quantified as described by [19] . For analysis of anthocyanin accumulation , the seeds were sown directly on 1/2 MS agar plates ( 0 . 6% agar ) with different levels of Pi and were placed horizontally in the growth room . The shoots of the seedlings were photographed at 10 DAG using a camera attached to a stereomicroscope ( Olympus SZ61 , Japan ) . The mapping population was generated by crossing the mutant hps5 to a plant of the Landsberg erecta ecotype . F2 progeny that displayed the dark-blue BCIP staining phenotype were selected , and DNAs from these seedlings were isolated for molecular mapping . A set of simple sequence length polymorphism ( SSLP ) and cleaved-amplified polymorphic sequence ( CAPS ) markers was used to map the HPS5 gene . The sequences and chromosomal positions of the molecular markers are listed in S8 Table . For the test of triple responses , seeds were sown directly on P+ medium ( 0 . 6% agar ) containing 1 μM ACC . After the seeds were stratified at 4°C for 3 days , they were placed horizontally in growth room at 22°C in the dark . After 96 h , the seedlings were photographed and the length of hypocotyl was measured . At least 20 seedlings ( n> 20 ) were scored for each measurement . For observation of the accumulation of EIN3-GFP proteins , seeds were first germinated on P+ medium ( 1 . 2% agar ) and grown for 4 days . The seedlings were then transferred to P+ and P- media with or without 10 μM ACC . After 6 days on these media , the roots were excised from the seedlings and were examined with a confocal laser scanning microscope ( Zeiss LSM710 , Germany ) . Excitation wavelengths of 488 nm and emission wavelengths of 507 nm were used to visualize GFP signals . To generate the plant transformation vector 35S::ERS1 for genetic complementation , the genomic sequence of the ERS1 gene was amplified from genomic DNA of WT Arabidopsis plants using primers 5’-AGAACACGGGGGACTCTAGAGGATCCATTGGTTTCTTCTTTATCACAC-3’ and 5’-TTGAACGATCGGGGAAATTCGAGCTCGAATCTGCCACAACCACA-3’ . The ERS1 genomic DNA was cloned into the site after the CaMV 35S promoter of the plant expression vector pZH01 , which carries a hygromycin-resistant gene . To generate the vector ERS1::ERS1 , the promoter and genomic DNA sequences of AtERS1 were amplified from WT Arabidopsis plants using primers 5'- TCTAGAGGCTAAGAAGTCCGAGAGTA -3' and 5'-CCCGGGGAATCTGCCACAACCACA-3' . The PCR products were cloned into the vector pBI101 , which carries a kanamycin-resistant gene . For the vector 35S:: mERS1 , genomic sequences of mutated ERS1 were amplified from the hps5 mutant using the same primers that were used for vector 35S::ERS1 . These constructs were then transformed into the hps5 mutant plants ( 35S::ERS1 and ERS1::ERS1 ) and WT plants ( 35S:: mERS1 ) by the floral dip method . To generate the vector EXPA7::RHS15 and EXPA7::RHS19 , the CaMV 35S promoter of the plant expression vector pZH01 was replaced by the EXPA7 promoter sequences . The genomic DNA sequences of RHS15 and RHS15 were amplified from WT Arabidopsis plants using primers of RHS15-F 5'- AAGAGGCTAGAATGGGATCC CACCATTGTAGTTGAAGAAC -3' , RHS15-R 5'- CGATCGGGGAAATTCGAGCTCAGGTAATCTTATTCTCATTC -3' , RHS19-F 5'- AAGAGGCTAGAATGGGATCC ACCATTCCTGTAAACACAAA -3' , and RHS19-R 5'- GATCGGGGAAATTCGAGCTCGGATTTATTAAACAAGTTTA -3' . The PCR products were cloned into the modified pZH01 vector . For all of the constructions , the PCR-amplified gene fragments were directly ligated to the plant transformation vector as described in [65] . The 7-d-old Seedlings were ground to fine powders in liquid nitrogen . One volume of ice-cold extraction buffer ( 0 . 1 M K-acetate , 20 mM CaCl2 , 2 mM EDTA , 0 . 1 mM phenylmethylsulfonyl fluoride , 20% glycerol ( v/v ) , pH 5 . 4 ) was added to the powders . Samples were gently agitated on ice for 0 . 5 h and then centrifuged at 13 , 000 rpm at 4°C for 15 min . The supernatant was transferred to a fresh tube , and the centrifugation was repeated . About 30 μg of denatured protein was separated on 10% sodium dodecyl sulfate ( SDS ) -polyacrylamide gels . After electrophoresis , the proteins were transferred to a polyvinylidene difluoride ( PVDF ) membrane in transfer buffer ( 25 mM Tris , 43 mM glycine , 20% methanol ) and subjected to western blot analysis using anti-EIN3 antibodies as described in [19] . The RNA-seq analyses were performed at the Core Sequencing Facility of Tsinghua University . Total RNAs were isolated from 7-d-old seedlings using the RNeasy Plant Mini Kit ( Qiagen ) . Poly ( A+ ) RNAs were purified using the Next Poly ( A ) mRNA Magnetic Isolation Module ( New England BioLabs ) and were reversely transcribed into cDNAs using M-MuLV Reverse Transcriptase and Next® mRNA Second Strand Synthesis Module ( New England BioLabs ) . The cDNAs were fragmented and sequencing libraries were prepared by end repairing , “A” tail addition , and adaptor ligation . The libraries were sequenced as 50-mers using HiSeq2500 ( Illumina ) with standard settings . The bioinformatics analyses of the sequencing data were essentially carried out as described in [66] . The raw data of Illumina reads are available at the National Center for Biotechnology Information Sequence Read Archive browser ( http://ncbi . nlm . nih . gov/sra; accession no . SRP071563 ) . ChIP-qPCR assays were performed as described by [67] . A 2-g quantity of 10-d-old 35S::EIN3-GFP seedlings was used for ChIP assays . The chromatins were precipitated using GFP-Trap A beads . The enriched DNA fragments were released by 200 μM NaCl and subjected to qPCR analysis . The primers used for ChIP-qPCR assays are listed in S9 Table . To construct a plasmid for the expression of recombinant EIN3 protein ( amino acids 141 to 352 ) in E . coli cells , the corresponding DNA fragment was amplified by PCR using primers 5’-ACTGGATCCAAGGTTAGGTTTGATCGT- 3’ and 5’-ACTCTCGAGTCAGAAGAATTCATAACTTTT- 3’ . The PCR-amplified fragment was inserted into BamHI and XhoI sites of the pGEX-6p-1 vector ( Pharmacia ) . The biotin-labeled probes of the promoter sequences were generated by annealing the biotin-labeled complementary oligonucleotides . The sequences of various DNA probes are listed in S6 Table . Each binding reaction ( 20 μL ) contained 0 . 1 μg of recombinant protein , 20 fM of labeled DNA probe , and 1 μg of poly ( dI-dC ) in buffer ( 25 mM HEPES-potassium hydroxide , pH 7 . 5 , 100 mM KCl , 0 . 1 mM EDTA , 10% [v/v] glycerol , 1 mM DTT ) ; binding reactions were performed at room temperature for 30 min . The binding reaction products were resolved on 5 . 5% polyacrylamide gels run in 0 . 5X Tris-borate-EDTA . The free and protein-bound DNAs were detected using the Light Shift Chemiluminescent EMSA kit ( Thermo Scientific ) according to the manufacturer’s instructions . | Phytohormone ethylene has previously been known to play an important role in mediating root hair development induced by phosphate starvation; however , the underlying molecular mechanism is not understood . Using combined molecular , genetic , and genomic approaches , we identify a group of genes that affect root hair development by regulating cell wall modifications . Pi starvation increase the stability of EIN3 protein , a key component in the ethylene signaling pathway . The expression of the identified root hair-related genes is enhanced in the EIN3-overexpressing line , but suppressed in the ein3 mutant . Furthermore , EIN3 protein directly binds to the promoter of these genes which are also targeted by a key transcription factor that regulates root hair development . This work not only explains how ethylene mediates root hair responses to phosphate starvation , but may provide a general mechanism for how ethylene regulates root hair development under both stress and non-stress conditions . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"plant",
"anatomy",
"chemical",
"compounds",
"ethylene",
"signaling",
"cascade",
"ethylene",
"gene",
"regulation",
"regulatory",
"proteins",
"brassica",
"dna-binding",
"proteins",
"organic",
"compounds",
"hormones",
"root",
"hairs",
"mutation",
"plant",
"science",
"model",
"organisms",
"plant",
"hormones",
"transcription",
"factors",
"seedlings",
"plants",
"arabidopsis",
"thaliana",
"research",
"and",
"analysis",
"methods",
"proteins",
"gene",
"expression",
"chemistry",
"biochemistry",
"plant",
"biochemistry",
"signal",
"transduction",
"point",
"mutation",
"plant",
"and",
"algal",
"models",
"plant",
"roots",
"organic",
"chemistry",
"cell",
"biology",
"genetics",
"biology",
"and",
"life",
"sciences",
"physical",
"sciences",
"cell",
"signaling",
"organisms",
"signaling",
"cascades"
] | 2016 | The Molecular Mechanism of Ethylene-Mediated Root Hair Development Induced by Phosphate Starvation |
The parameters involved in human cytomegalovirus ( HCMV ) latent infection in CD14 ( + ) and CD34 ( + ) cells remain poorly identified . Using next generation sequencing we deduced the transcriptome of HCMV latently infected CD14 ( + ) and CD34 ( + ) cells in experimental as well as natural latency settings . The gene expression profile from natural infection in HCMV seropositive donors closely matched experimental latency models , and included two long non-coding RNAs ( lncRNAs ) , RNA4 . 9 and RNA2 . 7 as well as the mRNAs encoding replication factors UL84 and UL44 . Chromatin immunoprecipitation assays on experimentally infected CD14 ( + ) monocytes followed by next generation sequencing ( ChIP-Seq ) were employed to demonstrate both UL84 and UL44 proteins interacted with the latent viral genome and overlapped at 5 of the 8 loci identified . RNA4 . 9 interacts with components of the polycomb repression complex ( PRC ) as well as with the MIE promoter region where the enrichment of the repressive H3K27me3 mark suggests that this lncRNA represses transcription . Formaldehyde Assisted Isolation of Regulatory Elements ( FAIRE ) , which identifies nucleosome-depleted viral DNA , was used to confirm that latent mRNAs were associated with actively transcribed , FAIRE analysis also showed that the terminal repeat ( TR ) region of the latent viral genome is depleted of nucleosomes suggesting that this region may contain an element mediating viral genome maintenance . ChIP assays show that the viral TR region interacts with factors associated with the pre replication complex and a plasmid subclone containing the HCMV TR element persisted in latently infected CD14 ( + ) monocytes , strongly suggesting that the TR region mediates viral chromosome maintenance .
Human cytomegalovirus ( HCMV ) is a ubiquitous herpesvirus that infects 60–90% of the population and is usually subclinical , however virus infection can cause severe disease and mortality in immune compromised patients [1] . Disease manifestations include retinitis , pneumonia and hepatitis [2] . HCMV lytic phase of infection is typically studied in cell culture using human fibroblasts ( HFs ) and viral encoded genes are expressed in a temporally regulated manner . HCMV lytic DNA replication requires cis and trans acting factors and results in the production of infectious virus [3]–[9] . Recent high-resolution transcriptome mapping during a lytic HCMV infection , revealed a complex pattern of transcription [10] . This analysis also showed that during lytic infection most viral RNA production is concentrated in four long non-coding RNAs ( lncRNAs ) , RNA2 . 7 ( also known as ß2 . 7 ) , RNA1 . 2 , RNA4 . 9 , and RNA5 . 0 [10] . The high expression level of viral encoded lncRNAs suggests that that these transcripts may be significant factors for regulating viral and cellular processes required for efficient viral replication . Herpesvirus latency is defined as the persistence of the viral genome in the absence of production of infectious virus . Certain properties of latency have emerged from study of the gamma herpesviruses including maintenance of the viral chromosome is as a circular episome [11] , which is controlled by virus-encoded proteins that interact with viral and host cell chromatin [12]–[23] . With respect to maintenance and replication of the HCMV DNA episome , previous studies have quantified the number of genomes during experimental and natural infection [24] , [25] . However the cis acting element required for maintenance of the viral genome is unknown . Lifelong HCMV latency is established in myeloid lineage , from bone marrow-derived CD34 ( + ) progenitors through peripheral blood to CD14 ( + ) monocytes [26]–[33] . Latently infected cells contain HCMV DNA without supporting lytic replication although virus can be reactivated and recovered through differentiation [34]–[40] . The regulation and maintenance of latency is poorly understood although reactivation of latent virus is a major source of virus associated with serious disease and mortality in immunocompromised hosts , hence a better understanding of HCMV latency may lead to treatments to resolve latent HCMV genomes from infected cells . Differentiation of CD34 ( + ) or CD14 ( + ) cells , to macrophages or dendritic cells through the use of various cytokines results in reactivation and subsequent production of infectious virus [41]–[45] . Several in vitro experimental systems have been developed to study HCMV latency . These systems use either cultured CD14 ( + ) monocytes or CD34 ( + ) hematopoietic stem cells . CD14 ( + ) or CD34 ( + ) cells cultured in vitro can be infected with HCMV clinical strains , resulting in latent infection and efficient reactivation of latent virus [26] , [37] , [43] , [46] , [47] . The mechanism of HCMV latent infection is unknown and the mechanism involved in the establishment and maintenance of the latent virus genome has not been addressed in either natural or experimental models of latency . In experimental models , as well as during natural latency in CD34 ( + ) and CD14 ( + ) cells , a limited number of latency specific HCMV transcripts have been identified [41] , [42] , [48] , [49] . Three consistently identified viral genes expressed during latency are UL138 , a TNF modulator , UL111A , a variant cmvIL-10 cytokine and the UL81–82 antisense transcript encoding LUNA [41] , [50] , [51] . HCMV UL138 is expressed during , though dispensable for HCMV lytic replication such that UL138 mutant viruses altered latency [41] , [52] , [53] . This may result from the fact that UL138 has been shown to upregulate TNFR1 and sensitizes infected cells to TNFα [54] , [55] . UL138 has been shown to interact with proteins encoded by nearby genes within the UL133–UL138 locus [52] . HCMV latency UL111A transcript encodes a 139 aa protein , which is a homolog to cellular IL-10 [51] , [56] . Recombinant viruses that lack the ability to express UL111A are capable of establishing latency and efficiently reactivate , however latently infected cells express higher levels of surface MHC class II in the absence of UL111A expression and suggests that UL111A may function to inhibit the recognition of latently infected cells by CD4 ( + ) T cells [57] . The expression of LUNA during lytic infection is regulated by IE72 expression and the interaction with hDaxx [58] . LUNA has recently been implicated as having a role in virus reactivation [59] . To elucidate the factors involved in HCMV latency , we infected CD14 ( + ) monocytes or CD34 ( + ) progenitor cells with an HCMV clinical isolate [37] , [46] . We have now used these experimental latency protocols , coupled with next generation sequencing , to reveal the complete high-resolution HCMV transcriptome ( RNA-Seq ) during early and latent infections . RNA-Seq analysis shows that during HCMV experimental and natural latency in CD14 ( + ) monocytes , viral transcripts encoding UL44 , UL84 , UL95 , UL87 , UL52 , UL50 , LUNA , UL138 and the lncRNAs 2 . 7 and 4 . 9 were detected . For latently infected CD34 ( + ) cells , all of the mRNAs detected in latently infected CD14 ( + ) cells were present , however additional transcripts encoding UL28/29 , UL37/38 , UL114 , UL133/135 and US17 were also detected . Using chromatin isolation by RNA purification ( ChIRP ) , lncRNA4 . 9 was shown to physically interact with the HCMV major immediate early promoter region and results in an enrichment of the repressive H3K27me3 mark at the MIEP during latency suggesting that HCMV latent genomes are silenced by PRC2 interaction . To address the HCMV cis requirements for latency , we show that a plasmid containing the terminal repeat ( TR ) element persisted in latently infected cells , strongly suggesting that this element mediates viral genome maintenance .
Cord blood was received from the Colorado Cord Blood Bank ( Univ . of Colorado ) . Pooled peripheral whole blood for natural infection studies was obtained from Renown Medical Center ( Reno , NV ) and processed with in 5 hours to isolate CD14 ( + ) or CD34 ( + ) cells . All protocols to obtain blood products were approved by IRB and Office of Human Research Protection . The pooled blood samples represented approximately 25 individual HCMV seropositive donors . Cells were isolated using human cord blood CD34 positive selection kit ( Stemcell technologies ) according to manufacturer's instructions . Briefly , samples were incubated with a pre-enrichment cocktail containing antibodies directed against CD66b and glycophorin A . This was step was performed for negative selection of granulocytes and erythrocytes . CD34 selected cells were retained from the non-selected cells by the use of the EasySep magnet . Cells were resuspended in culturing media or for natural infection studies immediately processed to extract total RNA . Human CD14 ( + ) were isolated using positive selection MACS bead and LS columns ( Miltenyi Biotec ) according to manufacturer's instructions and cells were resuspended in culturing media or immediately processed to extract total RNA . Purity of isolated CD14 ( + ) and CD34 ( + ) was assessed by flow cytometry using antibodies for human CD14-FITC or CD34-FITC ( Miltenyi Biotec ) , along with human CD45-Pacific Blue ( BioLegend ) for total cell staining . Isotype controls included IgG2a-FITC ( Miltenyi Biotec ) and IgG1-Pacific Blue ( BioLegend ) . Approximately 0 . 5×106 cells were stained for 30 minutes at 4°C with the fluorescently labeled antibody , washed once and resuspended in 1× PBS with 0 . 5%FBS , 2 mM EDTA and 1% methanol-free formaldehyde . Cells were determined to be free from red blood cells and >95% CD34 ( + ) or CD14 ( + ) . HCMV natural infection was evaluated from pooled blood from 25 seropositive individual donors . From 120 ml of pooled blood approximately 175 , 000 CD34 ( + ) cells and 32×106 CD14 ( + ) were isolated , total RNA was extraction and DNase treated , resulting in 300 ng of RNA for CD34 ( + ) and 28 . 5 mg of RNA for CD14 ( + ) . The RNA was used generate a sequencing library as described below . CD34 ( + ) cells were cultured in IMDM supplemented with 10% BIT serum substitute ( StemCell Technologies ) , 2 mM L-glutamine , 20 ng/ml low-density lipoprotein ( Sigma Aldrich ) , 50 mM 2-mercaptoethanol , 10 ng/ml Stem Cell factor , 10 ng/ml IL-3 , 10 ng/ml G-CSF ( R & D Systems ) [26] , [41] , [60] or cultured in X-Vivo 15 ( lonza ) [27] . Media was refreshed every three days until latency had been established and verified . To reactive latent virus , CD34 ( + ) cells were stimulated to proliferate and differentiate with the addition of 10 ng/ml GM-CSF , 10 ng/ml Flt-3 ligand , 10 ng/ml TPO , and 10 ng/ml TNF for three days , followed by the addition of 50 ng/ml lipopolysaccharide ( LPS ) ( Sigma-Aldrich ) for an additional four days . Human CD14 ( + ) cells purchased ( Lonza and ReachBio ) or isolated from cord blood were maintained in Iscove DMEM ( Hyclone ) supplemented with 20% heat-inactivated FBS ( Atlanta Biologicals ) , 50 ng/mL M-CSF , 50 ng/mL stem cell factor ( SCF ) , 50 ng/mL G-CSF , 50 ng/mL GM-CSF , 50 ng/mL IL-3 ( R&D Systems ) at a density of 1×106 cells/mL on low cell-binding plates ( Nunc Hydrocell ) . Medium was replaced every 3 days . BAC-derived FIX strains were propagated in human foreskin fibroblasts ( HF ) cells . After 12–14 days post infection cells were scraped and subjected to a freeze-thaw to release virus from the cells . Virus titer was determined with standard plaque assay on HF cells . Unless otherwise stated , infections of CD14 ( + ) cells were done at a multiplicity of 5 pfu/cell . Cells were incubated with virus for 1 hr . Cells were then washed twice with Hanks Balanced Salt solution ( HBSS ) . HCMV infected CD14 ( + ) or CD34 ( + ) cells were maintained and monitored for HCMV gene expression by qPCR analysis . Latency was determined to be established when no IE2 gene expression was detected by qPCR , detection of the virus genome and expression of LUNA and UL138 . For reactivation of latent virus , CD14 ( + ) cells were differentiated by adherence to plastic tissue culture dishes supplemented with 100 ng/mL IL-6 ( R&D Systems ) at 16–18 days post infection . The reactivation from latency in either CD14 ( + ) or CD34 ( + ) was monitored for IE2 gene expression using qPCR . For UV inactivation , 120 mls of 5×106 pfu/ml of FIX BAC virus was evenly distributed in a thin layer onto a 150 mm tissue culture dish . The virus was irradiated on ice in a Stratalinker 2400 ( Stratagene ) for 4 min at 9 . 9×105 mJ . UV inactivation was confirmed by the absence of immediate-early ( IE ) gene expression in infected human foreskin fibroblasts using qPCR . Human foreskin fibroblast cells , maintained in DMEM supplemented with 10% FBS , were plated in a 6-well tissue culture plate at 0 . 1×106 cells per well . 24 hours after plating half of the media was removed and replaced with reactivated CD34 ( + ) cells ( approximately 0 . 5×106 ) along with its' media . Cells were co-cultured together for 10–12 days and monitored by the appearance of green plaque formation in the HF cells . Total RNA from experimentally or naturally infected ( as determined above ) CD14 ( + ) or CD34 ( + ) cells was isolated using PureLink RNA mini kit ( Life Technology ) followed by removal of genomic DNA using Turbo DNA-free ( Life Technology ) . Poly-A RNA was enriched from total RNA by Dynabeads oligo ( dT ) 25 ( Life Technology ) according to manufacturers instructions . The resulting Poly-A RNA was used in dUTP based NEXTflex Directional RNA-Seq Kit with Illumina compatible adaptors ( Bioo Scientific ) according to manufacturers instructions . The resulting libraries were verified on a Bioanalyzer High sensitivity DNA chip ( Agilent ) and quantified with real-time PCR using Illumina compatible kit and standards ( KAPA ) . Final libraries were sequenced using an Illumina MiSeq instrument . HCMV transcript discovery and alignment was performed using CLC Genomics Workbench software and strand specific RNA-Seq parameters . Transcripts were aligned to the Fix strain ( VR1814 ) reference genome . Tiling of RNA 4 . 9 with biotin labeled DNA probes retrieves specific RNA 4 . 9 bound proteins and DNA sequences . Original protocol is from Chu et al [61] . All probes were biotinylated at the 3′ end with an 18-carbon spacer arm; probes were designed against RNA 4 . 9 full-length sequence using an online designer at http://www . singlemoleculefish . com , and synthesized at Protein and Nucleic Acid Facility ( Standford University ) . Samples were processed as described previously [62] . Eluted DNA was resuspended in 50 µl of water and used for end point PCR . Primers for the PCR include MIEP-1 forward: GTGTTTGTCCGAAATACGCG , reverse: GCCTCATATCGTCTGTCACC; MIEP-2 forward: GTTACATAACTTACGGTAAATGGCC , reverse: CCAAAACCGCATCACCATG; MIEP-3 forward: GATTTCCAAGTCTCCACCCC , reverse: GCGGTACTTACGTCACTCTTG; MIEP-4 forward: CCCCGCTTCCTTATGCTATAG , reverse: AAGAACCCATGTCCGGAAC; MIEP-5 forward: CTCCTTGCTCCTAACAGTGG , reverse: GTACTGCTCAGACTACACTGC; UL19 forward: CCTGTATGAGCTGTTTCGACG , reverse: GACTCACATCTAGCTCGTCTTC . ChIRP PCR primers are shown in Table S2 in Text S1 . For CD14 ( + ) and CD34 ( + ) culturing media was changed every three days in order to replenish nutrients and cytokines necessary to maintain the cell in culture . The supernatant that was removed from the cells was collected and used for quantitative real-time PCR to obtain relative values for the amount of virus being produced . The supernatant was subjected to two low speed spins , 400×g for 10 minutes in a table top centrifuge to remove any residual cells . For the real-time PCR reaction 5 µl of supernatant was used in a total reaction volume of 20 µl using Taqman Universal 2× master mix ( Life Technology ) and 20× primer-probe ( IDT ) . Relative Ct values of viral supernatant DNA were correlated to standard curve from a known concentration of purified FIX-BAC DNA . Cell viability , density and growth was monitored by using trypan blue exclusion test . Briefly , cells were gently resuspended and a small aliquot was removed . A 1∶1 dilution of cell suspension and 0 . 4% Trypan Blue solution ( MP Biomedicals ) was counted in a hemacytometer to determine and monitor total number of live cells . Evaluation of changes in histone modifications was performed as previously described [62] . Fold-enrichment of histone marks at various genomic loci was calculated as IgG-subtracted %Input of the locus normalized by the IgG-subtracted %Input of the reference gene GAPDH . 3 separate experiments were performed . PCR primers are listed in Table S4 in Text S4 . 5×106 latently infected CD14 ( + ) cells were transfected with 2 . 5 µg each of JMJD3-HA and UTX-HA expression plasmids ( Addgene ) , or 5 milligrams of GFP-control plasmid , and a non-transfected control group . The cells were transfected using Nucleofector device and Amaxa Human Monocyte Nucleofector kit ( Lonza ) according to manufactures instructions . Transfection efficiency was monitored by GFP expression . 72 hours post transfection RNA and protein was harvested using RNA/DNA/Protein purification kit ( Norgen BioTek Corp . ) . Total RNA was subjected to removal of genomic DNA using Turbo DNA-free ( Life Technology ) according to manufacturers instructions . The purified RNA was then used for cDNA synthesis as previously described . The resulting cDNA was quantified using real-time PCR and Taqman primers targeting specific HCMV transcripts . Expression of UTX and JMJD3 was confirmed by Western blot where total cell protein extracts were resolved by SDS-PAGE gel that was subsequently transferred to a polyvinylidene difluoride ( PVDF ) membrane , blocked with 5% nonfat dry milk powder in 1× TBST buffer and reacted with antibodies specific for HA-tag ( Sigma-Aldrich ) . After one hour incubation with the primary antibody the membrane was washed in 1× TBST three times 5 minutes each . The secondary antibody donkey anti-mouse IgG conjugated to Alexa Fluor-680 ( Life Technology ) was diluted in blocking buffer and added to the membrane for 30 minutes as which time it was again washed with 1× TBST three times 5 minutes each . Specific proteins bands were visualized using the Odyssey by LI-COR . Total RNA was isolated from cells using PureLink RNA mini kit ( Life Technologies ) , followed by removal of genomic DNA using Turbo DNA-free ( Life Technologies ) . cDNA was synthesized from 1 µg of total RNA in the presence of random hexamers , dNTPs , and Superscript III reverse transcriptase ( Life Technologies ) . The resulting cDNA was then used along with Taqman Universal PCR Master Mix ( Life Technologies ) and specific primers and FAM labeled probes ( IDT ) in an Eppendorf RealPlex . The following real-time PCR program was used: one cycle 95°C hot start for 5 minutes , and forty cycles of 95°C for 15 seconds and 60°C for 1 minute . Primers used for detection of specific gene expression are shown in Table S3 in Text S1 . 4–5×106 latently infected CD14 ( + ) monocytes were cross-linked by the addition of formaldehyde to a final concentration of 1% , incubated at room temperature for 10 minutes and quenched with 0 . 125 M glycine . After washing with ice cold PBS , cells were lysed in 1 ml/5×106 cells ice-cold lysis buffer ( 0 . 5% NP-40 , 150 mM NaCl , 50 mM Tris , pH 7 . 4 , 1 mM EDTA , and protease inhibitors ) . Lysed cells were dounced followed by collection of the crude nuclear extract by centrifugation . The nuclear pellet was resuspended in 1 ml of RIPA ( 50 mM Tris , pH 7 . 4 , 150 mM NaCl , 1% Triton-X , 0 . 1% SDS , 0 . 5% sodium deoxycholate , 1 mM EDTA ) and sonicated with a Fisher Scientific Sonic Dismembrator and micro-tip at 40% amplitude for 40 cycles of 20 s on followed by 20 s off in a wet ice bath . The sonicated chromatin was collected by centrifugation at 20 , 000×g for 15 min at 4°C to remove cellular debris . Chromatin shearing to 150–200 bp fragments was confirmed by agarose gel electrophoresis . 100 ul of chromatin was reserved for input library preparation and the remainder was pre-cleared for 1 hour at 4°C with 100 ul normal mouse IgG sepharose beads followed by incubation overnight with 20 ug UL84 or 100 ug UL44 mAb . 200 ul of Active Motif magnetic IgG coated beads were prepared by blocking overnight with 5 mg/ml BSA and 200 ug/ml yeast tRNA ( Ambion ) in PBS . Beads were washed once with 5 mg/ml BSA in PBS and added to immunoprecipitated chromatin . Immunoprecipitation was performed for 4–5 hours at 4°C . The beads and immunoprecipitated complexes were washed at room temperature with rotation twice for 1 min with RIPA , five times for 5 minutes each with LiCl2 buffer ( 500 mM LiCl2 , 100 mM Tris pH 7 . 4 , 1% NP-40 , 1% sodium deoxycholate ) , and one brief wash with TE ( 10 mM Tris , pH 8 . 1 , 1 mM EDTA ) . DNA was eluted from the beads by the addition of 200 ul elution buffer ( 1% SDS , 0 . 1 M NaHCO3 ) and reversed cross-linked overnight at 65°C . Reserved input DNA was also reverse cross-linked and treated the same as ChIP samples from this point forward . Reversed cross-linked UL84 ChIP DNA or UL44 Chip DNA and respective Input DNA were purified with a Qiagen min-Elute kit . Purified DNA was quantified using an Invitrogen Qubit Flourometer and sequencing libraries were created from 10 ng of ChIP or Input DNA with a Bioo Scientific NEXTflex ChIP-Seq library preparation kit ( #5143-01 ) . Library integrity was analyzed with an Agilent 2100 Bioanalyzer and quantified using a KAPA q-PCR kit . 12 pM each of UL84 ChIP or UL44 ChIP library and 6 pM of respective Input library were sequenced on an Illumina MiSeq . ChIP-Seq data analysis , including read mapping to FIX genome ( VR1814 ) and peak calling was performed using CLC Genomics Workbench with a maximum false discovery ( FDR ) rate of 5% , a window size of 250 bp , and reads were shifted based on a length of 200 bp . The data is the result of four separate experiments . 10×106 CD14 ( + ) or CD34 ( + ) cells were infected with FIX virus and cells were monitored for IE gene expression using qPCR . Once latency had been established , approximately 18 dpi , total RNA was isolated from cells using PureLink RNA mini kit ( Life Technologies ) , followed by removal of genomic DNA using Turbo DNA-free ( Life Technologies ) . The resulting RNA was used with Qiagen OneStep RT-PCR kit with primers designed to mRNA shown in Table S3 in Text S3 . 20×106 CD14 ( + ) cells were infected with FIX virus . 18 days post infection cells were harvested and processed as described previously [62] . RNA-protein complexes were precipitated by adding 300 µl lysate , 5 µl antibody , UL84 , UL44 , SUZ12 ( Active motif , cat . # 39357 ) , EZH2 ( Active motif , cat . # 39875 ) or GAPDH ( Abcam , ab128915 ) , 25 µl Protein G magnetic beads ( Active Motif ) , and 1 µl RNase Out . Primers used for PCR are shown in Table S1 in Text S1 . CD14 ( + ) monocytes were infected with BAC-derived FIX virus at a multiplicity of 5 pfu/cell . Cells were maintained and monitored for HCMV gene products by qPCR analysis . Latency was determined to be established when no IE gene expression was detected by qPCR analysis . After established latency pGEM7zf ( − ) , pGEM-oriLyt , or pGEM-TR plasmids ( 4 ug ) were added to 3×106 CD14 ( + ) latent or control non-infected cells using Nucleofector 2b device and Human Monocyte Nucleofector kit ( Lonza ) according to manufacture's instructions for program Y-001 on the device . Transfected cells were immediately transferred to 12-well low cell-binding plates ( Nunc Hydrocell ) containing 37°C prewarmed media and placed back into 37°C/5% CO2 incubator . Cells were evaluated for viability using trypan blue and for transfection efficiency using a spike of a plasmid expressing EGFP . Cells were determined to be 80% viable and a transfection efficiency of over 70% was achieved . For analysis of transfection input , 24 hrs post transfection total DNA from 2×105 cells was harvested using Norgen DNA/RNA/Protein purification kit according to manufacture's instructions . DNA was eluted from the column in 100 mls of DNA elution buffer , 50 mls was then used for EcoRI ( New England Biolabs ) restriction digest . The total sample of EcoRI digested DNA was loaded onto a 0 . 8% 1× TAE ( 40 mM Tris-HCl , 20 mM Acetic Acid , 1 mM EDTA ) agarose gel and electrophoresed in 1× TAE buffer until the dye front reached the bottom of the gel . For analysis of plasmid maintenance , approximately 2 million latently infected and transfected ( 15 dpt ) CD14 ( + ) cells/well were prepared , loaded , and electrophoresed on Gardella gels as previously described [63] . DNA was transferred to Zeta-Probe membrane ( Bio-Rad ) by the alkaline transfer method according to the manufacturer's instructions and hybridized to using a 32P-labled pGEM probe . Membranes were hybridized with 5 ng of probe in 10 ml of hybridization buffer ( 1 . 5× SSPE , 1 mM EDTA , 7% sodium dodecyl sulfate [SDS] , 10% [wt/vol] polyethylene glycol ) for 16 hrs at 65°C in a hybridization oven ( Robbins Scientific ) . Post-hybridization washed were performed with 2× SSC and 0 . 1% SDS twice for 15 minutes each at 65°C , and then with 0 . 1× SSC and 0 . 1% SDS twice for 30 minutes at 65°C . Southern blot were exposed and visualized using the Storm Scanner imaging system ( GE Healthcare ) . FAIRE was performed as previously described [64] . Briefly , 3×106 CD14 ( + ) monocytes infected with HCMV strain FIX at 4 dpi or 18 dpi were cross-linked with formaldehyde and lysed , followed by sonication to shear chromatin . 10% of the sheared cross-linked chromatin was reserved for input and the remainder was extracted twice with phenol∶chloroform to remove regions of DNA associated with nucleosomes . The aqueous layer containing open regions of the chromatin corresponding to nucleosome depletion was reserved , reverse cross-linked , and further purified by ethanol precipitation . Input DNA was also reverse cross-linked and purified by ethanol precipitation . Sequencing library preparation was performed using an Illumina Tru-Seq DNA library kit . Input and FAIRE libraries were sequenced by the University of California-Irvine High Throughput Genomics Facility on an Illumina Hi-Seq 2000 . Approximately 60 million reads per library were generated . Read alignment with FIX strain ( VR1814 ) was done using CLC Genomics Workbench . PCR primers were designed to specifically amplify the terminal repeat region using FIX BAC DNA as a template and PrimeSTAR GXL DNA polymerase ( Takara ) . After PCR was completed the product was precipitated with 1/10 volume 3 M NaAcetate and 2 . 5 X volume 100% ethanol . The DNA pellet was resuspended in 100 mls of water and 100 mls of 2× Easy-A Master mix ( Stratagene ) was added . The DNA was incubated at 70°C for 30 minutes and then precipitated with 1/10 volume 3 M NaAcetate and 2 . 5 X volume 100% ethanol . The DNA pellet was resuspended in 20 µl of water and 3 µl was used in pGEM T-easy ( Life Technology ) cloning reaction according to manufacturers instructions . The resulting clones were subjected to restriction enzyme digestion and DNA sequencing to determine clones with the correct insert . Each ChIP was performed with chromatin from approximately 200 , 000 latently infected CD14 ( + ) monocytes . DNA-protein complexes were immunoprecipitated using CDT1 ( Abcam , ab70829 ) , MCM3 ( Abcam , ab4460 ) , or an isotype control antibody . PCR for TR region was performed using PrimeSTAR GXL DNA Polymerase ( Takara ) . Primers to detect the TR region were: Set #1 Forward 5′-ACA CCT CCG ACG TCC ACT ATA TAC CA-3′/Reverse 5′-CGT CCA CAC ACG CAA CTC CAA TTT-3′ or Set #2 Forward 5′- GGT GAC GTC GGA GAC AGG G-3′/Reverse 5′- TGC TGT GGT GTA AGG GTA AGG TGT-3′ . Genomic DNA was isolated from approximately 2 million mock infected , 18 dpi , and 20 dpi FIX BAC latent CD14 ( + ) monocytes using a Norgen DNA/RNA/Protein purification kit . Primers amplifying a region within the TR or IR were Forward: 5′-aacgacagacgaagtacggcacaa-3′ and Reverse 5′-acaaacaccgcagaactccttgacg-3′ . Takara PrimeSTAR GXL polymerase was used for PCR .
Although previous studies investigated viral gene expression in HCMV latently infected CD14 ( + ) and primary hematopoietic progenitor cells [26] , [41] , [42] , [46] , [48] , [49] , [65] , [66] , new technologies now allow for the high-resolution evaluation of the HCMV latent virus transcriptome . While there is little or no understanding with respect to the mechanism involved to maintain or replicate the virus genome in a latent environment , we speculate that viral encoded factor ( s ) play a significant role in viral DNA maintenance . This assumption is solidly based on other herpesvirus systems , the most prominent being the gamma herpesviruses where at least one virus encoded factor is required for viral genome maintenance and replication [14] , [15] , [67]–[71] . Hence , the first step in identifying likely candidates involved in viral genome maintenance/replication is to identify all viral encoded transcripts present during latent infection . CD14 ( + ) monocytes were cultured as previously described using specific growth media and culture plates that did not allow attachment/differentiation of cells [46] . Undifferentiated CD14 ( + ) monocytes were infected with the FIX BAC clinical isolate virus strain and viral immediate early ( IE1/IE2 ) , LUNA and UL138 transcript levels were measured at various days post infection using qPCR ( Fig . 1 ) . Immediate early mRNA was detected at days 2 , 4 , 7 and 11 days post infection ( Fig . 1 ) . Immediate early gene expression was no longer detected by 14 days post infection ( Fig . 1 ) suggesting that HCMV virus genomes were not producing immediate early proteins and lytic replication had concluded . LUNA and UL138 transcripts were still detected , indicating that the virus entered the latent phase ( Fig . 1 ) . HCMV genomic DNA was measured from infected cells at 5 and 18 days post infection by qPCR . The HCMV genome was present in latently infected cells at approximately 4 copies per cell at 18 days post infection . Viral DNA was also detected using traditional PCR at 5 and 18 days post infection ( Fig . 1 , inset ) . We also measured viral genome copy number over a 26-day period of infection ( Fig . 1B ) . Viral genomes were present at approximately 4 copies per cell during latent infection at 16 days post infection ( Fig . 1B ) . Cells were monitored during infection and analyzed to ensure the absence of macrophage or dendritic cell markers ( CD11c+ , CD11b+ , CD141+ , CD303+ , CD11b+ , CD68+ , CD80+ ) . For RNA-Seq , total cellular RNA from latently infected CD14 ( + ) monocytes was extracted at 5 and 18 days post infection . Total cellular RNA was also extracted from HCMV latently infected CD14 ( + ) monocytes , after an 18-day incubation subsequent to treatment with IL-6 and re-plating of cells on a surface that allowed attachment . This lytic virus reactivated sample , along with the 5 and 18 day samples were subjected to next generation sequencing ( RNA-Seq ) . RNA-Seq was performed using an Illumina HiSeq 2000 instrument ( 54 million reads per sample , paired end directional sequencing ) . Data was analyzed using CLC Bio Genomics Workbench software ( RNA-Seq analysis ) using mock-infected RNA as a control , 4 independent biological replicates were used for analysis . Figure 2 is a “peak map” showing the location and relative abundance of transcripts from the HCMV FIX BAC viral genome identified by RNA-Seq . The height of the peaks represents the relative number of reads for the transcripts detected and correlates to the relative abundance of the mRNA in cells . At 5 days post infection almost all ( 99% ) of the HCMV ORFs present within the viral genome were expressed ( Fig . 2A ) . Interestingly , and partly consistent with what was previously reported from an RNA-Seq analysis of HCMV lytic infection , the highest amount of transcript accumulation after a 5-day infection of CD14 ( + ) monocytes occurred from the expression of UL22A , lncRNAs 2 . 7 and 4 . 9 loci ( Fig . 2A ) . UL22A is expressed in HCMV infected dendritic cells during lytic infection but has not been previously described in early infection of CD14 ( + ) monocytes . UL22A is responsible for immune evasion and selectively blocks CCL5 [72] . Immediate early transcripts encoding IE1/2 ( UL122–123 ) were also detected as well as all early transcripts encoding replication proteins ( UL44 , UL70 , UL105 , UL102 , UL54 and UL57 ) ( Fig . 2A ) . Late transcripts encoding viral glycoproteins and capsid proteins were also present at 5 days post infection . At 18 days post infection of CD14 ( + ) monocytes the transcriptome analysis showed that only a subset of viral-encoded RNAs were present in latently infected cells and confirmed that neither immediate early IE2 or IE1 mRNAs were expressed , nor were any other immediate early transcripts detected ( Fig . 2B ) . Transcripts were detected encoding LUNA , UL95 , UL138 , UL87 , UL84 , UL52 , UL50 and UL44 along with transcripts for the long noncoding RNAs RNA2 . 7 and RNA4 . 9 ( Fig . 2B ) . Very low levels of UL111A mRNA were detected , which cannot be seen on the peak map shown ( cutoff was 20 reads ) . A list of all transcripts detected during latent infection is shown in table 1 along with the number of reads from the RNA-Seq analysis . These data are the first reporting of the total virus encoded transcriptome from latently infected CD14 ( + ) monocytes using next generation sequencing . To confirm the presence of RNA transcripts we performed reverse transcriptase PCR ( RT-PCR ) using RNA isolated from 5 and 18 day infected CD14 ( + ) monocytes and primers specific for each of the transcripts detected as well as control transcripts . All transcripts identified from RNA-Seq analysis were detected using RT-PCR , whereas we were unable to detect IE2 , UL13 or US21 mRNAs ( Fig . 2B , Inset figure ) , confirming the results from the next generation sequencing . Lytic reactivation of latent virus was also performed by incubating latently infected ( 18 days post infection ) CD14 ( + ) monocytes with IL-6 and re-plating in tissue culture flasks that allowed for cell attachment . After a 7-day incubation with IL-6 , total cellular RNA was harvested and subjected to next generation sequencing . RNA-Seq analysis showed that HCMV gene expression was observed from almost the entire viral genome consistent with lytic virus replication ( Fig . 2C ) . Together these data show that HCMV latent infection in CD14 ( + ) monocytes is characterized by the absence of transcripts encoding immediate early proteins and the expression of a several specific transcripts , some of which were normally associated with lytic DNA replication and expression of two lncRNAs . To test for the production of virus , supernatants were collected at 18 days post infection from reactivated samples and measured the amount of HCMV DNA by qPCR . In reactivated cells , supernatant virus was detected indicating that lytic virus infection was efficiently reactivated in these cells ( Fig . 3 ) . For CD34 ( + ) cells , we evaluated mRNA expression at 3 days post infection as well as at a time point when IE2 transcripts were no longer detected . RNA was harvested at 3 and 10 days post infection and qPCR was performed . We measured the transcript abundance of IE2 , UL44 , UL84 , RNA 2 . 7 and RNA4 . 9 . At 10 days post infection of CD34 ( + ) cells , no IE2 mRNA was detected , however transcripts for UL44 , UL84 and the two lncRNAs were detected ( Fig . 4A ) . Based on the results of the qPCR showing that latency associated genes were expressed in the absence of detectable IE2 , we harvested total cellular RNA and performed next generation sequencing . The results of the RNA-Seq are shown in figure 4B . For HCMV latent infection of CD34 ( + ) cells we harvested RNA at 11 days post infection and the RNA-Seq peak map shows the presence of the same transcripts detected in latently infected CD14 ( + ) cells ( Fig . 4A , 11 day latent ) . Interestingly , several additional transcripts were detected . RNA-Seq of latently infected CD34 ( + ) cells shows that the mRNAs for UL28/29 , UL37/38 , UL114 , IE1 UL133/135 , UL111A and US17 were detected ( Fig . 4A ) . Hence these data show that there are common transcripts associated with HCMV latent infection in both CD14 ( + ) and CD34 ( + ) cells , including two lncRNAs and mRNAs that encode UL84 and UL44 . At 3 days post infection less than half of the total HCMV genes were expressed and at a very low levels ( Fig . 4B , 3 day PI ) . In reactivated cells , the entire HCMV genome was activated and the relative amount of transcription was about 5-fold higher ( Fig . 4B , reactivated ) . Cells were co-cultured with HFFs and cells were observed for the formation of green plaques . Green plaques were readily visible after a 12-day incubation indicating that efficient reactivation from latent infection was achieved ( Fig . 4B , reactivated inset image of green plaques on HFF cells ) . We also evaluated transcripts present in naturally latent infected CD14 ( + ) monocytes and CD34 ( + ) cells isolated from whole blood from pooled HCMV positive donors ( Renown Hospital , Reno NV ) . Cells were isolated , RNA was extracted and next generation sequencing libraries were generated . RNA-Seq analysis performed on naturally latent infected cells showed that , although the relative abundance of individual transcripts varied between experimental and natural infection , the transcripts present were mostly consistent with experimental latency models ( tables 1 and 2 , tables 3 and 4 ) . These data strongly suggest that the most of the transcripts detected during experimental latency in both CD14 ( + ) and CD34 ( + ) cells mirrors an HCMV natural latent infection and consequently validates in vitro HCMV latency infection protocols used for this study . Since we observed robust transcription at 5 days post infection of CD14 ( + ) monocytes , we wanted to determine if virion associated transcripts could be detected from UV inactivated virus at 5 days post infection and evaluate the ability of virus encoded mRNAs to persist in CD14 ( + ) monocytes . It was previously described that HCMV virions contain virus-encoded mRNAs [73] , [74] . Hence we infected CD14 ( + ) monocytes with UV inactivated virus and performed qPCR at 5 and 10 days post infection . We evaluated infected cells for the presence of transcripts detected during latency as well as IE2 mRNA . As expected all transcripts were detected at 1 hr post infection of CD14 ( + ) monocytes , suggesting that these transcripts are packaged within the HCMV virion ( Fig . 5 , 1 hr UV ) . At 5 and 10 days post infection the amount of fold increase for all transcripts from cells treated with non-UV inactivated virus was consistent with what was observed in the RNA-Seq , strongly suggesting lytic infection ( Fig . 5 , Day 5 and Day 10 UV ) . However in UV inactivated samples , at 5 or 10 days post infection , all transcripts were below 1-fold increase compared to mock infected samples suggesting that RNA transcripts observed during latency are not the result of input virus or from initial transcription observed at 5 days post infection ( Fig . 5 , Day 5 and 10 UV ) . These results indicated that there is a detectable amount of input mRNAs from virions in infected CD14 ( + ) monocytes , however these input transcripts are short lived and absent by 5-days post infection . The RNA-Seq data identified the presence of two transcripts in HCMV latently infected cells that encode UL84 and UL44 , two proteins that participate in virus lytic DNA replication in human fibroblasts [3] , [4] . UL84 and UL44 interact with oriLyt and other regions of the HCMV genome during lytic infection [5] , [6] , [75] , [76] . Also , UL84 interacts with UL44 , the DNA polymerase processivity factor in infected cells [77] . Since UL44 and UL84 were the only two transcripts encoding apparent DNA binding proteins observed in the RNA-Seq analysis in latently infected cells , it allows for the possibility that these two proteins may participate in HCMV latent DNA replication . The presence of these two transcripts in latently infected cells suggests that maintenance of the latent viral genome may involve some of the lytic virus machinery . Hence we investigated if these two proteins interacted with the viral genome under latent conditions in CD14 ( + ) monocytes . ChIP-Seq analysis for UL84 and UL44 was performed in latently infected CD14 ( + ) cells to determine if these proteins interact with the latent HCMV genome in CD14 ( + ) cells . Ten million CD14 ( + ) monocytes were infected with wt FIX BAC virus and mRNA levels were monitored until no IE gene expression was detected ( 18 days ) . Infected cells were then processed for ChIP-Seq and DNA-protein complexes were immunoprecipitated using antibodies specific for UL84 or UL44 . Immunoprecipitated DNA was used to generate libraries for next generation sequencing using input DNA as a reference . ChIP-Seq was performed using Illumina MiSeq instrument . Data was analyzed and peaks were identified using CLC Bio Genomics Workbench software using the ChIP-Seq analysis package compared to input DNA . ChIP-Seq analysis identified 5 major peaks corresponding to the binding domains for UL84 in the FIX BAC virus genome during latent infection ( Fig . 6A , Red peaks ) . Two of the peaks mapped to the promoter regions for UL84 and UL44 and suggests that UL84 may regulate the expression of these genes during latent infection ( Fig . 6A , Red Peaks . ) . Another peak was localized to the promoter region for UL112/113 and the upstream regions regulating LUNA and MIE gene expression . Also , the region containing oriLyt and encoding the lncRNA4 . 9 showed specific binding to UL84 in latently infected cells ( Fig . 6A , Red peaks ) . For UL44 , many of the binding domains in latently infected monocytes overlapped with those identified for UL84 . UL44 interacted most prominently with the upstream region encoding UL95 ( Fig . 6A , Blue peaks ) , one of the transcripts identified from the RNA-Seq analysis from latently infected CD14 ( + ) monocytes . Interestingly , UL44 also interacted with the region of the genome encoding lncRNA2 . 7 and , along with UL84 , the region upstream of its own coding sequence ( Fig . 6A , Blue peaks ) . Table 5 is shows the number of reads for UL44 and UL84 ChIP-Seq experiments corresponding to the peak map . These data indicate that UL84 and UL44 interact with the HCMV genome during latent infection in CD14 ( + ) monocytes and suggests these proteins may activate or suppress several genes during latency and allows for the possibility that viral encoded replication proteins may participate in viral genome maintenance . Two HCMV transcripts identified in latently infected CD14 ( + ) monocytes and CD34 ( + ) cells were long noncoding RNAs . Our previous studies investigating Kaposi's sarcoma-associated herpesvirus ( KSHV ) lncRNA PAN , showed that herpesvirus encoded lncRNAs can be regulators of both viral and cellular gene expression [62] . lncRNAs associate with chromatin modifying complexes [78] and interact with components of the polycomb repressive complex 2 ( PRC2 ) . The PRC2 complex is composed of EZH2 , SUZ12 and EED-1 . EZH2 is a protein that adds three methyl groups to lysine 27 of histone 3 [79] . SUZ12 is a protein that contains a zinc finger domain that is the point of contact with RNA [80] . EED-1 interacts with HDAC1 and histone deacetylase and various other proteins to mediate gene repression [81] . These PRC2 proteins can mediate changes in histone modifications ( methylation ) and subsequent repression of gene expression from various genetic loci . Hence the interaction of lncRNAs with PRC2 can globally influence gene expression . Since the RNA-Seq identified HCMV encoded lncRNAs 2 . 7 and 4 . 9 during latent infection , we investigated if the newly discovered RNA4 . 9 could interact with components of the PRC2 . Also , since it was previously described that HCMV UL84 is an RNA binding protein [82] and UL84 is also present during latent infection , we also evaluated if lncRNA4 . 9 interacted with UL84 in latently infected CD14 ( + ) monocytes . HCMV latently infected CD14 ( + ) monocytes were fixed and RNA crosslinking immunoprecipitation ( rCLIP ) was performed using antibodies specific for EZH2 , SUZ12 and UL84 . Also , as controls we used an isotype control antibody for immunoprecipitations . Immunoprecipitated RNA-proteins complexes were reverse cross-linked and cDNA was generated . cDNA was used for PCR amplification using primers specific for RNA4 . 9 , cyclophilin A or UL138 RNA . PCR amplification products were observed for immunoprecipitations using UL84 , EZH2 and SUZ12 antibodies and PCR primers specific for RNA4 . 9 ( Fig . 6B ) . No PCR amplification product was detected when the isotype control antibodies or an antibody to the cellular protein GAPDH was used or when using PCR primers specific for UL138 or cyclophilin A ( Fig . 6B ) . Also , no PCR product was observed when the reverse transcriptase was omitted from the PCR protocol ( Fig . 6B , RNA4 . 9 , no RT ) . These results show that HCMV RNA4 . 9 interacts with viral encoded UL84 and components of the PRC2 . Hence , RNA4 . 9 has the potential to function as a regulatory RNA in the context of HCMV latent infection of CD14 ( + ) monocytes . Since it was demonstrated that RNA4 . 9 interacts with components of the PRC2 we wanted to evaluate if these same factors bound to the MIEP under latent conditions . We performed ChIP assays using latently infected CD14 ( + ) cells and immunoprecipitated protein-DNA complexes with antibodies specific for EZH2 or SUZ12 . Immunoprecipitated DNA was subjected to PCR using primers specific for the core promoter region for the major immediate early gene locus ( PCR primer set MIEP-3 , see figure 7 and Table S4 in Text S1 ) . Both SUZ12 and EZH2 were shown to interact with the MIEP region under latent conditions ( Fig . 6C ) . We also evaluated the LUNA promoter during latency for the presence of SUZ12 and EZH2 . Consistent with the constitutive expression of LUNA during latency no specific interaction of PRC2 components was observed ( Fig . 6C , LUNA-Pr ) . Control immunoprecipitations using isotype specific antibodies shown no PCR product . These experiments , coupled with the observation that RNA4 . 9 interacts with polycomb proteins , suggest that RNA4 . 9 could mediate gene suppression at the MIEP during latency . As mentioned previously , one of activities of lncRNAs is to mediate changes in gene expression by an interaction with PRC and chromatin . Since we observed the presence of RNA4 . 9 during latent infection and the transcript was shown to interact with UL84 and PRC2 factors , it is logical to assume that RNA4 . 9 is involved in regulation of gene expression during latency . The ChIP-Seq analysis showed that the major interaction domain for UL84 was at the MIE promoter ( MIEP ) region . Therefore we investigated if RNA4 . 9 interacted with the latent HCMV genome at the MIEP . We performed chromatin isolation by RNA purification ( ChIRP ) using biotinylated oligonucleotides specific for RNA4 . 9 to determine if the transcript interacts with the MIEP region during both latent and initial infection of CD14 ( + ) monocytes . We evaluated the ability of RNA4 . 9 to interact with the MIEP across 5 regions that mapped to the unique , enhancer , core promoter , exon 1 and the first intron of the IE gene ( Fig . 7A ) . Five and 18 day infected CD14 ( + ) monocytes were used for ChIRP analysis of RNA4 . 9 binding to the MIEP . At 5 days , the ChIRP analysis showed only a slight interaction of RNA4 . 9 with the MIEP and enhancer region ( Fig . 7B CD14 ( + ) 5 dpi , lanes RNA4 . 9 ChIRP ) . However during latent infection a strong interaction with the HCMV MIEP , enhancer and a lesser interaction with the first exon was detected ( Fig . 7B CD14 ( + ) 18 dpi , lanes RNA4 . 9 ChIRP ) . This interaction , coupled with the observation that RNA4 . 9 interacts with components of the PRC2 strongly suggest that one mechanism for repression of immediate early gene expression during latent infection is the epigenetic modification of chromatin regulating MIEP region mediated by the virus encoded lncRNA4 . 9 . Since the interaction of PRC2 with chromatin is associated with the increase in the H3K27me3 repressive mark , we investigated the relative marking of this histone modification at the MIEP region . CD14 ( + ) cells were infected with FIX BAC virus and cells where harvested at 3-days post infection or during latency at 15 days post infection . ChIP assays were performed using antibodies specific for H3K27me3 or control proteins . The amount of H3K27me3 mark was evaluated for enrichment at MIEP-1 , MIEP-2 and MIEP-3 regions ( Fig . 7A ) as well as at the LUNA and GAPDH promoters . During latency , there was a significant increase in enrichment of the repressive H3K27me3 mark at regions MIEP-2 and MIEP-3 of the MIEP region ( Fig . 7C ) . These loci correspond to the regions where RNA4 . 9 binding is the most robust ( Fig . 7B ) . The enrichment of H3K27me3 at the LUNA promoter during latency changes only slightly during latent infection , which is consistent with the constitutive expression of LUNA during lytic and latent phases of infection ( Fig . 7C ) . We also evaluated the enrichment of the transcriptional activation H3K4me3 mark at the MIEP region . This mark does decrease during latent infection , consistent with less gene activation ( Fig . 7C ) . The results of these experiments suggest that RNA4 . 9 interacts with the MIEP region and mediates the enrichment of the repressive H3K27me3 mark to repress HCMV transcription of IE2 . Since we observed the enrichment of the repressive H3K27me3 mark at the MIEP region we investigated if the over expression of specific H3K27me3 demethylases could have an affect on reactivation , specifically the release of repression of IE2 gene expression . Plasmids expressing UTX and JMJD3 were cotransfected into latently infected CD14 ( + ) monocytes and 72 h post transfection total cellular RNA was harvested and qPCR was performed . We examined mRNA accumulation for IE2 , UL84 , RNA2 . 7 and UL54 ( polymerase ) . As a control , we transfected a plasmid that expressed EGFP . mRNA expression levels were compared to mock transfected latently infected CD14 ( + ) cells . In cells cotransfected with UTX and JMJD3 expression plasmids there was a 130 fold increase in IE2 mRNA accumulation ( Fig . 7D ) . Transcripts encoding RNA2 . 7 , UL84 and UL54 also increased markedly in cells transfected with demethylases expressing plasmids ( Fig . 7D ) . Latently infected CD14 ( + ) cells transfected with the EGFP expressing plasmid show no significant increase in mRNA accumulation ( Fig . 7D ) . These experiments strongly suggest that latent HCMV genomes are silenced , at least partly , by repressed chromatin marked with H3K27me3 . Also , our data suggests that RNA4 . 9 participates in transcriptional suppression of MIE gene expression . We postulate that the region ( s ) of the viral genome that could potentially serve as latent replication/maintenance elements should be depleted of nucleosomes and hence would define the major cis regulatory elements within the latent HCMV chromosome . This assumption is based on previous studies showing that origins of replication are depleted of bulk nucleosomes [68] , [83]–[86] . It was further demonstrated in Saccharomyces cerevisiae that nucleosomes are positioned as flanking replication origins [87] . We employed the method Formaldehyde Assisted Isolation of Regulatory Elements ( FAIRE ) to determine which regions within the latent HCMV chromosome are depleted of nucleosome structures [88]–[90] . FAIRE is a powerful approach to identify genome-wide active regulatory elements in vivo . FAIRE involves formaldehyde crosslinking of cells , followed by shearing of chromatin and subsequent phenol/chloroform extraction . The procedure is based on the fact that histones will crosslink with DNA and nucleosome free chromatin will be preferentially partitioned into the aqueous phase . These genomic regions are then mapped back to the HCMV genome using next generation sequencing ( Fig . 8A ) . To study the HCMV chromosome during latent infection we used the CD14 ( + ) monocyte experimental model [46] . In this experimental latency model , CD14 ( + ) monocytes are cultured using specific growth media and culture plates that retain CD14 ( + ) monocytes in an undifferentiated state . Cells were infected with the HCMV clinical isolate FIX BAC strain and monitored for the expression of immediate early gene expression and the presence of the previously described latency associated transcripts UL138 and LUNA [41] , [53] , [91] . At 14 days post infection of CD14 ( + ) monocytes , the level of IE1/2 mRNA was undetectable by qPCR , however UL138 and LUNA transcripts were still present ( not shown ) . Latently infected cells were harvested and subjected FAIRE followed by next generation sequencing . CD14 ( + ) monocytes were infected with FIX BAC virus and cells were harvested at 4 and 18 days post infection . At 18 days post infection cells expressed latency-associated transcripts and were absent for expression of immediate early transcripts . Cells were subjected to FAIRE and DNA was analyzed using next generation sequencing ( 60 million reads per sample , paired end sequencing ) . FAIRE data was analyzed using CLC Bio Genomics Workbench software ( ChIP-Seq analysis ) . At 4 days post infection of CD14 ( + ) monocytes over 130 peaks or nucleosome-depleted regions were elucidated ( Fig . 8B ) . These regions were distributed across the HCMV genome with a prominent peak at the major immediate early promoter region ( Fig . 8B ) . These “open” active regions across the HCMV genome are consistent with robust transcription and replication during lytic infection . The FAIRE profile was significantly different during a latent HCMV infection at 18 days . Nucleosome depleted regions of the viral genome were consistent with what was observed from the RNA-Seq analysis ( Fig . 2B ) , in that active regions were specific for loci of the viral chromosome where latent viral transcription was observed ( Fig . 8C ) . One obvious exception was the high degree of sequence reads detected for the inverted and terminal repeat region ( TR ) of the genome ( Fig . 8C , Red Arrow ) . Although nucleosome depleted regions were identified at both IR and TR regions of the genome , next generation sequencing could not distinguish between the two regions because of the high degree of homology . Also , although a portion of the IR region is not present in FIX BAC , the IR region was identified because the FAIRE-Seq reads were mapped to the annotated parent virus strain ( VR1814 ) , hence the nucleosome depleted regions were localized only to the TR regions of the genome . The presence of these highly nucleosome depleted loci , coupled with the observation that no viral transcripts were detected from this region by RNA-Seq during latent infection , strongly suggests that these regions of the genome could act as elements that mediate DNA replication/maintenance . Since FAIRE data strongly suggested that the TR region was depleted of nucleosomes we investigated if the TR region associated with the protein components of the pre-replication complex ( pre-RC ) . HCMV latently infected CD14 ( + ) monocytes were fixed and a chromatin immunoprecipitation ( ChIP ) assay was performed using antibodies specific for pre-RC proteins MCM3 and CDT1 proteins . We also performed immunoprecipitations using isotype control antibodies and an antibody to GAPDH . Figure 9A is a schematic showing the HCMV genome and the location of the terminal repeat ( TR ) region of the viral DNA . Also shown are primer sets used for amplification of DNA immunoprecipitated from the ChIP assay . Immunoprecipitated protein-DNA complexes were amplified using PCR primers designed to amplify the TR ( Fig . 9A , set 1 and 2 ) of the HCMV genome . As a control we also used primers specific for the UL25 genomic locus . PCR products were observed for ChIP samples using CDT1 or MCM3 specific antibodies when primer set 1 was used in the PCR mixture ( Fig . 9B ) . No specific amplification product was observed from samples using isotype control or GAPDH antibodies , or using primers designed to amplify the UL25 region ( Fig . 9B ) . These data strongly suggest that the TR region of the HCMV genome interacts with factors involved in cellular DNA replication and is consistent with previous results from other herpesvirus systems and latent origins . Since our data suggested that the terminal repeat region of the HCMV genome was involved in replication/maintenance of the latent viral genome , our next step was to develop an assay to evaluate the ability of the TR element to mediate genome maintenance in latently infected cells . We postulate that if the TR element mediates viral chromosome maintenance then a plasmid containing the TR element would persist in latently infected cells . To this end we subcloned the HCMV FIX BAC TR element , as it would exist in the circular genome form ( Fig . 10A ) , since previous data indicated that latent HCMV genomes exist as a circular episome [92] . This TR subclone , pTR , would be used to transfect HCMV latently infected CD14 ( + ) monocytes where required viral and cellular encoded factors would be supplied in trans from resident viral DNA and the host chromosome ( Fig . 10B ) . Latently infected CD14 ( + ) monocytes were transfected with pTR using Amaxa nucleofector and cells were processed as shown in figure 7C . As controls , we also transfected the parent vector pGEM and a plasmid that contains the HCMV origin of lytic DNA replication , oriLyt . After 15 days post transfection ( 26 days post infection ) cells were prepared and DNA was resolved using a Gardella gel [63] . The gel was transferred to a nylon membrane and hybridized to a radiolabeled pGEM probe . Several bands were detected in the lane that contained cells that were transfected with the TR-containing plasmid whereas lanes that contained samples that were transfected with either pGEM , cloned oriLyt or mock infected TR transfected cells failed to show a hybridization product ( Fig . 10D ) . Two bands were detected on the Southern blot from cells containing pTR . These two bands could be due to variable lengths of the repeat sequence present with the TR region of the genome . We also evaluated 26 day latently infected CD14 ( + ) that were transfected with oriLyt or TR containing plasmids . All transfected cells expressed latency-associated transcripts , in the absence of IE2 gene expression , detected in the RNA-Seq analysis at 26 days post infection ( Fig . 10D , inset graph ) . These data show that the cloned HCMV TR element is capable of persisting in latently infected CD14 ( + ) monocytes . This is the first report describing the existence of a DNA element within the HCMV genome that mediates maintenance of the viral chromosome during latency .
To date HCMV latency is defined as the lack of production of infectious virus , absence of immediate early gene expression and the presence of a few specific latency-associated transcripts , as well as the ability to reactivate latent resident viral DNA . Previous studies have identified the expression of some latency-associated transcripts in both naturally and experimentally infected CD14 ( + ) and CD34 ( + ) cells [41] , [42] , [48] , [49] . We utilized two experimental systems to study HCMV latency; the first system developed for CD14 ( + ) monocytes is where cells are cultured in an undifferentiated state by using specific cytokines and a growth surface that retards cell attachment [46] . Using CD14 ( + ) monocytes cultured under these specific conditions resulted in the presence of the HCMV genome in the absence of immediate early gene expression after approximately 18 days post infection . For the HCMV latency infection protocol used here , we treated CD14 ( + ) monocytes with media formulation previously shown to maintain monocytes in an undifferentiated state [46] . Although it was previously reported that IE gene expression was not detected at 11 days post infection , in our hands the loss of IE gene expression required the culturing of cells for approximately 14 days . Hence , to ensure that viral infection was indeed in a latent phase , we evaluated transcription at 18 days post infection and only when cells were negative for IE gene expression . RNA-Seq has major advantages over other previously employed methods to identify HCMV latency associated transcripts . RNA-Seq allows for a quantitative unbiased evaluation of transcripts present in infected cells . After a five-day infection , RNA-Seq showed the presence of transcripts originating from most of the HCMV genome . Although this observation suggests that a lytic infection precedes the establishment of latency , more experiments are needed to confirm this finding . Previous studies demonstrated that viral gene expression was required for establishment of latency and several virus-encoded genes were identified [39] , [48] . Although at present we do not know the significance of wide spread gene expression in CD14 ( + ) at early time points post infection , a lytic-type infection upon initial infection of primary cells was observed previously for KSHV and EBV [93]–[95] . In those systems , this initial burst of lytic replication may be critical for subsequent establishment of latency . The observed lytic infection for HCMV in monocytes warrants further investigation and will be the focus of future studies . The second experimental latency system used infection of CD34 ( + ) cells where we isolated CD34 ( + ) cell populations , since evidence suggests that these cell types support HCMV latent virus [96] . RNA-Seq showed that infected CD34 ( + ) cells did not undergo the same robust HCMV viral gene expression pattern as observed with initial infection of CD14 ( + ) monocytes . At three days post infection of CD34 ( + ) cells only a subset of genes were expressed , whereas in CD14 ( + ) cells almost the entire HCMV genome showed active transcription . Although the same core transcripts , including the lncRNAs RNA2 . 7 and RNA4 . 9 , were detected in latently infected CD34 ( + ) cells and CD14 ( + ) monocytes , some differences were also observed . One of the most notable was the presence of IE1 mRNA in latently infected CD34 ( + ) cells , which was also confirmed in naturally infected CD34 ( + ) cells . The presence of IE1 mRNA in latently infected CD34 ( + ) cells was observed previously [26] and allows for the possibility that this protein could tether the HCMV chromosome to genomic DNA since it was previously demonstrated that IE1 interacts with cellular DNA [97] , [98] . The lack of detection of IE2 , late gene expression or early genes involved in lytic DNA replication argue against the possibility that that a population of cells are undergoing lytic replication . Previous studies have examined the expression patterns in HCMV latently infected cells . All of these previous studies used microarrays to evaluate the presence of mRNAs [26] , [41] , [48] . Although these studies yielded important information regarding gene expression during latent infection , RNA-Seq has several advantages over microarrays [99]–[101] . One major advantage is that RNA-Seq is unbiased in that no prior knowledge about mRNA sequence is required . One example of this demonstrated in the present study is that most microarrays target known ORFs , hence the presence of expressed lncRNAs or antisense transcripts may be overlooked . Also , recent studies indicate that RNA-Seq is more sensitive than conventional microarrays [102] , [103] . One of the most striking findings from the study presented here is the observation that transcripts that encode proteins associated with lytic DNA replication were expressed during latency . These transcripts were present even though immediate early mRNAs were not expressed after approximately 14 days post infection of CD14 ( + ) monocytes . Hence the observed expression of transcripts identified was not the result of transactivation from IE2 . One other explanation for the presence of apparent lytic mRNAs is the detection of transcripts with long half-lives that were still present from the initial infection and lytic replication . This is unlikely since infection of CD14 ( + ) monocytes with UV-inactivated virus indicated that , although all latent RNAs were detected at 1 hr post infection due to apparent packaging in virions , we were unable to detect these transcripts at 5 or 10 days post infection . This suggests that these mRNA half-lives are less than 5 days in CD14 ( + ) monocytes . The RNA-Seq analysis showed that mRNAs encoding ORFs UL50 and 52 were present in latently infected cells . The expression of these ORFs during latency is interesting since UL52 is implicated in cleavage-packaging . UL52 a protein required for virus growth in human fibroblasts is localized to the nucleus and appears to enclose replication compartments . Although implicated in encapsidation/cleavage of virus DNA , UL52 was not associated with other proteins known to perform these functions [104] . However , UL52 is quite unique with respect to other proteins involved in cleavage and packaging in that it is localized to the nucleus and found in replication compartments [104] . Hence , it could be that UL52 supplies a function in latency that is related to replication of the virus genome during latent infection . For UL50 , this protein is implicated in nuclear egress and is associated with the nuclear lamina as part to the nuclear egress complex ( NEC ) [105] . UL50 associates with the cellular factors p32 and protein kinase C ( PKC ) . UL50 is localized to the inner nuclear membrane and associated with the nuclear lamina along with UL53 [105] , [106] . UL50 was shown to associate with BiP and this interaction was essential for phosphorylation of the nuclear lamina [106] . UL50 along with UL53 may act to remodel the nuclear lamina [107] . We speculate that UL50 , like UL52 , may play a role in maintaining the integrity of the HCMV latent virus genome . In both latently infected CD14 ( + ) and CD34 ( + ) cells the transcripts encoding UL95 and UL87 were among the most abundant . This observation also occurred in naturally infected cells . UL87 and UL95 encode essential early transcripts that produce proteins that apparently affect late gene expression including the UL44 late kinetic transcription and colocalize with UL44 prior to initiation of viral DNA synthesis [108] . Interestingly , the pre expression of UL87 inhibited the expression of MIE genes and virus DNA replication [108] . UL87 and UL95 are recruited to replication compartments during lytic infection [108] . Hence , during latent infection UL87 and UL95 could serve to enhance the expression of UL44 and/or UL50 and 52 . Another possibility is that UL87 and UL95 may serve to help suppress the expression of IE2 during a latent infection . UL84 is the putative initiator protein for lytic DNA replication and interacts with UL44 [5] , [76] , [82] , [109] , [110] . UL84 is a phosphoprotein that exhibits nucleocytoplasmic shuttling that is required for function [82] , [111] . UL44 is the DNA polymerase processivity factor , however these herpesvirus proteins have recently been shown to have diverse functions with respect to regulation of gene expression and initiation of DNA synthesis [112]–[117] . The observation that transcripts encoding UL44 and UL84 are produced in latently infected cells suggests that the maintenance of the HCMV virus genome may involve a mechanism that utilizes some of the HCMV lytic DNA synthesis machinery . UL84 interacts with both Ku70 and Ku80 and the Ku70/80 complex is involved in DNA repair [76] . Ku70/80 is and ATP dependent helicase that is involved with DNA repair and interacts directly with the RNA component ( hTR ) of telomerase [118] , [119] . These interactions suggest that stability of the HCMV genomic DNA in latently infected cells may also involve DNA repair enzymes and telomerase . We also demonstrated that UL84 protein interacted with the HCMV latent genome . We speculate that UL84 may act to suppress IE gene expression , however UL84 may also mediate its own expression and that of UL44 and LUNA . The interaction of UL84 with oriLyt could mean that the protein acts to suppress the activation of the lytic replication . However , it is just as plausible that UL84 ( possibly in cooperation with cellular factors ) activates the promoter region within oriLyt to facilitate the expression of RNA4 . 9 . Another possibility could be that UL84 is acting in the capacity of a replication factor at oriLyt to replicate the latent genome . Interestingly , UL84 is dispensable for lytic DNA replication in one clinical isolate , BACmid clone ( TB40/E ) [120] . Hence one possibility is that in clinical isolates UL84 is not required for lytic replication , but may mediate latent genome maintenance . LncRNAs have emerged as significant regulators of gene expression in human cells . Although the mechanisms used by lncRNAs vary , one well-defined mechanism involves their interaction with chromatin modifying complexes . lncRNAs can act as molecular scaffolds to mediate epigenetic changes in histones , which results in activation or suppression of gene expression [121]–[124] . Hence , lncRNAs can globally affect gene expression patterns . We have recently defined one mechanism of action for KSHV PAN RNA in lytically infected cells . PAN RNA is a highly abundant transcript and we show that expression of PAN RNA can result in the disregulation of genes involved in immune response and cell cycle [62] . This suppression of gene expression is most likely achieved by the interaction of PAN RNA with polycomb proteins . Activation of KSHV and cellular gene expression by PAN RNA appears to involve its interaction with the demethylases UTX and JMJD3 , which remove the repressive H3K27me3 mark [62] . Therefore evidence shows that viral lncRNAs are significant regulators of both cellular and viral gene expression . In HCMV , the lncRNA2 . 7 was shown to regulate the apoptosis pathway during lytic infection [125] . Since we also observed the expression of RNA2 . 7 in latently infected cells it is likely that this transcripts plays the same role in latently infected CD14 ( + ) cells . Early studies did not detect the presence of this transcript in latently infected CD14 ( + ) cells [35] . One possibility for the discrepancy between this early study and the present one is that we used highly sensitive RNA-Seq , where the early studies used traditional RT-PCR . The lncRNA4 . 9 was recently discovered by next generation sequencing of the HCMV transcriptome during lytic infection [10] . RNA4 . 9 initiates within oriLyt and extends upstream and terminates just downstream of UL69 [10] . Hence , since of the location of the 5 prime start of transcription for RNA4 . 9 is within oriLyt this suggests that the promoter region is also within oriLyt . We previously identified a bidirectional promoter within oriLyt just upstream of the RNA4 . 9 transcriptional start site [126] . We now show data that strongly suggests that RNA4 . 9 may act as a regulatory RNA with the potential to control cellular and viral gene expression during HCMV latency in CD14 ( + ) monocytes . This is the first reporting of an HCMV encoded RNA that interacts with the PRC . Using ChIRP we show that RNA4 . 9 physically interacts with the HCMV latent genome in the MIEP region . This observation , combined with the data showing that RNA4 . 9 interacts with PRC proteins that are also bound to the same locus suggests that this transcript may repress the expression of IE2 during latency . Previous studies have demonstrated epigenetic regulation of IE gene expression [27] , [37] , [47] , [127] . These earlier reports investigated the deposition of the acetylated H4 mark within the MIEP . PRC2 is associated with the repressive H3K27me3 mark; hence we examined the presence of this mark at the MIEP . Recent studies have shown that the presence of H3K27me3 is associated with herpesvirus latent genomes and repression of lytic gene expression [128]–[131] . One of the key factors for mediating gene silencing is lncRNAs [132] , [133] . Data presented here strongly implicates lncRNA4 . 9 as a targeting factor for suppression of IE gene expression during latent infection in CD14 ( + ) cells . For CD34 ( + ) latently infected cells , the presence of UL28/29 and UL37/38 is interesting . It was recently reported that this locus encodes a spliced transcript and stimulate the accumulation of immediate early RNAs [134] . UL28/29 proteins interact with nucleosome remodeling and deacetylase protein complex , NuRD along with UL38 and UL28/29 enhanced activity of the MIEP [135] . Recently , UL28/29 was shown to interact with UL84 , p53 and suppresses cellular gene expression during lytic infection [136] . Hence the expression of UL28/29 during latency suggests viral control of p53-regulated genes in the context of a latent infection . Although RNA-Seq showed some differences between the CD34 ( + ) and CD14 ( + ) latent transcriptome , many transcripts were common to both systems including the presence of lncRNAs , UL84 and UL44 . One explanation for the presence of different factors could be that specific viral factors may be required to maintain genome latency in a cell type specific manner . Our RNA-Seq evaluation of HCMV naturally latently infected CD14 ( + ) and CD34 ( + ) cells isolated from seropositive donors showed an almost exact match to the transcripts observed from experimental latency , although relative abundances differed . Since pooled blood from HCMV seropositive donors was used , we do not know the prevalence or degree of transcript expression in individual donors . Nevertheless , data shows that experimental latency closely matches natural latent infection . This is the first evaluation of natural latent infection by next generation RNA sequencing . Interestingly , RNA-Seq detected the previously transcript originating from UL126a in naturally infected cells [137] , [138] . This transcript was not detected in our experimental latency model . One explanation for the discrepancy is that some differences between natural and experimental latency may exist . This could be due the fact that natural infection data was obtained from pooled donors and hence variations in expression levels are reflected in the RNA-Seq analysis . Nevertheless , overall the transcripts present during natural and experimental infection vary little . This is a very significant finding in that HCMV CD14 ( + ) and CD34 ( + ) experimental latency systems are widely used . Our data suggests a strong link between transcripts expressed during HCMV natural latent infection and those that occur in cell culture . It was previously demonstrated that herpesvirus latent origins are marked by lack of nucleosome structure and the presence of cellular factors involved in DNA replication or licensing [139] . Also , it was shown that nucleosome assembly proteins interact with herpesvirus-encoded proteins to regulate replication at the latent origins replication [140] , [141] . For EBV , EBNA-1 was shown to destabilize nucleosomes at the latent origin [142] . Also , for EBV , cellular factors play a significant role in replication of the latent virus origin [143]–[145] . Hence previous evidence indicates that specific herpesvirus and cellular proteins bind to DNA domains to mediate latent viral genome replication and maintenance . Currently little data exists regarding the mechanism involved in maintenance of the viral chromosome in HCMV latently infected cells . Previous studies indicated the HCMV latent genome is in a circular form [25] . Based on the latency models from the gamma herpesviruses , it is assumed that the maintenance and replication of the HCMV latent viral genome requires trans and cis acting factors . In this report we show that , consistent with the presence of active chromatin , that the TR region of the HCMV genome are depleted of nucleosomes during latent infection of CD14 ( + ) monocytes . We developed a latent replication/maintenance assay where CD14 ( + ) monocytes harboring the latent HCMV chromosome were transfected with the TR containing plasmid . Control plasmids , including a plasmid that contained oriLyt , failed to persist in latently infected cells . The TR element of HCMV in the circular form contains two “a” sequences and many repetitive DNA sequence motifs . Hence , during replication/maintenance there is a potential for variability of these repeat sequences . The fact that we observed two bands in the Gardella gel may indicate the presence of variable repeated regions . This was also postulated for the KSHV latent origin , which also contains several repeated DNA sequences [15] . Although the Gardella gel clearly shows that the TR plasmid can persist in latently infected cells , further studies are needed to demonstrate if the TR element can direct plasmid replication . Nevertheless , this is the first reporting of a region of the HCMV chromosome that mediates viral DNA maintenance . The identification of cis and trans acting factors involved in HCMV latency in CD14 ( + ) monocytes and CD34 ( + ) cells now allows for a more in depth analysis of the factors required for viral chromosome maintenance/replication . | Human cytomegalovirus ( HCMV ) is a ubiquitous herpesvirus where infection is usually subclinical . HCMV initial infection is followed by the establishment of latency in CD34 ( + ) myeloid cells and CD14 ( + ) monocytes . Primary infection or reactivation from latency can be associated with significant morbidity and mortality can occur in immune compromised patients . Latency is marked by the persistence of the viral genome , lack of production of infectious virus and the expression of only a few previously recognized latency associated transcripts . Despite the significant interest in HCMV latent infection , little is known regarding the mechanism involved in establishment or maintenance of the viral chromosome . We have now identified the transacting factors present in latently infected CD14 ( + ) monocytes and CD34 ( + ) progenitor cells as well as identification of a region of the HCMV genome , the terminal repeat locus that mediates viral DNA maintenance . This is a major step toward understanding the mechanism of HCMV latent infection . | [
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] | [
"medicine",
"infectious",
"diseases"
] | 2013 | Cis and Trans Acting Factors Involved in Human Cytomegalovirus Experimental and Natural Latent Infection of CD14 (+) Monocytes and CD34 (+) Cells |
In the rice blast fungus Magnaporthe oryzae , the cAMP-PKA pathway regulates surface recognition , appressorium turgor generation , and invasive growth . However , deletion of CPKA failed to block appressorium formation and responses to exogenous cAMP . In this study , we generated and characterized the cpk2 and cpkA cpk2 mutants and spontaneous suppressors of cpkA cpk2 in M . oryzae . Our results demonstrate that CPKA and CPK2 have specific and overlapping functions , and PKA activity is essential for appressorium formation and plant infection . Unlike the single mutants , the cpkA cpk2 mutant was significantly reduced in growth and rarely produced conidia . It failed to form appressoria although the intracellular cAMP level and phosphorylation of Pmk1 MAP kinase were increased . The double mutant also was defective in plant penetration and Mps1 activation . Interestingly , it often produced fast-growing spontaneous suppressors that formed appressoria but were still non-pathogenic . Two suppressor strains of cpkA cpk2 had deletion and insertion mutations in the MoSFL1 transcription factor gene . Deletion of MoSFL1 or its C-terminal 93-aa ( MoSFL1ΔCT ) was confirmed to suppress the defects of cpkA cpk2 in hyphal growth but not appressorium formation or pathogenesis . We also isolated 30 spontaneous suppressors of the cpkA cpk2 mutant in Fusarium graminearum and identified mutations in 29 of them in FgSFL1 . Affinity purification and co-IP assays showed that this C-terminal region of MoSfl1 was essential for its interaction with the conserved Cyc8-Tup1 transcriptional co-repressor , which was reduced by cAMP treatment . Furthermore , the S211D mutation at the conserved PKA-phosphorylation site in MoSFL1 partially suppressed the defects of cpkA cpk2 . Overall , our results indicate that PKA activity is essential for appressorium formation and proper activation of Pmk1 or Mps1 in M . oryzae , and phosphorylation of MoSfl1 by PKA relieves its interaction with the Cyc8-Tup1 co-repressor and suppression of genes important for hyphal growth .
Magnaporthe oryzae is the causal agent of rice blast , which is one of the most important rice diseases worldwide . In the past two decades , M . oryzae has been developed as a model organism to study fungal-plant interactions because of its economic importance and the experimental tractability [1–3] . For plant infection , the fungus forms a highly specialized infection cell called an appressorium to penetrate plant cuticle and cell wall [4] . After penetration , the narrow penetration peg differentiates into bulbous invasive hyphae [5] that grow biotrophically inside penetrated plant cells [6] . Various apoplastic and cytoplasmic effectors are known to play critical roles in suppressing plant defense responses during different stages of invasive growth [7] . At late infection stages , lesions are formed and the pathogen produces conidiophores and conidia on diseased plant tissues under favorable conditions . Appressorium formation is initiated when conidia land and germinate on plant surfaces . On artificial hydrophobic surfaces that mimic the rice leaf surface , M . oryzae also forms melanized appressoria . On hydrophilic surfaces , appressorium formation can be induced by cAMP , IBMX , or cutin monomers [8] . Although late stages of appressorium formation is regulated by the Pmk1 MAP kinase , the cAMP-PKA ( protein kinase A ) pathway is involved in recognizing surface hydrophobicity to initiate appressorium formation , appressorium turgor generation , and invasive growth [9–11] . Deletion of the MAC1 adenylate cyclase ( AC ) gene results in mutants that are defective in appressorium formation [12] . In addition to Cap1 AC-interacting protein [13] , heterotrimeric G-proteins and Rgs1 have been shown to function upstream from the cAMP-PKA pathway [2 , 14] . The PdeH high-affinity cAMP phosphodiesterase is also important for successful establishment and spread of the blast disease [15] . The PKA holoenzyme consists of two regulatory subunits and two catalytic subunits . Binding of cAMP with the regulatory subunit results in the detachment and activation of the catalytic subunits [16] . In M . oryzae , the CPKA gene encoding a catalytic subunit of PKA is dispensable for hyphal growth but the cpkA mutant was delayed in appressorium formation and defective in appressorium turgor generation and plant penetration . In addition , the cpkA mutant still responds to exogenous cAMP for appressorium formation on hydrophilic surfaces [10 , 11] , suggesting that another PKA catalytic subunit gene must exist and play a role in surface recognition and infection-related morphogenesis in M . oryzae . In the budding yeast Saccharomyces cerevisiae , three genes , TPK1 , TPK2 , and TPK3 , encode PKA catalytic subunits and the triple mutant is inviable [17] . The fission yeast Schizosaccharomyces pombe has only one PKA catalytic subunit gene , PKA1 , that is important but not essential for normal growth [18] . In the human pathogen Aspergillus fumigatus , the pkaC1 pkaC2 double mutant is delayed in conidium germination in response to environmental nutrients and is significantly reduced in virulence [19] . In the wheat scab fungus Fusarium graminearum , deletion of both CPK1 and CPK2 caused severe defects in growth and conidiation . The cpk1 cpk2 double mutant was sterile in sexual reproduction and nonpathogenic [20] . In the basidiomycete Ustilago maydis , the phenotype of the adr1 uka1 double mutant has similar phenotype with the adr1 mutant and is defective in yeast growth , mating , and plant infection [21] . In S . cerevisiae , Sfl1 is one of the downstream transcription factors of the cAMP-PKA pathway . When functioning as a repressor , it is involved in the repression of flocculation-related genes , including FLO11 and SUC2 [22 , 23] . As an activator , SFL1 is involved in the activation of stress-responsive genes such as HSP30 [24] . The major PKA catalytic subunit Tpk2 negatively regulates its repressor function [25] . In M . oryzae , deletion of MoSFL1 has no obvious effect on vegetative growth but results in reduced virulence and heat tolerance [26] . Several Sfl1-interacting proteins have been identified in the budding yeast , including Cyc8 , Tup1 , and various mediator components [23 , 27] . Although it lacks intrinsic DNA-binding activities , the Cyc8-Tup1 ( also known as Ssn6-Tup1 ) co-repressor complex interacts with various transcription factors with sequence-specific DNA binding motifs , including Sfl1 , Mig1 , Crt1 , and α2 , to negatively regulate different subsets of genes [27 , 28] . In S . cerevisiae , Cyc8 functions as an adaptor protein required for the interaction between Tup1 tetramers and DNA-binding transcription factors [29] . To further characterize the roles of PKA in growth and infection , in this study we generated and characterized the cpk2 and cpkA cpk2 double mutants and spontaneous suppressors of cpkA cpk2 in M . oryzae . Our results demonstrate that CPKA and CPK2 have specific and overlapping functions . The cpkA cpk2 double mutant had severe defects in growth and conidiation and failed to form appressoria or infect plant through wounds . Spontaneous mutations or deletion and truncation mutations in MoSFL1 suppressed the defects of cpkA cpk2 in hyphal growth and appressorium formation but not invasive growth and lesion development . In affinity purification and co-IP assays , MoCyc8 interacted with the full-length but not truncated MoSfl1 . Treatment with exogenous cAMP also reduced the interaction of MoSfl1 with MoCyc8 and MoTup1 . Furthermore , the S211D mutation in MoSFL1 suppressed the growth defect of cpkA cpk2 . In F . graminearum , 29 of 30 suppressor strains of cpk1 cpk2 mutant had mutations in FgSFL1 , with 15 of them truncated of its C-terminal region . Taken together , our results indicate that PKA activity is essential for appressorium formation in M . oryzae , and phosphorylation of MoSfl1 by PKA likely relieves its interaction with the Cyc8-Tup1 co-repressor and suppression of genes important for hyphal growth and appressorium development . The inhibitory function of Sfl1 orthologs on hyphal growth is likely conserved in filamentous fungi because similar suppressor mutations in FgSFL1 were identified in spontaneous suppressor strains of the cpk1 cpk2 mutant in F . graminearum .
In M . oryzae , the PKA regulatory subunit is encoded by SUM1 , a suppressor of the mac1 deletion mutant [30] . In an effort to identify proteins interacting with known virulence factors , including Sum1 , we generated the SUM1-S-tag fusion and transformed it into the wild-type strain 70–15 . Total proteins were isolated from the resulting transformant B22 ( Table 1 ) and subjected to affinity purification and MS analysis after trypsin digestion as described [13 , 31] . CpkA and Cpk2 ( MGG_02832 ) were among the Sum1-interacting proteins identified in all three independent biological replicates ( S1 Table ) . Cpk2 shares 48% amino acid identity with CpkA but it has a shorter N-terminal region ( S1 Fig ) . To confirm their interaction by co-immunoprecipitation ( co-IP ) assays , the SUM1-S , CPKA-3×FLAG , and CPK2-3×FLAG constructs were generated and transformed into the wild-type strain Guy11 in pairs . Western blot analysis with the resulting transformants showed that both PKA catalytic subunits strongly interact with Sum1 ( S2 Fig ) . To determine the function of PKA catalytic subunits , we generated the cpk2 and cpkA cpk2 deletion mutants ( Table 1; S1 Fig ) . On complete medium ( CM ) plates , the cpkA mutant had no obvious growth defects but the cpk2 mutant was slightly reduced in growth rate . The cpkA cpk2 double mutant was viable but it was significantly reduced in growth rate ( Table 2; Fig 1A ) . Unlike cpkA and cpk2 , the cpkA cpk2 double mutant rarely produced conidia . In cultures induced for conidiation , the double mutant produced only a few conidia per plate . Under the same conditions , over 1×107 conidia/plate were produced by the wild type ( Table 2 ) . Unlike cpkA that was delayed in appressorium formation , cpk2 had no obvious defects in appressorium formation ( Table 2 ) . However , the cpkA cpk2 mutant failed to form appressoria on hydrophobic plastic coverslips and GelBond membranes although conidium germination was normal ( Fig 1B; Table 2 ) . Even after prolonged incubation up to 72 h , no appressorium formation was observed in the double mutant . The cpkA cpk2 mutant also failed to form appressoria on barley and rice leaves ( S3 Fig ) . In spray infection assays with two-week-old seedlings of rice cultivar CO-39 , numerous blast lesions were observed on leaves sprayed with Guy11 or the cpk2 mutant but no lesions were caused by the cpkA and cpkA cpk2 mutants ( Fig 1C ) . Because the cpkA cpk2 mutant failed to form appressoria , we conducted injection infection assays . Whereas the cpk2 mutant was as virulent as the wild type , the cpkA mutant failed to cause lesions on intact leaves but caused limited necrosis at the wounding sites . No lesions or necrosis at the wounding sites were observed on leaves inoculated with the cpkA cpk2 mutant ( Fig 1C ) . These results indicate that PKA activities are essential for appressorium formation and invasive growth after penetration in M . oryzae . Although the cpkA cpk2 mutant was blocked in appressorium formation , we noticed that majority ( over 83% ) of its germ tubes were curved one direction after incubation on hydrophobic side of GelBond membranes for 24 h ( Fig 1B ) . Interestingly , when assayed for appressorium formation on the hydrophilic side of GelBond membranes , the majority of the cpkA cpk2 conidia failed to germinate ( Fig 1B; Table 3 ) . For the ones ( <25% ) germinated , germ tubes of cpkA cpk2 mutant failed to form appressoria ( Fig 1B ) . Conidia of the wild-type , cpkA , and cpk2 strains germinated but failed to form appressoria under the same conditions . In the presence of 5 mM cAMP , over 76% of the wild-type germ tubes formed appressoria on hydrophilic surfaces . However , exogenous cAMP had no stimulatory effects on either conidium germination or appressorium formation in the cpkA cpk2 mutant ( Fig 1B; Table 3 ) . Over 75% of the double mutant conidia failed to germinate and the ones germinated failed to form appressoria or display germ tube curling defects in the presence of 5 mM cAMP . These results indicate that the cpkA cpk2 mutant still recognizes surface hydrophobicity for germination and germ tube growth but not for appressorium formation . We then assayed the intracellular cAMP level in vegetative hyphae harvested from liquid CM cultures . In the cpk2 mutant , the intracellular cAMP level was similar to that of the wild type . However , the cpkA and cpkA cpk2 mutants had higher intracellular cAMP levels than the wild type ( Fig 2A ) . In comparison with the wild type , the double mutant was increased approximately 3-fold in intracellular cAMP . These results suggest that reduced or lack of PKA activities results in an increase in intracellular cAMP in M . oryzae . Because the Pmk1 MAP kinase is essential for appressorium formation [34] , we assayed its activation with an anti-TpEY phosphorylation specific antibody . To our surprise , although its expression was not affected , Pmk1 phosphorylation was increased in the cpkA cpk2 mutant ( Fig 2B ) . However , the double mutant was reduced in the expression and phosphorylation levels of Mps1 MAP kinase ( Fig 2B ) that is required for appressorium penetration and conidiation [33] . These results suggest that over-activation of the Pmk1 MAP kinase pathway is not sufficient to stimulate appressorium formation in the absence of PKA activities , and reduced Mps1 activities may be related to conidiation defects of cpkA cpk2 . Interestingly , the cpkA cpk2 mutant was unstable when cultured on the oatmeal agar ( OTA ) plates and fast-growing sectors caused by spontaneous suppressor mutations often became visible in cultures older than 10 days ( Fig 3A ) . Twenty suppressor strains with faster growth rate than the original mutant were isolated . All of them had similar colony morphology and produced more aerial hyphae than the cpkA cpk2 mutant . On average , the growth rate of suppressor strains recovered to approximately 83% of that of the wild type ( S4 Fig ) . Conidiation also was partially rescued in these suppressor strains although to a much lesser degree than the recovery in growth rate ( S4 Fig ) . Although they varied slightly in growth rate and conidiation ( S4 Fig ) , all the 20 suppressor strains had similar defects as strain CCS1 that was described and presented in figures below . Besides having similar colony morphology , suppressor strains produced melanized hyphal tips in aerial hyphae of 10-day-old OTA cultures ( Fig 3B ) . In infection assays with rice seedlings , none of the suppressor strains caused lesions on intact or wounded leaves ( Fig 3C ) . Therefore , mutations occurred in these suppressor strains only suppressed the defects of the cpkA cpk2 mutant in hyphal growth but not plant infection . On artificial hydrophobic surfaces , over 95% of the conidia from the suppressor strains formed appressoria after incubation for 24 h . Interestingly , they also developed appressoria on the hydrophilic surface of GelBond membranes ( Fig 3D ) . However , approximately 40% of appressoria formed by suppressor strains were abnormal in morphology on hydrophobic or hydrophilic surfaces ( Fig 3E ) . Unlike normal dome-shaped appressoria , the majority of appressoria formed by suppressor strains had irregular shapes ( Fig 3E ) . Although they were still melanized , many appressoria formed by the suppressor strains had projections at one side ( Fig 3E ) . These results indicate that suppressor mutations in these strains also only partially rescued the defect of cpkA cpk2 in appressorium morphogenesis . To identify suppressor mutations , we selected eight genes ( S2 Table ) [27 , 35–40] that are orthologous to downstream targets of PKA in the budding yeast , including SOM1 and CDTF1 [40] for PCR and sequencing analysis in the selected suppressor strains CCS1 , CCS4 , CCS7 and CCS14 . Whereas suppressor stains CCS4 and CCS14 had no mutations in these candidate genes , both CCS1 and CCS7 had mutations in the MoSFL1 [26] gene that encodes a transcription factor with a conserved HSF ( heat shock factor ) DNA-binding domain in the N-terminal region ( residues 124–225 ) [26] . In suppressor strain CCS1 , 10 extra nucleotides CCCCCGCCGC were inserted in the coding region of MoSFL1 ( between 1556 and 1557 ) , resulting in a frameshift change at residue P414 . In suppressor strain CCS7 , a 1241-bp deletion occurred in the coding region of MoSFL1 ( Δ405–1645 ) , resulting in the truncation of 78% of its amino acids . In M . oryzae , deletion of MoSFL1 had no effects on hyphal growth although it was reduced in virulence [26] . To confirm whether insertion or truncation mutation in MoSFL1 has suppressive effects , the MoSFL1 gene replacement construct was transformed into the cpkA cpk2 mutant . Bleomycin-resistant transformants were screened by PCR for deletion of MoSFL1 and confirmed by Southern blot ( S5 Fig ) . The resulting cpkA cpk2 Mosfl1 mutant had similar phenotypes as spontaneous suppressor strains ( Fig 4; Table 4 ) , including recovered growth rate and increased conidiation in comparison with cpkA cpk2 . Melanized appressoria were efficiently formed by the triple mutant but it failed to cause lesions on rice leaves , further confirming that loss-of-function mutations in MoSFL1 rescue the growth defect of cpkA cpk2 . Therefore , MoSFL1 must function as a negative regulator of vegetative hyphal growth and phosphorylation of MoSfl1 by PKA relieves its suppressive effects . In suppressor strain CCS1 , the 10-bp insertion in MoSFL1 causes frameshift and results in the truncation of its C-terminal 414–588 aa . Sequence alignment showed that this C-terminal region of MoSfl1 is well conserved among its orthologs from other filamentous fungi , including Fusarium graminearum and Neurospora crassa ( S6 Fig ) . To verify its importance , we generated a MoSFL1ΔCT gene-replacement construct ( S5 Fig ) to delete residues 496–588 of MoSFL1 in the cpkA cpk2 mutant . The resulting cpkA cpk2 MoSFL1ΔCT triple mutants ( S5 Fig ) had the same phenotypes with the cpkA cpk2 Mosfl1 mutant and spontaneous suppressor strains ( Fig 4; Table 4 ) . These results suggested that the C-terminal region of MoSfl1 is essential for its negative regulator function although it has no known protein motifs . Whereas the N-terminal region of MoSfl1 is involved in DNA binding , the C-terminal region may be responsible for protein-protein interactions to suppress the expression of its target genes important for hyphal growth . Because deletion of residues 496–588 is suppressive to cpkA cpk2 , this C-terminal region of MoSfl1 is likely responsible for interacting with other proteins as a negative regulator . To identify proteins differentially interacting with MoSfl1 and MoSfl1ΔCT , the 3×FLAG-MoSFL1 and 3×FLAG-MoSFL1ΔCT constructs were generated and transformed into the cpkA cpk2 mutant . Total proteins were isolated from the resulting 3×FLAG-MoSFL1 and 3×FLAG-MoSFL1ΔCT transformants and used for affinity purification with anti-FLAG M2 beads . Proteins co-purified with MoSfl1 or MoSfl1ΔCT were identified by mass spectrometry ( MS ) analysis after trypsin digestion as described in previous studies [13 , 31] . Based on MS results from three biological replicates , MGG_03196 was the only protein that co-purified with MoSfl1 but not MoSfl1ΔCT ( Table 5 ) . Its ortholog in yeast , Cyc8 ( Ssn6 ) , forms a transcriptional co-repressor complex with Tup1 to regulate genes involved in a wide variety of physiological processes [28 , 41 , 42] . Interestingly , the Tup1 ortholog , MGG_08829 , was one of the proteins that were commonly co-purified with MoSfl1 and MoSfl1ΔCT ( Table 5 ) . However , the number of MoTup1 peptides identified by MS analysis was significantly lower in the MoSFL1ΔCT transformant than in the MoSFL1 transformant ( Table 5 ) , suggesting a weaker interaction of MoSfl1 with MoTup1 when its C-terminal region is deleted . Based on the conserved nature of Tup1 , Cyc8 , and other components , it is likely that the Cyc8 and Tup1 orthologs also form a transcriptional co-repressor complex with MoSfl1 in M . oryzae , which is consistent with the interaction of Sfl1 with the Cyc8-Tup1 complex in yeast [27] . To confirm the importance of the C-terminal region of MoSfl1 in its interaction with Cyc8 , the CYC8-S-tag construct was generated and co-transformed into the cpkA cpk2 mutant with 3×FLAG-MoSFL1 or -MoSFL1ΔCT . The resulting transformants CYS15 and CNC19 ( Table 1 ) were confirmed by western blot analyses for the expression of transforming constructs . In co-IP assays , the MoSfl1 band was detected in both total proteins and elution from anti-S-tag agarose beads in the transformant expressing the CYC8-S and 3×FLAG-MoSFL1 constructs ( Fig 5 ) . However , the MoSfl1ΔCT band was detected only in total proteins isolated from the transformant expressing CYC8-S and 3×FLAG-MoSFL1ΔCT ( Fig 5 ) . These results confirmed that Cyc8 interacts with the full-length MoSfl1 but not MoSfl1ΔCT in M . oryzae . Interestingly , additional bands smaller than MoSfl1 or MoSfl1ΔCT were detected by the anti-FLAG antibody in transformants CYS15 and CNC19 but not in Guy11 , suggesting that MoSfl1 proteins may be cleaved in vegetative hyphae . Because deletion of the C-terminal region of MoSFL1 suppressed the growth defect of cpkA cpk2 , phosphorylation of MoSfl1 by PKA may affect its interaction with the Cyc8-Tup1 complex . To test this hypothesis , the MoCYC8-S and 3×FLAG-MoSFL1 constructs were co-transformed into the wild-type strain Guy11 . In the resulting transformant , treatment with 5 mM cAMP significantly reduced the MoCyc8-MoSfl1 interaction compared to treatment with 10 μM PKA inhibitor ( PKI ) H-89 ( Fig 6A ) . These results indicate that stimulation of PKA activities by exogenous cAMP reduces the interaction of MoSfl1 with MoCyc8 . Therefore , phosphorylation of MoSfl1 by PKA likely reduced the interaction of MoSfl1 with the co-repressor MoCyc8 to negative regulation of hyphal growth-related genes . We also generated transformants of Guy11 expressing the MoTUP1-S and 3×FLAG-MoSFL1 constructs . MoTup1 was found to interact with MoSfl1 in co-IP assays with the resulting transformant ( Fig 6B ) . Treatment with exogenous cAMP also significantly reduced the interaction of MoSfl1 with MoTup1 ( Fig 6B ) . These results indicate that stimulation of PKA activities by exogenous cAMP reduces the interaction of MoSfl1 with MoCyc8 and MoTup1 . To determine the function of MoTUP1 , we generated the Motup1 deletion mutant in the wild-type strain Guy11 . The resulting deletion mutant MTU35 ( Table 1 ) was significantly reduced in growth rate and it rarely produced aerial hyphae and conidia ( Fig 6C ) . Unlike the cpkA cpk2 mutant , the Motup1 mutant was normal in appressorium formation . One distinct phenotype of the Motup1 mutant was the production of swollen bodies in the subapical regions in hyphae grown on CM ( Fig 6C ) , suggesting cell wall integrity defects . The phenotype differences between the Motup1 mutant and Mosfl1 or cpkA cpk2 mutant indicate that the MoCyc8-MoTup1 co-repressor is involved in regulating different sets of genes by interacting with transcription factors other than MoSfl1 . In S . cerevisiae , Sfl1 has two predicted consensus PKA phosphorylation sites S207 and S733 [43] that are conserved in MoSfl1 ( S211 and S554 ) and its orthologs from other fungi ( Fig 7A ) . PSITE analysis identified T441 as the only other consensus PKA phosphorylation site in MoSfl1 . To determine their functions , the MoSFL1S211D , MoSFL1T441D , and MoSFL1S554D alleles were generated and transformed into the cpkA cpk2 mutant . Whereas MoSFL1S211D transformants grew faster , the MoSFL1T441D and MoSFL1S554D transformants had similar growth defects with the original cpkA cpk2 mutant ( Fig 7B ) . Similar approaches were used to generate transformants of cpkA cpk2 expressing the MoSFL1S211A , MoSFL1T441A , and MoSFL1S554A alleles ( Table 1 ) . None of these S/T to A mutations had suppressive effects on the growth defect of cpkA cpk2 ( Fig 7B ) . These results indicate that phosphorylation of MoSfl1 at S211 may play a critical role to release its inhibitory functions . However , the MoSFL1S211D transformant failed to form appressoria on hydrophobic surfaces ( Fig 7C ) . Therefore , the S211D mutation in MoSFL1 could suppress the growth but not appressorium formation defect of cpkA cpk2 . To identify genes affected by deletion of both CPKA and CPK2 , we conducted RNA-seq analysis with RNA isolated from hyphae collected from 2-day-old CM cultures . Considering the significant reduction in growth rate , it was surprising that only 451 genes were down-regulated in the cpkA cpk2 double mutant in comparison with the wild type ( S3 Table ) . However , many of them are functionally important for growth , including several genes encoding ribosomal proteins ( MGG_00546 , MGG_03372 , MGG_09927 , MGG_01113 , MGG_06571 ) and enzymes important for cell wall synthesis ( MGG_00592 , MGG_03208 , MGG_07331 , MGG_01575 , and MGG_03883 ) . When the promoter regions ( 1000-bp upstream of the start codon ) of genes down-regulated in the double mutant were analyzed , 111 of them contain the putative HSF-binding element AGAA-n-TTCT ( n≤20 ) [27] . Among them , 29 genes have more than one HSF-binding elements in their promoter regions . These results indicate that the putative MoSfl1-binding element is enriched among the genes significantly down-regulated in the double mutant . The cpk1 cpk2 double mutant of F . graminearum also had severe growth defects [20] . Similar to the cpkA cpk2 mutant of M . oryzae , fast-growing sectors were often observed in V8 cultures of cpk1 cpk2 mutant that were older than 10 days . We isolated 30 suppressor strains that had similar growth rate with the wild type ( Fig 8A ) . In infection assays with corn silks , suppressor strains of cpk1 cpk2 were still defective in plant infection ( Fig 8B ) . When the FgSFL1 gene was amplified and sequenced , 29 of them had mutations in the open reading frame ( ORF ) ( Fig 8C; S4 Table ) . Suppressor strain HS29 lacked mutations in the ORF of FgSFL1 although its phenotype was similar to that of other suppressor strains with mutations in FgSFL1 , suggesting possible mutations at its interacting site on FgCYC8 . The most common mutation is the C1717 to T ( Q501 to stop codon ) mutation that resulted in the truncation of C-terminal 91 amino acids . A total of 15 suppressor strains had this C1717T mutation . Therefore , truncation of the C-terminal region also suppressed the cpk1 cpk2 mutant in F . graminearum . These results indicate that the function of SFL1 orthologs in hyphal growth is well conserved in M . oryzae , F . graminearum , and possibly other filamentous ascomycetes .
Like many other filamentous ascomycetes , the rice blast fungus has two genes encoding catalytic subunits of PKA . Whereas the cpkA mutant is defective in appressorium formation and pathogenesis , deletion of CPK2 had no effects on plant infection and appressorium morphogenesis . Interestingly , unlike cpkA , the cpk2 mutant was slightly reduced in growth rate . Therefore , although CpkA plays a more critical role than Cpk2 for pathogenesis , CPK2 may be more important during vegetative growth . In F . graminearum and A . fumigatus , deletion of the CPK2 ortholog had no detectable phenotype [19 , 20] , which differs slightly from cpk2 in M . oryzae . However , the cpkA cpk2 mutant had more severe defects than the single mutants in growth and conidiation , which is similar to F . graminearum and A . fumigatus [19 , 20] . Therefore , the overlapping functions of CpkA and Cpk2 orthologs during vegetative growth and asexual reproduction may be evolutionally conserved in filamentous ascomycetes . Although cAMP signaling is known to be important for appressorium formation in a number of fungal pathogens , including Colletotrichum species [44 , 45] , no mutants deleted of both catalytic subunits have been reported in plant pathogenic fungi except in U . maydis . In U . maydis , the mutant deleted of both catalytic subunits was defective in plant infection but its defects in appressorium formation was not examined [21] . Our results showed that PKA activities are essential for appressorium formation in M . oryzae , which has not been previous reported in plant pathogens . The cAMP-PKA pathway is responsible for surface recognition in M . oryzae . To our surprise , although no tip swelling or appressorium formation was observed , the majority of cpkA cpk2 germ tubes curled or rotated on hydrophobic surfaces . Therefore , germ tubes of the double mutant still responded to surface hydrophobicity although they were blocked in appressorium formation . Sensing of surface hardness and hydrophobicity likely involves mechanosensor proteins , which may trigger polarity disturbance and germ tube curling independent of cAMP signaling . Several putative mechanosensor genes have been shown to be up-regulated during appressorium formation [46] . It is puzzling that most cpkA cpk2 conidia failed to germinate on hydrophilic surfaces , and curling germ tubes were not observed in a few of them that germinated . In filamentous ascomycetes such as A . nidulans and C . trifolii , cAMP signaling is known to regulate conidium germination [44 , 47] . However , the cpkA cpk2 mutant was normal in conidium germination on hydrophobic surfaces . Because surface attachment is a cue for stimulating conidium germination in M . oryzae , one likely explanation is that deletion of both CPKA and CPK2 may affect the attachment of conidia to hydrophilic surfaces . The Pmk1 MAP kinase pathway is essential for appressorium formation in M . oryzae and other plant pathogens [9 , 34] . The cpkA cpk2 mutant had an increased phosphorylation level of Pmk1 but was defective in appressorium formation . It is possible that PKA activity is required to release the suppressive effect of MoSfl1 on genes important for germ tube tip swelling and appressorium formation . Spontaneous suppressors of cpkA cpk2 produced melanized tips in aerial hyphae and appressoria on hydrophilic surfaces , which is similar to transformants expressing the dominant active RAS2 [48] . Therefore , releasing the repressor role of MoSfl1 or MoCyc8-MoTup1 in the cpkA cpk2 mutant in which Pmk1 is over-activated is sufficient to activate appressorium formation under non-conducive conditions . One likely explanation is that some genes required for tip deformation or appressorium formation are only expressed when MoSfl1 is phosphorylated by PKA although the essential role of Pmk1 in appressorium formation involve other downstream targets . Because deletion of MoSFL1 had no effect on appressorium formation [26] , it will be interesting to determine the effects of expressing the dominant active MST7 allele in the Mosfl1 deletion mutant . In M . oryzae , the mac1 mutant is known to produce spontaneous suppressors and some of them had suppressor mutations in the SUM1 gene [30] . In fact , instability of adenylate cyclase mutant and suppressor mutations in regulatory subunit genes are well characterized in S . cerevisiae , U . maydis , and other fungi [30 , 49 , 50] . However , to our knowledge , spontaneous suppressors of PKA mutants have not been reported in other fungi . In S . cerevisiae , the tpk1 tpk2 tpk3 triple mutant is not viable . It will be important to assay whether the mutants deleted of both catalytic subunits also produce spontaneous suppressors in U . maydis and A . fumigatus [19 , 21] . Because the cpk1 cpk2 mutant of F . graminearum [20] was also found to be unstable and had mutations in FgSFL1 in 29 of the 30 suppressor strains sequenced , it is likely that SFL1 orthologs have a conserved role in the repression of genes important for growth and conidiation , at least in filamentous ascomycetes . For the one F . graminearum and two M . oryzae suppressor strains without mutations in the SFL1 ortholog , identification of the suppressor mutations by whole genome sequencing [51] and characterization of the corresponding genes in these mutants in the future will be helpful to better understand the cAMP-PKA pathway in filamentous fungi . In yeast , SFL1 can function as either a transcriptional activator or repressor [23 , 24] . In M . oryzae , deletion of MoSFL1 by itself did not affect vegetative growth but resulted in a reduction in virulence [26] . Phenotype characterization of the Mosfl1 mutant is suitable for characterizing its activator but not repressor functions . In this study , we showed that deletion or truncation of MoSFL1 could suppress the defects of cpkA cpk2 mutant in vegetative growth and appressorium formation but not plant infection and conidiation . Interestingly , the suppressor mutants with mutations in FgSFL1 were also recovered in vegetative growth but not pathogenicity in F . graminearum . Whereas M . oryzae forms melanized appressoria for plant penetration , F . graminearum produces infection cushions and hyphopodia [52 , 53] . However , after plant penetration , both of them form invasive hyphae inside plant cells that are different from vegetative hyphae in hyphal morphology and possibly cell cycle regulation [52 , 54 , 55] . The cAMP-PKA pathway is important for plant infection in M . oryzae [34 , 56] and F . graminearum [20 , 53] , possibly by regulating the growth of invasive hyphae after penetration in these two plant pathogenic fungi with different infection mechanisms . It is possible that MoSFL1 plays an activator role in regulating genes important for invasive growth but negatively regulates genes important for vegetative growth in M . oryzae . However , it is more likely that the cAMP-PKA pathway regulates genes important for plant penetration and invasive growth via other transcription factors . In yeast , Sfl1 inhibits the transcription of its target genes by interacting with the Cyc8-Tup1 co-repressor [27] . However , it is not clear which region of Sfl1 interacts with Cyc8 or Tup1 . Our data clearly showed that the C-terminal 93 amino acids of MoSfl1 is essential for its interaction with Cyc8 ( Fig 9 ) . This C-terminal region of MoSfl1 is well conserved in its orthologs from filamentous ascomycetes . In F . graminearum , 15 of the 29 suppressor strains of cpk1 cpk2 mutant had the nonsense mutation at Q501 ( S4 Table ) in FgSFL1 resulting in the truncation of its C-terminal region . Other 12 suppressor strains had either nonsense or frameshift mutations upstream from Q501 . These results indicate that the C-terminal region of Sfl1 orthologs likely plays a conserved role in regulating the expression of genes important for hyphal growth via its association with the Cyc8-Tup1 co-repressor complex . The difference between MoSfl1 and yeast Sfl1 in the C-terminal region may be directly related to the importance of PKA activities in hyphal growth in M . oryzae , F . graminearum , and possibly other filamentous fungi . Besides the Cyc1-Tup1 co-repressor , Sfl1 also interacts with the mediator proteins Ssn2 , Ssn8 , Sin4 , and Rox3 in S . cerevisiae [23] . Although their orthologs are conserved in M . oryzae , none of them was identified by affinity purification . One possibility is that phosphorylation by PKA is necessary for MoSfl1 to interact with these mediator components but the cpkA cpk2 mutant was used to identify proteins that differentially interacted with MoSfl1 and MoSfl1ΔCT and responsible the suppression of PKA deficiency . Nevertheless , it is also possible that their interactions with MoSfl1 is mediated by the Cyc8-Tup1 complex , which may be too dynamic or transient in M . oryzae . Among the MoSfl1-interacting proteins identified by affinity purification , MGG_01588 and MGG_06958 are orthologous to BMH1 and SSA1 , respectively that interact with Sfl1 in S . cerevisiae . However , several MoSfl1-interacting proteins in M . oryzae ( Table 5 ) such as the putative pathway-specific nitrogen regulator MGG_03286 are unique to filamentous fungi . It will be important to determine their functions in hyphal growth and pathogenesis in M . oryzae . In S . cerevisiae , Sfl1 is a substrate of Tpk2 [25 , 57] . However , it is not clear whether S207D mutation ( = S211 of MoSfl1 ) is sufficient to suppress the growth defect of inviable tpk1 tpk2 tpk3 PKA-deficient mutant [58] . In M . oryzae , expression of the MoSFL1S211D allele in the cpkA cpk2 mutant rescued its growth but not appressorium formation defects . It is possible that phosphorylation of MoSfl1 by PKA disrupts the interaction of MoSfl1 with the Cyc8-Tup1 co-repressor , which in turn activates the expression of genes important for hyphal growth ( Fig 9 ) . Residue S211 of MoSfl1 is well conserved in its orthologs from other filamentous ascomycetes , suggesting that its phosphorylation by PKA likely has a conserved role in the regulation of hyphal growth by the cAMP-PKA pathway . Considering the fact that CpkA and Cpk2 are highly similar in the kinase domain and they have overlapping functions in hyphal growth , it is possible that both of them can phosphorylate MoSfl1 at S211 . Nevertheless , MoSfl1 may be phosphorylated by CpkA and Cpk2 at different amino acid residues . Therefore , it will be important to characterize the phosphorylation sites of MoSfl1 in the cpkA and/or cpk2 deletion mutants .
All the M . oryzae strains used in this study ( Table 1 ) were cultured on oatmeal agar ( OTA ) or complete medium ( CM ) plates at 25°C and stored on desiccated Whatman #1 filter paper at -20°C [59] . Protoplast preparation and PEG-mediated transformation were performed as described [60] . Transformants were selected with 250 μg/ml hygromycin B ( CalBiochem ) , 250 μg/ml geneticin G418 ( Sigma ) , or 200 μg/ml zeocin ( Invitrogen ) in the top agar . Growth rate and conidiation were assayed with OTA cultures as described [60 , 61] . To generate the CPK2 gene replacement construct by double-joint PCR , its 1 . 2-kb upstream and downstream flanking sequences of CPK2 were amplified with primer pairs 1F/ 2R and 3F/4R ( S5 Table ) , respectively ( S1 Fig ) . The hph cassette was amplified with primers Hyg/F and Hyg/R from pCX63 [62] . The resulting products of double-joint PCR were transformed into protoplasts of the wild-type strain Guy11 . Putative cpk2 mutants were screened by PCR with primers 5F and 6R and further confirmed by Southern blot analyses with its downstream flanking sequence as the probe . Vegetative hyphae harvested from two-day-old CM cultures were used for DNA and protein isolation as described [63] . The same strategy was used to generate the CPKA gene replacement construct . The 1 . 2-kb upstream and downstream flanking sequences of CPKA were amplified with primer pairs A1F/A2R and A3F/A4R , respectively ( S1 Fig ) . The G418 cassette was amplified with primers G418/F and G418/R from pFl7 . The products of double-joint PCR were transformed into protoplasts of the cpk2 mutant to generate the cpkA cpk2 mutant . Conidia were harvested from 10-day-old OTA cultures and resuspended to 5×104 conidia/ml in sterile water . For appressorium formation assays , 50 μl droplets of conidium suspensions were placed on glass cover slips ( Fisher Scientific ) or GelBond membranes ( Cambrex ) and incubated at 25°C for 24 h as described [13 , 64] . To assay its stimulatory effects , cAMP was added to the final concentration of 10 μM to conidium suspensions [65] . For infection assays , conidia were resuspended to 5×105 conidia/ml in 0 . 25% gelatin . Two-week-old seedlings of rice cultivar CO-39 were used for spray or injection infection assays as described [66 , 67] . Lesion formation was examined 7 days post inoculation ( dpi ) . Vegetative hyphae were harvested from 2-day-old CM cultures and used for protein extraction as described [68] . Total proteins ( approximately 20 mg ) were separated on a 12 . 5% SDS-PAGE gel and transferred to nitrocellulose membranes for western blot analysis [62] . Expression and phosphorylation of Pmk1 and Mps1 were detected with the PhophoPlus p44/42 MAP kinase antibody kits ( Cell Signaling Technology ) following the manufacturer’s instructions . Intracellular cAMP was extracted from vegetative hyphae harvested from two-day-old CM cultures as described [69] and detected with the cAMP enzyme immunoassay ( EIA ) system ( Amersham Pharmacia Biotech ) following the manufacturer’s instructions . The yeast gap repair approach was used to generate the S-tag and 3×FLAG fusion constructs [68] . To generate the SUM1- , TUP1- , and CYC8-S-tag constructs , each gene was amplified and cloned into vector pXY203 [66 , 70] . MoSFL1 was cloned into vector pFL6 to generate the 3×FLAG-MoSFL1 construct . CPKA and CPK2 were cloned into vector pFL7 [70] to generate the CPKA-3xFLAG and CPK2-3×FLAG fusion constructs . All of the resulting S-tag and 3×FLAG fusion constructs were confirmed by sequencing analysis and transformed into Guy11 or the cpkA cpk2 mutant in pairs . Total proteins were isolated from the resulting transformants and incubated with the anti-S-Tag Antibody Agarose beads ( Bethyl Laboratories ) . Proteins bound to anti-S-tag agarose were eluted and used for western blot analysis [13 , 31] . The presence of related fusion proteins was detected with the anti-FLAG ( Sigma-Aldrich ) or anti-S ( Abcam ) antibody as described [31] . Fast-growing sectors of the cpkA cpk2 mutant were transferred with sterile toothpicks to fresh oatmeal agar plates . After single-spore isolation , each subculture of spontaneous suppressors was assayed for defects in growth , conidiation , and plant infection [71 , 72] . To identify suppressor mutations , all the candidate downstream target genes of PKA were amplified with primers listed in S5 Table and sequenced . Mutation sites were identified by sequence alignment with sequences of target genes in the reference genome [1] and their PCR products . To generate the cpkA cpk2 Mosfl1 mutant , the upstream and downstream flanking sequences of MoSFL1 were amplified with primer pairs Sfl1ko1F/Sfl1ko2R and Sfl1ko3F/ Sfl1ko4R ( S5 Table ) , respectively , and fused with the ble cassette amplified from pFL6 [73] by double-joint PCR [74] . To generate the cpkA cpk2 MoSFL1ΔCT mutant , the flanking sequences of MoSFL1 were amplified with primer pairs CCSko1F/CCSko2R ( S5 Table ) and Sfl1ko3F/ Sfl1ko4R ( S5 Table ) . The resulting MoSFL1 and MoSFL1CT gene replacement PCR products were transformed into protoplasts of the cpkA cpk2 mutant . Putative Mosfl1 or MoSFL1CT mutants were screened by PCR analysis and verified for the deletion of MoSFL1 or its C-terminal region ( 496–588 aa ) . The full length MoSFL1 and MoSFL1ΔCT fragments were amplified with primers listed in S5 Table and cloned into vector pFL6 by yeast gap repair [70] . The 3×FLAG-MoSFL1 and 3×FLAG-MoSFL1ΔCT fusion constructs were rescued from the resulting yeast transformants and transformed into the cpkA cpk2 mutant . Hyphae of the 3×FLAG-MoSFL1 and 3×FLAG-MoSFL1ΔCT transformants were homogenized with a glass beater at 4°C for protein extraction [31 , 68] . Proteins eluted from anti-FLAG resins ( Sigma-Aldrich ) were digested with trypsin and the resulting tryptic peptides were analyzed with nanoflow liquid chromatography tandem mass spectrometry ( MS ) as described [31 , 75–77] . Proteins were identified by searching MS data against NCBI non-redundant F . graminearum protein database with the SEQUEST™ algorithm [78] . At least three independent biological replicates were analyzed to identify proteins that interact with MoSfl1WT and MoSfl1ΔCT . To generated the MoSFL1S211D allele , PCR fragments amplified with primer pairs MoSfl1-FL5F/MoSfl1-S211D1R and MoSfl1-S211D2F/MoSfl1-FL5R ( S5 Table ) were connected by overlapping PCR [79] and cloned into vector pFL5 [70] by yeast gap repair . The MoSFL1S211D construct was rescued from Trp+ yeast transformants and verified for the S211D mutation by sequencing analysis . Similar approaches were used to generate the MoSFL1S211A , MoSFL1T441D , MoSFL1T441A , MoSFL1S554D , and MoSFL1S554A constructs . All the MoSFL1 mutation alleles were transformed into the cpkA cpk2 deletion mutant . The resulting transformants were characterized for defects in growth , conidiation , appressorium formation , and plant infection as described [71] . | The cAMP-PKA signaling pathway plays a critical role in regulating various cellular processes in eukaryotic cells in response to extracellular cues . In the rice blast fungus , this important pathway is involved in surface recognition , appressorium morphogenesis , and infection . However , the exact role of PKA is not clear due to the functional redundancy of two PKA catalytic subunits CPKA and CPK2 . To further characterize their functions in growth and pathogenesis , in this study we generated and characterized the cpkA cpk2 double mutant and its suppressor strains . Unlike the single mutants , cpkA cpk2 mutant had severe defects in growth and conidiation and was defective in appressorium formation and plant infection . Interestingly , the double mutant was unstable and produced fast-growing suppressors . In two suppressor strains , mutations were identified in a transcription factor gene orthologous to SFL1 , a downstream target of PKA in yeast . Deletion of the entire or C-terminal 93 residues of MoSFL1 could suppress the growth defect of cpkA cpk2 . Furthermore , the terminal region of MoSfl1 was found to be essential for its interaction with the MoCyc8 co-repressor , which may be negatively regulated by PKA . Therefore , loss-of-function mutations in MoSFL1 can bypass PKA activity to suppress the growth defect of cpkA cpk2 . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] | [
"phosphorylation",
"gene",
"regulation",
"fungal",
"structure",
"fungi",
"plant",
"science",
"model",
"organisms",
"experimental",
"organism",
"systems",
"gene",
"types",
"appressoria",
"plant",
"pathology",
"molecular",
"biology",
"techniques",
"mutagenesis",
"and",
"gene",
"deletion",
"techniques",
"research",
"and",
"analysis",
"methods",
"saccharomyces",
"mycology",
"artificial",
"gene",
"amplification",
"and",
"extension",
"proteins",
"gene",
"expression",
"molecular",
"biology",
"yeast",
"biochemistry",
"rice",
"blast",
"fungus",
"post-translational",
"modification",
"plant",
"fungal",
"pathogens",
"polymerase",
"chain",
"reaction",
"plant",
"pathogens",
"genetics",
"biology",
"and",
"life",
"sciences",
"yeast",
"and",
"fungal",
"models",
"saccharomyces",
"cerevisiae",
"suppressor",
"genes",
"organisms",
"deletion",
"mutagenesis"
] | 2017 | PKA activity is essential for relieving the suppression of hyphal growth and appressorium formation by MoSfl1 in Magnaporthe oryzae |
Biological function of proteins is frequently associated with the formation of complexes with small-molecule ligands . Experimental structure determination of such complexes at atomic resolution , however , can be time-consuming and costly . Computational methods for structure prediction of protein/ligand complexes , particularly docking , are as yet restricted by their limited consideration of receptor flexibility , rendering them not applicable for predicting protein/ligand complexes if large conformational changes of the receptor upon ligand binding are involved . Accurate receptor models in the ligand-bound state ( holo structures ) , however , are a prerequisite for successful structure-based drug design . Hence , if only an unbound ( apo ) structure is available distinct from the ligand-bound conformation , structure-based drug design is severely limited . We present a method to predict the structure of protein/ligand complexes based solely on the apo structure , the ligand and the radius of gyration of the holo structure . The method is applied to ten cases in which proteins undergo structural rearrangements of up to 7 . 1 Å backbone RMSD upon ligand binding . In all cases , receptor models within 1 . 6 Å backbone RMSD to the target were predicted and close-to-native ligand binding poses were obtained for 8 of 10 cases in the top-ranked complex models . A protocol is presented that is expected to enable structure modeling of protein/ligand complexes and structure-based drug design for cases where crystal structures of ligand-bound conformations are not available .
Interactions between proteins and small molecules are involved in many biochemical phenomena . Insight into these processes relies on detailed knowledge about the structure of protein/ligand complexes , e . g . how enzymes stabilize substrates and cofactors in close proximity . Moreover , almost all drugs are small-molecule ligands that interact with enzymes , receptors or channels . Accordingly , ligand-bound receptor complex structures are a critical prerequisite for understanding biological function and for structure based drug design . However , structure determination of protein/ligand-complexes can be difficult , time-consuming and expensive . Crystal structures of protein/ligand complexes are usually obtained either by co-crystallization or soaking and it is a common problem that even when conditions for crystallizing the apo-protein are well established these might not be transferable to the protein/ligand complex [1]–[4] . Particularly , conformational transitions of the receptor associated with ligand binding pose a severe challenge to the structure elucidation of holo complexes [5]–[8] . When structures of ligand-bound protein conformations are not available , structure-based drug design becomes highly challenging . Several studies showed that virtual screening to an apo-structure usually results in a poor enrichment factor ( the ability to discriminate between binders and non-binders ) compared to the holo-structure even when the structural difference between both is comparably small [9]–[11] . Therefore , the development of docking programs aims at allowing a certain degree of receptor flexibility either by using an ensemble of structures instead of a single receptor conformation [12]–[15] or by explicitely modeling flexibility such as sidechain variations ( Autodock4 [16] , [17] , Gold [18] , [19] , FlexX [20] , RosettaLigand [21] ) , predefined flexibility of certain parts of the structure ( FlipDock [22] ) and also small variations of the backbone ( Glide/Prime [23] , RosettaLigand [24] , ICM [25] , [26] ) . Incorporating receptor flexibility in molecular docking is a substantial progress and has been shown to enhance both enrichment factors and the ability to predict correct binding poses , particularly in cases when docking a compound to a receptor structure that has been crystallized with a different ligand ( cross-docking ) which is usually the case when searching for novel drugs . However , the degree of flexibility thus far is limited to either sidechain motions or small variations of the backbone and thus , the availability of a holo-structure or an apo-structure that is highly similar to the holo conformation is currently a prerequisite for a successful docking , severely limiting structure-based drug design . Particularly , receptors that undergo a substantial conformational transition upon ligand binding are currently precluded from structure based drug design . Although protein-ligand crystals suitable for diffraction might not be accessible , several experimental techniques exist to detect conformational changes . In many cases where proteins undergo domain reorientations upon ligand binding they adopt a different shape in the ligand bound state , corresponding to a change in the radius of gyration that can be studied either by NMR , where a more compact shape causes a descrease in the rotational correlation time [27]–[29] or by small-angle scattering of x-rays ( SAXS ) or neutrons ( SANS ) [30]–[33] . These shape descriptions provide invaluable information for modeling of structures [34]–[36] and macromolecular assemblies [37] , [38] as well as insight into protein dynamics . Here , we present a method to predict the structure of protein/ligand complexes for proteins that undergo a large conformational change upon ligand binding . The protocol solely requires the apo-structure , a known ligand and experimental data on the shape of the holo-structure . Here , we apply the radius of gyration as shape information , a quantity that can frequently be readily assessed more easily than an x-ray structure . We developed a simulation protocol that combines biased conformational sampling , docking and molecular dynamics simulations and applied it to ten ligand-binding proteins ( see Table 1 ) . We chose cases where both , the unbound conformation and the bound conformation are known from x-ray crystallography in order to be able to a-posteriori validate the predicted receptor conformations and docking poses . The conformational changes involved range from 2 . 1 to 7 . 1 Å backbone RMSD ( see table 1 and fig . 1 ) and the binding site geometries differ substantially between the apo and the holo conformations . In nine of the ten cases we predict holo receptor conformations close to the native ligand-bound conformation and in eight cases we predict ligand binding poses close to the native state , rendering our method suitable for blind predictions of protein/ligand complexes involving large conformational transitions . | Structure-based drug design has become a powerful tool in modern drug discovery pipelines . A critical prerequisite is a structure of the target protein close to its ligand bound conformation which is often difficult to determine experimentally . In many cases , a structure of the unbound receptor is available , but conformational changes with respect to the ligand-bound form preclude it from being used as a basis for structure-based drug design . We have developed a computational approach to predict the structure of protein/ligand complexes based solely on the unbound conformation , the ligand , and easy-to-assess experimental data . We tested our protocol on proteins that undergo substantial structural rearrangements upon binding a ligand and were able to predict structures of protein/ligand complexes which are in good agreement with experimentally determined structures . The ability to predict ligand bound receptor conformations based on structures in the unbound state enables structure-based drug design for cases where crystallization of the complex has not been successful so far . | [
"Abstract",
"Introduction"
] | [
"biophysics/theory",
"and",
"simulation",
"biophysics/biomacromolecule-ligand",
"interactions",
"computational",
"biology/protein",
"structure",
"prediction",
"computational",
"biology/molecular",
"dynamics",
"biochemistry/drug",
"discovery"
] | 2010 | Conformational Transitions upon Ligand Binding: Holo-Structure Prediction from Apo Conformations |
Group A streptococcus ( GAS; Streptococcus pyogenes ) is a common pathogen that invades non-phagocytic human cells via endocytosis . Once taken up by cells , it escapes from the endocytic pathway to the cytoplasm , but here it is contained within a membrane-bound structure termed GAS-containing autophagosome-like vacuoles ( GcAVs ) . The autophagosome marker GFP-LC3 associates with GcAVs , and other components of the autophagosomal pathway are involved in GcAV formation . However , the mechanistic relationship between GcAV and canonical autophagy is largely unknown . Here , we morphologically analyzed GcAV formation in detail . Initially , a small , GFP-LC3-positive GcAV sequesters each streptococcal chain , and these then coalesce into a single , large GcAV . Expression of a dominant-negative form of Rab7 or RNAi-mediated knockdown of Rab7 prevented the initial formation of small GcAV structures . Our results demonstrate that mechanisms of GcAV formation includes not only the common machinery of autophagy , but also Rab7 as an additional component , which is dispensable in canonical autophagosome formation .
Streptococcus pyogenes , also known as Group A Streptococcus ( GAS ) is a common pathogen that causes a variety of acute infections including pharyngitis , skin infections , acute rheumatic fever and life-threatening necrotizing fasciitis [1] . The bacterium enters non-phagocytic human cells via endocytosis and subsequently escapes the endolysosomal pathway by inserting bacteria-derived streptolysin O , a cholesterol-dependent pore-forming cytolysin , into the host endosomal membrane [2] . Following bacterial escape into the cytoplasm , GAS is engulfed by a unique structure named GAS-containing autophagosome-like vacuoles ( GcAVs ) [2] . GcAVs acquire lysosomal enzymes leading to GAS degradation . When GAS release from cells is blocked with tannic acid treatment , the number of organisms recovered from wild-type ( WT ) cells is reduced by 80% , but there is no reduction in the number of bacteria isolated from autophagy-deficient cells . Macroautophagy , hereafter referred to simply as autophagy , is a highly conserved cellular process induced by nutrient-starvation that transfers some cytoplasmic components to lysosomes for degradation and recycling of constituent macromolecules [3] . Autophagy involves the formation of a specialized membrane structure , the autophagosome , which is a spherical structure enclosed by two lipid bilayers [4] . The GcAV is considered an “autophagosome-like” structure because of several characteristics shared with autophagosomes . Both structures are labeled with the auophagosomal marker GFP-LC3 [2] . LC3 is the mammalian homologue of yeast Atg8 , one of over 30 autophagy-related Atg proteins identified in yeast [5] . Its carboxy terminus is conjugated to the lipid phosphatidylethanolamine by a ubiquitination-like reaction , and this modification leads to LC3 localization on the autophagosomal membrane [6] . Additionally , the formation of both structures depends on Atg5 , another protein essential for autophagy [7] . In Atg5 knockout cells , few GcAV are formed [2] . Recently , we showed that the protein complex containing Atg5 , Atg12 and Atg16L acted as an E3-like enzyme during LC3 lipidation [8] . We hypothesize that both GcAV formation and autophagy require this protein complex . Despite these similarities , there are several important differences between autophagosomes and GcAVs . The diameter of autophagosomes is 0 . 5 to 1 . 0 µm , but GcAV measure nearly 10 µm across [2] . Furthermore , majority of canonical autophagy is currently thought to non-selectively take up cytosolic components , but GcAV formation is highly specific for GAS sequestration . Thus , although they share some common mechanisms of formation , these two processes are distinct physiologic phenomena . The autophagic machinery has recently been implicated to play a role in several host-pathogen interactions and host defense which included Toxoplasma gondii , Listeria , Mycobacterium tuberculosis , and Shigella [9] , [10] , [11] , [12] , [13] . Mycobacterium tuberculosis has the ability to survive within the phagosomal compartment by interfering with phagosome maturation in Mycobacterium-containing vacuoles ( MCVs ) , but the autophagic pathway can deliver MCVs to the lysosomal degradative pathway for eventual pathogen elimination [11] . Additionally , ligation of a Toll-like receptor ( TLR ) at the cell surface leads to its internalization and subsequent recruitment of the autophagic machinery to TLR-containing phagosome [14] . However , the dynamic processes regulating vesicular trafficking and membrane fusion during GcAV formation are largely undefined . In this paper , we morphologically characterize GcAV formation and identify several mechanistic events that occur during GcAV biogenesis . In particular , we show that Rab7 is required for the early phase of GcAV formation . Based on these findings , we develop a membrane dynamic model of GcAV formation and its relation to canonical autophagy .
GAS internalized into HeLa cells eventually come to be contained within membrane delimited intracellular structures called GcAVs [2] . We wished to morphologically characterize GcAV formation in greater detail . HeLa cells expressing GFP-LC3 were infected with GAS , fixed 0 . 5 h after infection , and examined by fluorescence microscopy . A number of different GcAV morphologies were seen that we subcatergorized into four types as follows ( Figure 1A–D ) : type A , a linear chain-like structure; type B , aggregates of GAS-containing vacuoles with a clear GFP-LC3 boundary between each cocci; type C , a larger aggregate of GAS-containing structures with a partially discontinuous inner boundary; and type D , a semi-spherical structure lacking a clear inner boundary . The frequencies of these different structures changed as the time of infection increased . At 0 h , type A was the most frequently observed , but the number of type A structures decreased with time . Conversely , type D structures increased with time from less than 10% of GAS containing structures to nearly 50% . The observed frequencies of structure types B and C were intermediate , suggesting that these morphologic forms are developmental intermediates ( Figure 1E ) . To confirm this hypothesis , we performed time-lapse imaging of the membrane dynamics in living cells . Several linear chain-like GcAVs ( type A ) assembled into a collection of grape-like GcAVs ( types B ) , and the boundary between the GAS chains then gradually disappeared ( Figure 2A ) . Indeed , as seen in Video S1 , two GAS-containing structures fused to form a larger structure with the appearance of type C structures . We next performed immunoelectron microscopy to analyze the nature of structure type D in greater detail ( Figure 3 ) . HeLa cells expressing green fluorescent protein ( GFP ) -LC3 were infected with GAS and after a 2-h infection , cells were fixed and subjected to immuno electron microscopy ( EM ) analysis using anti-GFP . GAS was surrounded by single membrane labeled with anti-GFP . Furthermore , these images show that the membrane encompassing the smaller GcAVs is lost upon formation of the larger structures . Taken together , these results indicate that GcAV formation is mediated by the fusion of GAS-containing structures into large , membrane-enclosed compartment . Our results suggest that homotypic membrane fusion may be involved in GcAV enlargement . We hypothesized that Rab7 , a member of the small GTPase Rab family , may be involved in GcAV formation because Ypt7 , the yeast orthologue of Rab7 , catalyzes the homotypic fusion of vacuoles/lysosomes [15] . To investigate the role of Rab7 in GcAV formation , we used a Rab7 mutant that is constitutively GDP bound and acts as a dominant-negative during the later stages of endocytosis in mammalian cells [16] . HeLa cells expressing GFP-LC3 were transiently transfected with either the dominant-negative Rab7 ( T22N ) tagged with monomeric red fluorescent protein ( mRFP ) or mRFP alone as a control , cultured overnight , and then infected with GAS . The cells were fixed and examined by fluorescence microscopy at different time points after infection ( Figure 4A–C ) . Consistent with several previous reports [17] , [18] , [19] , canonical autophagosomes accumulated in the cytoplasm of Rab7 ( T22N ) -expressing cells ( Figure 4B and C , arrowheads ) , reflecting the blockage of fusion between the lysosome and autophagosomes possibly formed by constitutive autophagy . We expected decreased formation of large GcAV and an accumulation of small GAS-containing structures in Rab7 ( T22N ) -expressing cells , but there was quite little accumulation of GFP-LC3 signal around GAS , indicating that LC3 recruitment to GcAV or GcAV formation is defected ( Figure 4D ) . GcAv formation was not substantially affected in cells overexpressing WT Rab7 or a GTPase deficient mutant ( Q67L ) ( data not shown ) . To confirm these observations , we extended our analysis with electron microscopy . As before , there was significant accumulation of canonical autophagosomes within the cytoplasm of Rab7 ( T22N ) -expressing cells after infection with GAS ( Figure 4E , arrowheads ) . However , no GAS were enclosed in GcAV-like membrane delimited structures ( Figure 4E , arrows ) characterized by canonical autophagosome or isolation membrane like double membrane ( Figure 4E , arrowheads and small arrows ) , indicating that GcAVs were not formed . Thus , we conclude that Rab7 inactivation disrupts an early step in GcAV formation . To further define the role of Rab7 in GcAV formation , we examined the effects of Rab7 inactivation on the killing of GAS within GcAV . HeLa cells were transduced with adenovirus encoding Rab7 ( T22N ) , and these cells were then infected with GAS and treated with tannic acid to prevent GAS escape to the cytoplasm [2] . The number of surviving bacteria was counted using a colony formation assay . Expression of Rab7 ( T22N ) decreased the efficiency of endocytic internalization of GAS by 35 . 9±2 . 1% ( average ± SD of four independent experiments ) compared to control cells , possibly reflecting negative feedback within the endocytic pathway . Despite the decreased uptake , however , greater numbers of GAS were recovered from Rab7 ( T22N ) -expressing cells at all time points ( Figure 5A ) . To confirm this finding , we used RNAi to knockdown Rab7 expression in HeLa cells , and virtually identical results were obtained ( Figure 5B ) . Taken together , these results clearly establish a direct requirement for Rab7 in GcAV formation and GAS killing . To better understand the mechanism ( s ) regulating GcAV formation , we next wished to identify the GcAV precursor structure . We used correlative fluorescence and electron microscopy to examine HeLa cells expressing GFP-LC3 . After GAS infection and fixation , a GFP-LC3 positive structure , morphologically distinct from types A , B , C , or D described above , was identified incompletely surrounding several bacteria by fluorescence microscopy . By electron microscopy , this corresponded to a sac-like structure aligned along the GAS bacteria at the sites of GFP-LC3 signal ( Figure 6A ) . To better understand the spatial relationship of this sac-like structure , we performed serial sectioning and electron microscopy . In Figure 6B and C , two membrane sacs appear associated with the same bacterium ( white arrowheads ) . Moreover , successive electron microscopic images revealed that a double-membrane bound structure encompassing a GAS bacterium in one area ( Figure 6B , white arrows ) was connected to another such structure in the next section ( Figure 6C , red arrowheads ) . Atg5 is transiently associated with isolation membranes , the intermediate form of autophagosome [7] , and we next determined the localization of Atg5 during GcAV formation . When HeLa cells expressing GFP-Atg5 were infected with GAS , GFP-Atg5 was seen around GAS bacteria ( Figure 7A ) . Video microscopy of living cells demonstrated the transient appearance of multiple punctate spots of GFP-Atg5 near individual bacteria ( Figure 7B control and Video S2 ) . These spots were highly mobile , suggesting that they correspond to individual membrane-bound structures , but not specialized domain in continuous membrane . These results support the model that multiple , smaller membrane-bound structures surround a GAS chain and coalesce to form GcAVs . In cells expressing Rab7 ( T22N ) , GFP-Atg5 positive structures were also observed adjacent to GAS ( Figure 7B Rab7 ( T22N ) , Video S3 ) , and , when the number of contacts between GAS and GFP-Atg5 was quantified , 26 ( RFP ) and 27 cells ( T22N ) contained such sites of contacts ( Table 1 ) . We next examined the localization of endogenous Rab7 . NIH3T3 cells expressing GFP-LC3 were infected with GAS , and the cells were fixed and processed for immunofluorescent confocal microscopy using anti-Rab7 antibodies . In these cells ( Figure 8A–F ) , Rab7 localized to GcAVs labeled by GFP-LC3 . Rab7 localizes to late endosomes and lysosomes [16] , and , because GcAVs terminally fuse with lysosomes [2] , the colocalization of Rab7 with GFP-LC3 positive GcAVs is not entirely unexpected . However , as shown in Figure 8A–F , there is a sub-population of Rab7 that localized to GcAVs that were not well stained by the lysosomal membrane marker , Lamp-1 . Observation of live cells using fluorescently labeled Rab7 ( mRFP-Rab7 ) and GFP-Atg5 demonstrated colocalization of mRFP-Rab7 and GFP-Atg5 at DAPI-stained GAS ( Figure 8G ) . Thus , a population of Rab7 is recruited to GFP-Atg5 positive membranes during the early phase of GcAV formation . These results indicate that Rab7 plays an important role in the early phase of GcAV formation .
In this study , we mechanistically characterized GcAV formation and demonstrated that this process is biphasic . Each bacterium is enclosed by a membrane and these structures fuse into a membrane-bound structure to form a large spherical GcAV . Rab7 is involved in the early phase of GcAV formation , though it may play a role in later stages as well . The final phase of GcAV formation is the appearance of a semi-spherical large GcAV enclosed by a single membrane . However , because each bacterium is enclosed in a membrane , the generation of a border composed of a single membrane requires the loss of the internal membranes surrounding each GAS . An identical phenomenon is observed during canonical autophagy when the double-membrane lined autophagosome is converted into the single membrane enclosed autolysosome . Additionally , as in canonical autophagy , the inner membrane of the GcAV becomes degraded following the acquisition of lysosomal contents including lipases . Indeed , we previously showed that GcAVs fused with lysosomes during the late stages of their formation [2] . Thus , GcAV formation is a highly complex process involving dynamic membrane rearrangements . It is likely that the physiochemical properties including associated proteins govern the size and structure of the terminal GcAV , but the precise factors governing this structure are unknown . We identified the precursor structure of the GcAV , and it is a membrane-bound organelle very similar to the isolation membrane of canonical autophagy . Both structures are labeled with LC3 and Atg5 , and , although the exact composition of the isolation membrane remains unclear , it is possible that these two structures are identical or , at the very least , share a number of common lipid and protein components . One key difference between GcAV formation and canonical autophagy , however , is the specificity of the process , and it is likely that GAS recognition factors are incorporated into the nascent GcAV . Additionally , multiple membrane bound structures are recruited to a GAS chain , and they coalesce to form a larger structure encompassing all of the bacteria . The recruitment and fusion of these vesicles into a larger GcAV may proceed without specific adaptor molecules , or , conversely , another organelle ( e . g . lysosomes ) may mediate the fusion of two or vesicles simultaneously . In either case , the membrane fusion event is similar to that leading to autophagosome formation . In autophagy , isolation membranes undergo two distinct fusion events: the closing of the isolation membrane during autophagosome formation , and the fusion between autophagosome and lysosome . Therefore , both of our alternative hypotheses are possible and consistent with events known to occur during autophagy . Indeed , we occasionally observed two morphologically distinct membrane structures connected in serial EM sections , consistent with ongoing fusion events . Interestingly , inactivation of Rab7 impaired GcAV formation and as a result , intracellular survival of GAS increased . The impact of Rab7 inactivation to the GAS killing is milder than that of the Atg5 knock out , but it may be due to impairment of other cellular process because Rab7 is also involved in broad range of physiological responses through the regulation of endocytosis . Or in the presence of intact canonical autophagosome dependent protein sequestration , some unknown bacterial killing system other than autophagy might be induced to compensate for GcAV function . Members of the Rab family including Rab7 are involved in membrane fusion events throughout the secretory and endocytic pathways . During GcAV formation , however , the pre-fusion intermediate structures did not accumulate in cells expressing the Rab7 dominant-negative mutant . Thus , Rab7 may be necessary for the generation of the precursor membrane structure . However , we observed comparable amounts of GFP-Atg5 associated with GAS in cells expressing the Rab7 dominant-negative mutant , and this event presumably occurs later in GcAV formation . In canonical autophagy , Rab7 is not required for autophagosome formation , but it is essential for the fusion of lysosomes and autophagosomes [17] , [18] . Consequently , if GcAV formation largely recapitulates autophagy , the role of Rab7 in the early stages of GcAV formation may be more complicated than simply mediating precursor formation . Alternatively , Rab7 may be involved in the stabilization of precursor vesicles and/or its delivery to GcAV immediately prior to their association with GAS . The presence of Rab7 on GAS-containing structures during the early phases of GcAV formation is consistent with this possibility . We should also stress the possibility that Rab7 is also involved in the later steps of these processes , such as fusion of isolation membranes , homotypic fusion of small GcAV , fusion between GcAV and lysosome , although our experimental system was unable to directly demonstrate the defects in these subsequent events . Further analyses of the composition and formation of GcAV and all precursor structures will help refine these models . In this report , we have further characterized the morphologic features and membrane fusion events associated with GcAV formation . We demonstrated that GcAV formation , while similar to canonical autophagy , has several unique features . The GcAV pathway likely represents a Rab7-dependent evolutionary variation on canonical authophagy to sequester and degrade microorganisms that have evolved strategies to escape from lysosomal degradation . Further studies will identify the mechanism of not only GcAV formation but also of canonical autophagy .
GAS ( S . pyogenes ) strain JRS4 ( M6+ F1+ ) was grown in Todd-Hewitt broth ( BBL , Cockeysville , MD , USA ) supplemented with 0 . 2% yeast extract ( THY ) as described previously [20] . HeLa cells , HeLa cells stably expressing GFP-LC3 [5] or GFP-Atg5 [7] , NIH3T3 cells and 293A ( Invitrogen ) were maintained in DMEM [Dulbecco's modified Eagle's medium ( DMEM , Sigma-Aldrich , MO , USA ) containing 9% fetal bovine serum ( FBS , Invitrogen , CA , USA ) and 2 mM L-glutamine ( Invitrogen ) ] in a 5% CO2 incubator at 37°C . Transfection was done with LipofectAmine2000 reagent ( Invitrogen ) according to the manufacturer's instructions . To obtain stable transformants , cells were selected in the presence of 0 . 5 mg/ml G418 . For knock-down of Rab7 , RNAi targeting human Rab7 ( CGGTTCCAGTCTCTCGGTGTT; corresponding to human Rab7 cDNA of 205–225 bp ) and siPerfect Negative control RNAi ( Sigma Genosys ) were transfected using jetPEI ( Polyplus-transfeciton , Inc ) . For transient expression of the Rab7 ( T22N ) mutant by adenovirus-mediated transfer , ViraPower adenoviral Gateway expression kit ( Invitrogen ) was used according to the manufacturer's instructions . For confocal microscopy analysis , cells ( 8×104 cells ) were seeded in 500 µl of medium in 24-well tissue culture plates containing 12-mm-diameter glass coverslips . After cells were grown overnight , the spent medium was removed and replaced with fresh medium . Bacteria were harvested and washed twice with DMEM , then were added to cell cultures at a multiplicity of infection ( MOI ) of 100 ( cells:bacteria = 1∶100 ) , without antibiotics . After incubation with GAS for 1 h , cells were washed twice with DMEM to remove nonadherent bacteria , and cells were further incubated for the indicated times in DMEM/FBS containing antibiotics [100 µg/ml of gentamicin ( Sigma ) and 100 µg/ml of streptomycin ( Invitrogen ) ] to kill extracellular bacteria . For live cell imaging , GAS bacteria were stained with 0 . 4×10−2 ng/ml DAPI ( 4′ , 6-Diamidino-2-phenylindole dihydrochloride , Sigma ) for 15 min before infection . HeLa cells ( 2×104 cells/well ) were cultured in 24-well culture plates . Transfection of Rab7-RNAi or negative control of RNAi ( 20 pmol ) or adenovirus-mediated transfer ( moi = 100 ) were performed 48 h before infection experiments . Cells were infected as described above . To prevent the intracellular release of GAS , 0 . 5% tannic acid was added to the medium twice , both at 15 min before antibiotic treatment and at 2 h after infection . After an appropriate incubation time , infected cells were lysed in sterile distilled water and serial dilutions were plated on THY agar plates . The number of viable intracellular bacteria was determined and presented as the ratio of “intracellular live GAS at the indicated time” to “intracellular and adhered GAS at 0 h” , with ± s . e for six independent experiments . LC3 and Rab7 ( T22N ) were fused to the C-terminus of mRFP [21] and transiently expressed under the control of the CMV IE promoter . The point mutation ( T22N ) was introduced by Quick change system ( Invitrogen ) . For construction of Rab7 ( T22N ) expressing adenovirus vector , the WT Rab7 or Rab7 ( T22N ) fragment were cloned into pENTR-D-TOPO , and converted to adenovirus vector according to the manufacturer's instructions ( Invitrogen ) . Rabbit serum targeting Rab7 was a kind gift from Dr . E . Kominami ( Juntendo University , Tokyo , Japan ) . Antibodies were used at the following concentrations: rabbit anti-Rab7 , 100x dilution of a stock; mouse monoclonal anti-human Lamp1 ( clone H4A3; Santa Cruz Biotechnology , CA , USA ) , 10 µg/ml; rabbit anti-GFP [7] , 1000x dilution of a stock . The secondary antibodies were Alexa568-conjugated anti-rabbit IgG ( Molecular Probes , OR , USA ) and Alexa647-conjugated anti-mouse IgG ( Molecular Probes ) . All secondary antibodies were diluted 1∶400 . To label bacterial and cellular DNA , cells were stained with 1 µg/ml DAPI dissolved in PBS . For Immunofluorescence microscopy , NIH3T3 cells expressing GFP-LC3 were infected with GAS for 0 . 5 h and pre-permeabilized for 5 min in PBS containing 50 µg/ml digitonin . The cells were then washed once in PBS and fixed for 15 min in 4% paraformaldehyde . The cells were then blocked in PBS containing 0 . 1% gelatin ( gelatin-PBS ) for 0 . 5 h , followed by incubation with primary antibodies diluted in gelatin-PBS for 2 h . After the cells were washed six times in PBS ( 5 min for each ) , they were incubated with the secondary antibodies diluted in gelatin-PBS for 1 . 5 h and washed six times in PBS ( 5 min for each ) . All steps were done at room temperature . The cells were then mounted with Slow Fade ( Molecular Probes ) and observed under a fluorescence laser-scanning microscope ( FV1000 , Olympus , Japan ) . HeLa cells expressing GFP-LC3 were transfected with mRFP or mRFP-Rab7 ( T22N ) , cultured overnight and infected with GAS . At 0 , 2 , and 4 h after the infection period , cells were fixed and stained with DAPI . The number of cells containing GcAVs was quantified at each time point . Transfected cells were identified by mRFP fluorescence . The rate of cellular GcAV formation was expressed as a percentage of the total LC3- , mRFP-positive and GAS-infected cells . Approximately 50 cells were analyzed in each assay . For morphological analysis , HeLa cells expressing GFP-LC3 were incubated with GAS and observed at the indicated time points . The ratio of each morphologically characterized GcAV was calculated as a percentage of total observed GcAVs . Approximately 60 cells were analyzed at each time point . HeLa cells expressing GFP-Atg5 were grown , transfected and incubated infected with GAS on glass-bottom culture dishes ( Synapse Fine-View dish , SF-T-D12 , Ivic , Japan ) . At the indicated time after infection , cells were analyzed by video microscopy at 37°C on a IX81 ( Olympus ) equipped with a cooled CCD camera ( Roper Scientific , Japan ) . Time-lapse recording was operated with the SlideBook Imaging System ( Intelligent Imaging Innovations . Inc . , CO , USA ) . To observe live cells , we used a CO2 incubator ( MI-IBC , Olympus ) that allows accurate control of the microenvironment . The recorded images were processed using Image J software . The presented images are representative of approximately 45 cells from at least 8 independent experiments . The percentage of cells with GFP-Atg5 punctate dots was quantified as follows: During 1- to 2-h period after infection with GAS , the cells expressing GFP-Atg5 and mRFP/mRFP-Rab7 ( T22N ) were routinely recorded with a 1-min interval for 20 min . We scored cells in which GFP-Atg5 staining colocalized with or contacted DAPI-stained GAS in at least 2 successive frames . For individual dots , we sorted them according to whether or not the dot would colocalize with/contact GAS , when the dot first appeared in the frame . For Correlative FM-EM , cells expressing GFP-LC3 were cultured on uncoated glass bottom culture dishes with grid patterns ( Mat Tek , Ashland , MA , USA ) in 1 ml DMEM . At 0 . 5 or 1 h after infection , an equal volume of the first fixative [1 . 25% glutaraldehyde ( Distilled EM grade , EM sciences , PA , USA ) and 1% formaldehyde ( EM grade , Nacarai , Japan ) dissolved in 0 . 1 M cacodylate buffer ( pH , 7 . 4 ) ] was added , followed by 5-min incubation . After the fixative was aspirated , the cells were subsequently incubated with the second fixative ( 1 . 25% glutaraldehyde and 1% formaldehyde dissolved in the same buffer ) for 1 h . The cells were further incubated with the third fixative ( 2 . 5% glutaraldehyde and 2% formaldehyde dissolved in the same buffer ) for 4 h . All fixations were performed at room temperature . The fixed specimens were washed three times with 0 . 1 M cacodylate buffer ( pH , 7 . 4 ) containing 7% sucrose at 4°C ( 5 min for each ) . Then , the cells bearing GcAVs were identified and photographed by phase-contrast and fluorescence microscopy . Next , the samples were fixed with 1% osmium dissolved in cacodylate buffer and block-stained in 0 . 5% uranyl acetate . After dehydration in a graded ethanol series , they were embedded in Epon 812 resin ( TAAB , UK ) . Ultra thin sections were double-stained with uranyl acetate and Reynolds' lead citrate [22] , and the cells of interest were analyzed using a JEOL JEM-1011 transmission electron microscope . For conventional EM , the cells were cultured on collagen-coated plastic coverslips ( Sumitomo Bakelite , Japan ) in 500 µl DMEM in 24-well tissue culture plates . Subsequent procedures were done as described above . After a 2-h infection , HeLa cells expressing GFP-LC3 via adenoviral transduction [2] were fixed with 4% paraformaldehyde and 0 . 1% glutaraldehyde in 0 . 1 M Na-phosphate buffer , pH 7 . 4 for 1 h at room temperature . The pre-embedding silver enhancement immunogold method was performed as previously described [23] . | Autophagy has become one of the leading edge subjects in science . Autophagy occurs when a cell eats some of its cellular components and digests them . These cellular components may include cytosol and organelles as well as bacteria that has invaded the cell . Thus , autophagy plays an important role in killing pathogens . Here , we introduce an anti-bacterial autophagy called xenophagy . Group A Streptococcus ( GAS ) enters HeLa cells and escapes from the endosome into the cytoplasm for its growth . However , autophagy kicks in and traps GAS , thus preventing its survival path . Detailed morphological observation of this process reveals several specific features which were not found in canonical autophagy . These results provide key information about not only anti-bacterial autophagy , but also canonical autophagy . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"cell",
"biology"
] | 2009 | An Initial Step of GAS-Containing Autophagosome-Like Vacuoles Formation Requires Rab7 |
Unique purine-rich mRNA sequences embedded in the coding sequences of a distinct group of gammaherpesvirus maintenance proteins underlie the ability of the latently infected cell to minimize immune recognition . The Epstein-Barr virus nuclear antigen , EBNA1 , a well characterized lymphocryptovirus maintenance protein has been shown to inhibit in cis antigen presentation , due in part to a large internal repeat domain encoding glycine and alanine residues ( GAr ) encoded by a purine-rich mRNA sequence . Recent studies have suggested that it is the purine-rich mRNA sequence of this repeat region rather than the encoded GAr polypeptide that directly inhibits EBNA1 self-synthesis and contributes to immune evasion . To test this hypothesis , we generated a series of EBNA1 internal repeat frameshift constructs and assessed their effects on cis-translation and endogenous antigen presentation . Diverse peptide sequences resulting from alternative repeat reading frames did not alleviate the translational inhibition characteristic of EBNA1 self-synthesis or the ensuing reduced surface presentation of EBNA1-specific peptide-MHC class I complexes . Human cells expressing the EBNA1 frameshift variants were also poorly recognized by antigen-specific T-cells . Furthermore , a comparative analysis of the mRNA sequences of the corresponding repeat regions of different viral maintenance homologues highlights the high degree of identity between the nucleotide sequences despite very little homology in the encoded amino acid sequences . Based on these combined observations , we propose that the cis-translational inhibitory effect of the EBNA1 internal repeat sequence operates mechanistically at the nucleotide level , potentially through RNA secondary structural elements , and is unlikely to be mediated through the GAr polypeptide . The demonstration that the EBNA1 repeat mRNA sequence and not the encoded protein sequence underlies immune evasion in this class of virus suggests a novel approach to therapeutic development through the use of anti-sense strategies or small molecules targeting EBNA1 mRNA structure .
Members of the viral family Herpesviridae , which are widely distributed throughout the animal kingdom , are characterized by their large double-stranded , linear DNA genomes . The gammaherpesviruses , one of three sub-families of Herpesviridae , predominantly replicate and persist in lymphoid cells with the distinguishing characteristic that they are able to establish lifelong latent infections of their hosts [1] . Gammaherpesviruses are of particular interest mainly due to the two human viruses , Epstein-Barr virus ( EBV ) and Kaposi's sarcoma-associated herpes virus , ( KSHV ) and the diseases they cause; Burkitts lymphoma , Nasopharyngeal carcinoma and Hodgkins lymphoma in the case of EBV and Kaposi's sarcoma , primary effusion lymphomas and AIDS-related lymphoproliferative disorders in the case of KSHV [2]–[4] . Latent infection of host cells by many gammaherpesviruses is dependent upon the expression of a viral maintenance protein , which ensures persistence of the viral episome within actively dividing cells , yet simultaneously evades immune surveillance [5]–[8] . Recent studies have investigated the unique properties of gammaherpesvirus maintenance proteins that allow the virus to restrict detection by host CD8+ cytotoxic T lymphocytes ( CTLs ) at crucial times during latency [1] , [9]–[16] . Expression of the EBV nuclear protein 1 ( EBNA1 ) is widespread in all forms of EBV infection , accentuating its central role in the maintenance of the viral DNA episome , a process essential for viral persistence and associated oncogenic potential [17] , [18] . A wide range of studies have demonstrated that EBV latently infected B cells are able to escape immune recognition , due in part to an internal glycine–alanine repeat ( GAr ) domain within EBNA1 , which significantly limits MHC class I-restricted presentation of EBNA1 epitopes linked in cis [9]–[15] , [19]–[22] . An earlier report suggested that the GAr polypeptide directly interfered with the translational machinery [10] . However , more recent studies including reports from the Hoeben group , have proposed that the EBNA1 purine-rich mRNA secondary structure encoding the GAr , rather than the protein sequence , is the critical component underlying the regulation of self-synthesis and evasion of immune recognition by cytotoxic T-cells [9] , [11] , [12] , [15] , [22] . Similar to EBNA1 , the latency-associated nuclear antigen 1 ( LANA1 ) maintenance protein of the closely related KSHV virus also acts to tether the viral episome to the host genome , thereby permitting the necessary segregation of viral DNA during cell division [9] , [15] . Studies have also demonstrated that LANA1 inhibits MHC class I peptide presentation in cis as a means of immune evasion [9] , [15] . Interestingly , studies of several other members of the gammaherpesvirus family have also reported similar immune evasive properties for the maintenance proteins of these viruses [23] , [24] . To define the underlying mechanism influencing the cis-translational inhibition responsible for minimizing the exposure of EBNA1 epitopes to immune surveillance , we have designed a series of EBNA1 expression constructs encoding alternative repeat reading frames to assess their impact on self-synthesis and antigen presentation . As well as these genetic experiments , we have also undertaken a detailed comparative analysis of the mRNA and protein sequences of the repeat regions of different gammaherpesvirus maintenance protein homologues . Based on these analyses , we conclude that the cis-inhibitory effect of the internal repeat sequences of gammaherpesviruses operates at the nucleotide level and is unlikely to be mediated through the direct action of the GAr polypeptide .
The internal GAr sequence within the EBV maintenance protein , EBNA1 , has been shown to inhibit self-synthesis , which in turn significantly restricts in cis antigen presentation [10]–[14] , [19]–[22] . To assess the functional importance of the mRNA sequence versus the protein sequence of the EBNA1 internal repeat in inhibiting self-synthesis , we compared both the mRNA and encoded protein sequences of similar internal repeat structures within the viral maintenance proteins of several gammaherpesviruses . Similar to EBNA1 , these maintenance proteins are critical for the persistence of the viral genome within latently infected cells . Gammaherpesviruses have been subdivided into four genera: Lymphocryptovirus , Rhadinovirus , Macavirus and Percavirus ( Table 1 ) [25] . Lymphocryptoviruses ( LCVs ) include the well-characterized EBV or Human herpesvirus 4 [26] , [27] , Lymphocryptovirus of rhesus monkeys , and Herpesvirus papio of baboons [26] , [28] , [29] . The Rhadinoviruses include the second human gammaherpesvirus KSHV or Human herpesvirus 8 [30] , [31] , Herpesvirus saimiri ( HVS ) [32] and Rhesus monkey rhadinovirus ( RRV ) [33] . The genera Macavirus includes the Alcelaphine herpesvirus 1 [34] and a newly defined species [25] the Ovine herpesvirus 2 ( Table 1 ) [35] . The coding mRNA sequence and deduced protein sequence of the viral maintenance proteins of these gammaherpesviruses were extracted from GenBank [36] . The overall homology between the EBNA1 coding mRNA sequence and coding mRNAs for different gammaherpesvirus maintenance proteins was investigated by performing mRNA dot-plot pair wise sequence alignments [37] to visualize local alignments of repeated regions between the maintenance protein homologues and EBNA1 ( Fig . 1 ) . The over-all homology between sequences is shown as a straight line on the diagonal , while regions of repeats are shown as lots of lines in the same region , allowing visualization of where the repeated regions are between sequences . In each panel the intensity of the dot plots indicate the level of homology between the sequences being compared . As illustrated in Figure 1 , the EBNA1 internal mRNA repeat sequence is highly identical to regions of similar repeat sequences , albeit in different positions within the coding sequences of the maintenance proteins from other gammaherpesviruses . The plot in Panel A highlights a highly repetitive homologous region between Human Herpes virus 4 EBNA1 ( 280–1180 bp ) and Human Herpes virus 8 LANA1 ( 1000–2800 bp ) , while Panel B highlights a highly repetitive homologous region towards the 5′ ends of both the Human Herpes virus 4 EBNA1 ( 280–1180 bp ) and Papiine Herpes virus 1 baboon EBNA1 ( 290–580 bp ) sequences . All six viral maintenance protein mRNAs showed varying sized repeated regions that have strong homology with the internal repeat present within the EBNA1 mRNA . In Table 1 it is apparent that the identity between these purine-rich mRNA repeat sequences of EBNA1 and other viral maintenance proteins is relatively high ( 50–75 . 6% ) , whilst strikingly the corresponding repeat amino acid sequences showed markedly reduced identity levels and in some cases the complete absence of any similar conservation . For example , there is less than 1% homology between Human Herpes virus 4 EBNA1 and Human Herpes virus 8 LANA1 repeat amino acid sequences and only 2 . 1% homology between Human Herpes virus 4 EBNA1 and Macacine herpesvirus 5 Rhesus rhadinovirus ORF73 repeat amino acid sequences , despite corresponding repeat mRNA identities of 76 . 2% and 66 . 5% , respectively . Three EBNA1 expression constructs were designed comprising identical mRNA sequences whilst encoding three , alternative repeat reading frames . The constructs were used to assess the impacts of the EBNA1 repeat region mRNA and protein sequence on self-synthesis and antigen presentation . Three DNA fragments were synthesized to generate the alternative EBNA1 repeat reading frames encoding either glycine/alanine residues , referred to as E1-GA ( wild-type ) ; glycine/glutamine/glutamic acid residues , referred to as E1-GQE ( frameshift 1 ) ; or glycine/arginine/serine , referred to as E1-GRS ( frameshift 2 ) . The synthesized DNA fragments were cloned into an EBNA1 expression construct lacking the internal GAr sequence ( E1ΔGA/pcDNA3 ) to generate the EBNA1 protein sequences outlined in Figure 2 . This strategy maintained the wild-type protein sequences in the regions flanking the internal repeat . As illustrated in Figure 2 , a single nucleotide deletion near the start of the EBNA1 repeat sequence generated a strongly acidic ( GQE ) repeat domain , whilst the deletion of two nucleotides at the same position resulted in a third repeat reading frame encoding a repetitive peptide with both basic and neutral residues ( GRS ) . The corresponding insertion of either one or two nucleotides at the end of the repeat sequence allowed the contiguous encoded C-terminal domains for these constructs to maintain wild-type EBNA1 protein sequence ( Figure S1 in Text S1 ) . Thus , the three proteins generated by these constructs were highly dissimilar in their repeat regions in terms of amino acid composition and charge . For intracellular localization studies , the EBNA1 frameshift expression sequences were also sub-cloned in-frame with a sequence coding for green fluorescent protein , generating fusion proteins with GFP at the C-terminus . In addition , the H-2Kb-restricted SIINFEKL epitope from ovalbumin , was inserted in-frame into the different EBNA1-GFP frameshift variants for endogenous processing studies thereby generating the following expression constructs E1ΔGA-SIIN-GFP , E1-GA ( wild-type ) -SIIN-GFP , E1-GQE ( frameshift 1 ) -SIIN-GFP and E1-GRS ( frameshift 2 ) -SIIN-GFP . A microscopic analysis presented in Figure 3 ( panels A and B ) demonstrated an unaltered nuclear staining pattern for all three EBNA1-SIIN-GFP frameshift variants as well as for E1ΔGA-SIIN-GFP , which lacks the internal repeat . Whilst constructs encoding the wild-type GA repeat or GQE repeat resulted in similar expression levels following transfection , there was reduced EBNA1-GFP expression for the construct encoding GRS repeat sequences ( Fig . 3 , panel A ) . The reduced expression of E1-GRS ( frameshift 2 ) -SIIN-GFP transfectants was not due to differential transfection efficiencies as all three alternative reading frame constructs contain virtually identical DNA sequences which differ by only one or two nucleotides . The lower expression of E1-GRS ( frameshift 2 ) -SIIN-GFP transfectants was consistent with decreased cell viability observed in the phase contrast microscopic analysis of the GRS transfectants ( Figure 3 , panel C ) , demonstrating a reduced percentage of GFP-expressing cells following transfection . Also of note , the GFP+ve GRS-transfectants are very low GFP expressers ( MFI of 705 ) compared to an MFI of 4692 for the EBNA1 wild-type transfectants and 11 , 385 for E1ΔGA transfectants , indicating less EBNA1-GFP is being synthesized in GRS transfectants ( Figure 3 panel D ) and possibly resulting in a percentage of the GRS GFP-expressing cells being below the threshold level for GFP detection . Expression levels of the EBNA1-SIIN-GFP frameshift variants were also confirmed by flow cytometry ( Fig . 3 , panel D ) . Thus , the three EBNA1-GFP frameshift variants are all expressed and demonstrate similar nuclear localization . To discount the possibility that altered peptide sequences due to alternative reading frames within the EBNA1 repeat domain may have changed protein stability , we determined the intracellular kinetics of degradation of the EBNA1-SIIN-GFP frameshift variants following cycloheximide treatment of 293KbC2 cells transiently transfected with E1ΔGA-SIIN-GFP , E1-GA ( wild-type ) -SIIN-GFP , E1-GQE ( frameshift 1 ) -SIIN-GFP , E1-GRS ( frameshift 2 ) -SIIN-GFP or E1ΔGA-SIIN-GFP expression constructs ( in the presence or absence of 10 µM of the proteasomal inhibitor MG132 ) over a 30 hour time course . Both the E1-GA ( wild-type ) -SIIN-GFP and E1-GQE ( frameshift 1 ) -SIIN-GFP transfectants displayed a similar pattern of degradation ( Fig . 4A ) which was slightly lower than that observed for E1ΔGA-SIIN-GFP transfectants . The E1-GRS ( frameshift 2 ) -SIIN-GFP transfectants displayed a less pronounced decrease in EBNA1-SIIN-GFP expression at 24 hours ( Fig . 4A ) . The degradation kinetics carried out in the presence of the proteasomal inhibitor MG132 demonstrate that the observed loss of EBNA1-GFP fluorescence following cycloheximide treatment is due to turnover ( Figure 4A ) . In vitro translation assays of the EBNA1/pcDNA3 frameshift variants demonstrated similarly low translational efficiencies for EBNA1 sequences encoding either GQE or GRS repeat domains as observed for the wild-type GA repeat domain ( Fig . 4B ) . In contrast , the translational efficiency of the EBNA1 sequence lacking the internal repeat domain , E1ΔGA , is 10-fold higher ( p<0 . 05 ) ( Fig . 4B ) . The different migration rates observed for the EBNA1 frameshift variants is due to the nature of the highly repeated residues within the frameshift repeat domains leading to varying amounts of bound SDS per unit mass of protein . The strongly acidic protein , E1-GA ( GQE ) , binds less SDS and hence migrates slower than expected . The E1-GA ( GRS ) protein is strongly basic and binds excess SDS causing the protein to migrate faster than expected . All of the expression constructs containing the purine-rich repeat also display premature termination products arising from the difficulty in ribosome transit through this sequence . In summary , neither of the two frameshift variants E1-GA ( GQE ) or E1-GA ( GRS ) were able to override the translational inhibition observed for the EBNA1 sequence encoding the wild-type GA repeat domain . Several experiments were undertaken to examine the impact of alternative EBNA1 repeat peptide sequences on the endogenous processing of MHC class I-restricted epitopes within EBNA1 . In the first set of experiments , the endogenous loading of MHC class I molecules with a H-2Kb-restricted epitope from ovalbumin ( SIINFEKL residues 257–264 ) inserted at the C-terminus of the EBNA1 sequence was assessed [11] . H-2Kb expressing HEK293 cells were transiently transfected with either E1-GA ( wild-type ) -SIIN-GFP , E1ΔGA-SIIN-GFP , E1-GQE ( frameshift 1 ) -SIIN-GFP or E1-GRS ( frameshift 2 ) -SIIN-GFP expression constructs with transfection efficiencies of 47% , 50% , 42 . 8% and 8% respectively . Following transfection ( 24 h ) , cells were assessed by flow cytometry for GFP expression and surface expression of H-2Kb-SIINFEKL complexes using a monoclonal antibody ( 25-D1 . 16 ) that recognizes the SIINFEKL epitope bound to H-2Kb molecules [38] . Flow cytometry results shown in Figure 5 demonstrate that the percentage of cells expressing surface H-2Kb-SIINFEKL complexes was similar for all three EBNA1 repeat reading frames and ranged from 1 . 9%–2 . 0% . In contrast , a 4–4 . 3-fold increase in the surface expression of H-2Kb-SIINFEKL complexes was observed for transfectants expressing EBNA1-GFP lacking the GAr domain ( E1ΔGA-SIIN-GFP ) , indicating that all three repeat reading frames inhibited the endogenous processing of MHC class I-restricted epitopes within EBNA1 to a similar extent ( Fig . 5 ) . Transfection of a control parent plasmid without SIINFEKL provided a baseline ( 0 . 67% cells expressing surface H-2Kb-SIINFEKL complexes ) above which an increase in fluorescence would indicate specific surface expression of H-2Kb-SIINFEKL complexes . In the next set of experiments , the influence of EBNA1 repeat frameshifts on the T-cell recognition of a H-2Kb-restricted SIINFEKL epitope encoded within EBNA1 was evaluated . H-2Kb expressing HEK293 cells were transiently transfected with either E1-GA ( wild-type ) -SIIN-GFP , E1ΔGA-SIIN-GFP , E1-GQE ( frameshift 1 ) -SIIN-GFP or E1-GRS ( frameshift 2 ) -SIIN-GFP expression constructs with transfection efficiencies of 45 . 5% , 49% , 43% and 9 . 8% , respectively . Twenty-four hours post-transfection , cells were incubated with CD8+ T-cells specific for the SIINFEKL epitope and stimulation was assessed by intracellular cytokine staining assays ( ICS ) . Data presented in Figure 6 ( panels A and B ) show that both E1-GA ( wild-type ) -SIIN-GFP and E1-GQE ( frameshift 1 ) -SIIN-GFP transfectants stimulated a similar number of IFN-γ producing SIINFEKL-specific T-cells ( 4 . 2% and 4 . 4% , respectively ) . Transfectants expressing E1-GRS ( frameshift 2 ) -SIIN-GFP stimulated 4-fold less IFN-γ producing SIINFEKL-specific T-cells ( 1 . 1% ) than transfectants expressing GA ( wild-type ) -SIIN-GFP or GQE ( frameshift 1 ) -SIIN-GFP , consistent with reduced EBNA1 expression levels following transfection of EBNA1-GFP constructs expressing repeat sequences encoding GRS residues . Cells expressing EBNA1-GFP lacking the GAr domain ( E1ΔGA-SIIN-GFP ) stimulated 2–2 . 1-fold more IFN-γ producing SIINFEKL-specific T-cells ( 8 . 9% ) compared to GA ( wild-type ) ( 4 . 2% ) or GQE ( frameshift 1 ) ( 4 . 4% ) repeat domains ( Fig . 6 , panels A and B ) . The endogenous processing of a second CD8+ T-cell epitope , this time encoded within EBNA1 ( HLA B*3508-restricted , HPVGEADYFEY residues 407–417 ) was similarly assessed . EBV-negative DG75 B-cells were transiently co-transfected with E1ΔGA-SIIN-GFP or EBNA1-GFP expression vectors encoding alternative EBNA1 repeat peptide sequences; E1-GA ( wild-type ) -SIIN-GFP , E1-GQE ( frameshift 1 ) -SIIN-GFP , E1-GRS ( frameshift 2 ) -SIIN-GFP and a HLA B*3508-GFP expression construct . Transfection efficiencies were similar for all constructs ranging from 63 . 2%–68 . 8% . Co-transfection with the HLA B*3508-GFP expression construct allowed evaluation of endogenous processing of EBNA1 using HPV-specific T-cell clones . At twenty-four hours post-transfection , cells were incubated with CD8+ T-cells specific for HPV epitopes and stimulation assessed using ICS . Data presented in Figure 6 ( panels C and D ) demonstrate that cells expressing E1-GQE ( frameshift 1 ) -SIIN-GFP or E1-GRS ( frameshift 2 ) -SIIN-GFP showed no increased stimulation of IFN-γ producing HPV-specific T-cells compared to cells expressing E1-GA ( wild-type ) -SIIN-GFP . This result was identical to our earlier endogenous processing and presentation data with the SIINFEKL epitope . Collectively , the endogenous processing results demonstrate that the poor immunogenicity of EBNA1 is unlikely to be due to the alanine/glycine repetitive peptide sequence within the EBNA1 repeat since alternative repeat peptide sequences also failed to enhance the presentation of MHC class I-restricted epitopes within EBNA1 to the level observed for cells expressing E1ΔGA-SIIN-GFP , where the GAr had been deleted . To assess the impact of a purine-rich repeat sequence from another gammaherpesvirus maintenance protein on both EBNA1 expression and T-cell recognition of a SIINFEKL epitope fused to EBNA1 , a 519 nucleotide repeat sequence from the Ateline herpesvirus 3 ORF73 was synthesized and cloned into the E1ΔGA-SIIN-GFP expression vector to generate ( E1-Ateline-SIIN-GFP ) . The resulting EBNA1-Ateline expression vectors encoded a 173 amino acid acidic repeat domain of predominantly glycine and aspartic acid ( GD ) residues ( Figure S2 in Text S1 ) . A microscopic analysis presented in Figure 7A demonstrated the expected nuclear staining pattern for the E1-Ateline-SIIN-GFP variant , similar to the EBNA1 wild-type and frameshift variants . Following transfection of the E1-Ateline-SIIN-GFP expression construct , we observed a notably reduced EBNA1-SIIN-GFP expression level , as evidenced by an MFI of 1224 ( Fig . 7A ) , which was 9 fold lower than the MFI for E1ΔGA-SIIN-GFP and 3 . 5 fold lower than the MFI observed for E1-GA ( wild-type ) -SIIN-GFP ( Fig . 3 panel D ) . Next , the influence of the Ateline herpesvirus 3 repeat sequence in the context of EBNA1 on T-cell recognition of the H-2Kb-restricted SIINFEKL epitope fused to EBNA1 was also evaluated . H-2Kb expressing HEK293 cells were transiently transfected with either E1-Ateline-SIIN-GFP , E1ΔGA-SIIN-GFP or E1-GA ( wild-type ) -SIIN-GFP expression constructs . Twenty-four hours post-transfection , cells were incubated with CD8+ T-cells specific for the SIINFEKL epitope and stimulation assessed by intracellular cytokine staining assays . Data presented in Figure 7B demonstrate that E1-Ateline-SIIN-GFP transfectants stimulated only 1 . 2% of IFN-γ producing SIINFEKL-specific T-cells compared to 4 . 04% for E1-GA ( wild-type ) -SIIN-GFP transfectants and 7 . 59% for E1ΔGA-SIIN-GFP transfectants , at a responder∶stimulator ratio of 2 . 5∶1 . This result is consistent with the reduced EBNA1-SIIN-GFP expression levels observed for E1-Ateline-SIIN-GFP transfectants in Figure 7A and demonstrates that the purine-rich mRNA repeat of the Ateline herpesvirus 3 ORF73 is able to inhibit both protein expression and T-cell recognition .
Viruses that establish chronic latent infections of host cells have evolved numerous mechanisms to evade the host immune system . One such example is the Epstein-Barr virus nuclear antigen 1 , EBNA1 , which is responsible for maintenance of the viral episome within latently infected B cells . The synthesis of EBNA1 is tightly regulated to achieve levels sufficient to maintain viral infection , but low enough so as to minimize EBNA1's exposure to EBNA1-specific T-cells . The regulated inhibition of EBNA1 synthesis has been shown to occur in cis as a result of an internal purine-rich repetitive mRNA sequence that dramatically reduces the rate of EBNA1 protein synthesis [12] . Removal of the repeat sequence leads to increased EBNA1 synthesis and enhanced recognition of MHC class I-restricted epitopes within EBNA1 . The current studies demonstrate that this regulated inhibition of EBNA1 synthesis and the resultant restriction of antigen presentation and host immune recognition is independent of alternative repeat protein sequences embedded in EBNA1 mRNA . Combined microscopic analyses , in vitro translation assays and intracellular cytokine presentation experiments , investigating frameshift changes within the EBNA1 internal repeat demonstrate that altered peptide sequences within the repeat do not override the repeat's cis-inhibitory effect on EBNA1 translation and antigen presentation . The results show that the repetitive purine-rich mRNA sequence itself is responsible for the inhibition of EBNA1 protein synthesis and subsequent poor immunogenicity . When taken together with other studies [12] , [22] , these results suggest that an unusual RNA secondary structure within the repeat region may interfere with translation of the EBNA1 mRNA by inhibiting ribosome transit through the purine-rich sequence , thereby leading to a reduction in the levels of EBNA1 such that the infected cell evades the normal host immune surveillance mechanisms . Comparison of the mRNA sequences of related viruses encoding corresponding proteins responsible for maintenance of latent infections reveals the presence of highly homologous purine-rich repetitive sequences interspersed within the functional coding regions of these proteins . Although highly conserved in mRNA sequence , these repeat regions encode very different peptide sequences in the different viruses . Moreover , substituting the native EBNA1 mRNA repeat sequence with the purine-rich mRNA repeat sequence from the related viral maintenance protein Ateline herpesvirus 3 ORF73 demonstrated that the mRNA repeat of the Ateline herpesvirus 3 ORF73 is able to inhibit both EBNA1-GFP expression and T-cell recognition . These observations strongly support the conclusion that the purine-rich mRNA sequence , rather than its encoded protein sequence , is responsible for the reduced expression of these viral mRNAs . The immune suppressive effects of these mRNA repeat sequences on antigenic epitope generation may represent a more general immune evasive strategy as hundreds of eukaryotic viral mRNAs have evolved with a purine bias [39] . The loss of conservation of protein sequence in the face of evolutionary conservation of the purine-rich mRNA sequence needed for translational repression and avoidance of immune surveillance may be the result of the tendency for “translational recoding” or frame-shifting that has been shown to be induced by G-rich mRNA sequences [40] . Although the overall purine-rich mRNA repeat sequence regions are strongly conserved , such an evolutionary mechanism would lead to random frame shifting and different repeat protein sequences . This suggests that the repeat sequences are subject to strong purifying selection acting at the level of the nucleotide sequence and not the protein sequence . Separate from the repeat region in EBNA1 , there is a nuclear localization signal , two short domains flanking the internal GAr involved in binding to host cell chromosomes and also an overlapping DNA-binding and dimerization domain required for EBNA1 dimerization and binding to the OriP region of the viral genome , [8] [41] . Therefore , it is likely that the design of this protein serves two primary functions – viral genome maintenance and immune evasion , with the latter involving translational repression mediated by the repeat region mRNA sequence . The identification of mRNA repeats which inhibit EBNA1 translational efficiency and endogenous antigen presentation suggests a novel approach to potential new therapeutic interventions involving the use of specific “antisense” therapeutics aimed at the putative structures in the purine-rich mRNA sequence . Such strategies would increase the amount of EBNA1 protein in latently infected cells , thus facilitating normal immune recognition and thereby elimination of the virus by the immune system .
The Queensland Institute of Medical Research Ethics Committee approved all experiments ( P353 ) . All patients provided written informed consent for the collection of blood samples and subsequent analysis . The EBV negative cell line , DG75 was maintained in RPMI 1640 supplemented with 2 mM L-glutamine , 100 IU/ml penicillin , and 100 µg/ml streptomycin plus 10% FCS ( referred to as Growth Medium ) and used as targets for T-cell assays . HEK293 cells stably expressing the mouse class I allele H-2Kb ( 293KbC2 ) were maintained in DMEM supplemented with 5 . 56 mM D-glucose , 4 mM L-glutamine , 1 mM sodium pyruvate , 100 IU/ml penicillin and 100 µg/ml streptomycin plus 10% foetal calf serum ( referred to as DMEM/10FCS ) and were used for EBNA1 localization studies , intracellular degradation analysis and CTL assays . Comparisons between the coding sequence mRNA of EBNA1 and a number of other gammaherpesvirus maintenance proteins were performed using a pair wise sequence alignment visualized as dot plots . The alignments were performed using zPicture , which is a dynamic alignment and visualization tool based on the BLASTZ alignment program [37] . The Genbank accession numbers for the viral mRNAs were: ( HHV-4 ) EBNA1 ( NC_007605 ) ; ( HHV-8 ) Lana1 ( U75698 . 1 ) ; ( Papiine HV-1 ) baboon EBNA1 ( HPU23857 ) ; ( Macacine HV-4 ) rhesus EBNA1 ( NC_006146 . 1 ) ; ( Alcelaphine HV-1 ) ORF73 ( AF005370 . 1 ) ; ( Ovine HV-2 ) ORF73 ( AY839756 . 1 ) and ( Saimirine HV-2 ) ORF73 ( NC_001350 . 1 ) . An EBNA1 expression construct encoding native GAr sequence was generated by synthesizing a 615 nucleotide DNA fragment corresponding to EBNA1 nucleotides 209–814 and incorporating a 3′ Cla1 site ( DNA 2 . 0 , Menlo Park CA ) . This DNA fragment was cloned into the Bspe1 and a mutagenized Cla1 site ( position 250 ) of a previously generated E1ΔGA/pcDNA3 expression vector [12] to generate the expression construct E1-GA ( wild-type ) corresponding to native EBNA1 sequence encoding a 175 amino acid ( aa ) glycine/alanine repeated peptide sequence . An alternative EBNA1 frameshift expression construct was generated by altering the reading frame of the internal EBNA1 repeat sequence to encode a glycine/glutamic acid/glutamine ( GQE ) repeated peptide sequence ( 175 aa ) . This was achieved by synthesizing a second DNA fragment similar to that described above but with a single ( A ) nucleotide deletion at position 56 within the synthesized DNA fragment ( corresponding to EBNA1 nucleotide position 264 ) to generate the expression construct E1-GQE ( frameshift 1 ) . Likewise , a second EBNA1 frameshift expression construct was generated to encode a glycine-arginine-serine ( GRS ) repeated peptide sequence ( 175 aa ) by synthesizing a third DNA fragment ( again similar to the first DNA fragment described above ) but with two ( A ) nucleotides deleted at positions 56–57 within the synthesized DNA fragment ( corresponding to EBNA1 nucleotide positions 264–265 ) to generate the expression construct E1-GRS ( frameshift 2 ) . To maintain the wildtype EBNA1 reading frame immediately following the internal repeat , either a single ( G ) nucleotide was inserted at EBNA1 nucleotide position 809 by mutagenesis in the E1-GQE ( frameshift 1 ) construct or two nucleotides ( AG ) were inserted at EBNA1 nucleotide positions 813–814 in the E1-GRS ( frameshift 2 ) construct . The DNA sequences of all three frameshift expression constructs did not encode stop codons as verified by DNA sequencing . The three EBNA1 frameshift DNA sequences , E1-GA ( wildtype ) , E1-GQE ( frameshift 1 ) and E1-GRS ( frameshift 2 ) in addition to E1ΔGA were also sub-cloned in-frame with a sequence coding for green fluorescent protein ( pEGFP-N1 , CLONTECH , Palo Alto , CA ) to generate E1-GA ( wild-type ) -GFP , E1-GQE ( frameshift 1 ) -GFP , E1-GRS ( frameshift 2 ) -GFP and E1ΔGA-GFP . For the assessment of endogenous loading of MHC class I molecules , a H-2Kb-restricted epitope from ovalbumin , ( Ser–Ile–Ile–Asn–Phe–Glu–Lys–Leu , residues 257–264 ) , referred to as SIINFEKL [38] was inserted in-frame into all three EBNA1-GFP frameshift expression constructs as well as into the E1ΔGA-GFP expression construct between the 3′ end of the EBNA1 sequence and the start of the GFP sequence to generate E1-GA ( wild-type ) -SIIN-GFP , E1-GQE ( frameshift 1 ) -SIIN-GFP , E1-GRS ( frameshift 2 ) -SIIN-GFP and E1ΔGA-SIIN-GFP . Endogenous processing and surface presentation of EBNA1 was also assessed using a second epitope encoded within the EBNA1 sequence and restricted through HLA B*3508 , HPVGEADYFEY ( His–Pro–Val–Gly–Glu–Ala–Asp–Tyr–Phe–Glu–Tyr , residues 407–417 ) and referred to as HPV . HEK293KbC2 cells ( 2×105 ) were transiently transfected with 0 . 4 µg of the expression constructs E1ΔGA-SIIN-GFP , E1-GA ( wild-type ) -SIIN-GFP , E1-GQE ( frameshift 1 ) -SIIN-GFP or E1-GRS ( frameshift 2 ) -SIIN-GFP in the presence or absence of the proteasomal inhibitor MG132 ( 10 µM; Merck Biosciences ) using Effectene ( QIAGEN , Hilden , Germany ) according to the manufacturer's instructions . At 24 hours post-transfection , cycloheximide ( 50 µg/ml ) was added to each sample well . Equal aliquots of cells were trypsinized , washed and processed to measure EBNA1-GFP expression by flow cytometry at time points 0 h , 3 h , 6 h , 24 h and 30 h . EBNA1-pcDNA3 frameshift expression constructs E1-GA ( wild-type ) , E1-GQE ( frameshift 1 ) and E1-GRS ( frameshift 2 ) ; E1ΔGA and E1-Ateline were transcribed and translated in vitro with T7 RNA polymerase using a coupled transcription/translation reticulocyte lysate system ( Promega , Madison WI ) supplemented with 10 µCi 35[S]-methionine ( Perkin-Elmer Pty Ltd . , Boston , MA . ) . Lysates were subjected to SDS-PAGE followed by autoradiography and band intensities were quantified by densitometric analysis using Imagequant software ( Molecular Dynamics ) . 293KbC2 cells ( 2×105 ) , which stably express H-2Kb [38] , were transfected with 0 . 4 µg of the EBNA1-SIIN-GFP frameshift expression constructs using Effectene . A separate transfection of the parent construct without SIINFEKL was also performed to provide a negative control . Cells were harvested after an overnight transfection and stained with mAb 25D1 . 16 [38] conjugated to Allophycocyanin ( Molecular Probes , Invitrogen ) for 30 min at 4°C . Cells were washed and analyzed by flow cytometry on a FACSCanto II ( BD Biosciences ) for GFP expression and 25D1 . 16 binding . HEK293KbC2 cells ( 2×105 ) transiently transfected with EBNA1-SIIN-GFP frameshift expression constructs ( 24 h ) were incubated with ovalbumin-specific T-cells ( OT-1 ) for 3 hours at 37°C at responder to stimulator ratios of 2 . 5∶1 , 5∶1 , 10∶1 and 20∶1 and 40∶1 in DMEM/10FCS medium supplemented with Brefeldin A ( BD Pharmingen , San Diego , USA ) . Cells were washed and incubated with Allophycocyanin ( APC ) -conjugated anti-CD3 and PerCP-conjugated anti-CD8 for 30 min , rewashed , then fixed and permeabilized with cytofix/cytoperm ( BD Pharmingen ) at 4°C for 20 minutes . Cells were washed in perm/wash ( BD Pharmingen ) , incubated with PE-conjugated anti-IFN-γ ( BD Pharmingen ) at 4°C for 30 mins , rewashed and analyzed for IFN-γ production by OT-1 T-cells by flow cytometry on a FACSCanto II . DG75 cells ( 5×106 ) co-transfected ( 24 h ) with 1 . 2 µg of EBNA1-SIIN-GFP frameshift expression constructs E1-GA ( wild-type ) -SIIN-GFP , E1-GQE ( frameshift 1 ) -SIIN-GFP , E1-GRS ( frameshift 2 ) -SIIN-GFP or E1ΔGA-SIIN-GFP and 0 . 8 µg of a HLA B*3508-GFP expression construct using the Amaxa Cell Line Nucleofector Kit V ( Lonza , Cologne , Germany ) were incubated with HPV-specific T-cells overnight ( 37°C ) at responder to stimulator ratios of 2 . 5∶1 , 5∶1 , 10∶1 and 20∶1 in Growth Medium supplemented with Brefeldin A . IFN-γ production by HPV-specific T-cells was determined by intracellular cytokine staining as described above with FITC-conjugated anti-CD4 , PerCP-conjugated anti-CD8 , APC labeled B*3508 HPV Pentamer ( ProImmune , Oxford , UK ) and PE-conjugated anti- IFN-γ . HEK293KbC2 cells seeded on glass coverslips were transfected with the EBNA1-SIIN-GFP expression constructs as described above . At twenty-four hours post-transfection the cells were fixed in 4% paraformaldehyde for 20 mins , washed , permeabilized in 1% Triton-X100 in PBS for 20 mins , washed and then mounted in Pro Long Gold antifade reagent with DAPI ( Molecular Probes , Invitrogen ) . GFP fluorescence in cells was detected using a laser-scanning Bio-Rad ( Hercules , CA ) MRC600 confocal microscope with original magnification ×63 . | Viruses establishing persistent latent infections have evolved various mechanisms to avoid immune surveillance . The Epstein-Barr virus-encoded nuclear antigen , EBNA1 , expressed in all EBV-associated malignancies , modulates its own protein levels at quantities sufficient to maintain viral infection but low enough so as to minimize an immune response by the infected host cell . This evasion mechanism is regulated through an internal purine-rich mRNA repeat sequence encoding glycine and alanine residues . In this study we assess the impact of the repeat's nucleotide versus peptide sequence on inhibiting EBNA1 self-synthesis and antigen presentation . We demonstrate that altered peptide sequences resulting from frameshift mutations within the repeat do not alleviate the immune-evasive function of EBNA1 , suggesting that the repetitive purine-rich mRNA sequence itself is responsible for inhibiting EBNA1 synthesis and subsequent poor immunogenicity . Our comparative analysis of the mRNA sequences of the corresponding repeat regions of different gammaherpesvirus maintenance homologues to EBNA1 highlights the high degree of identity between the nucleotide sequences despite very little homology in the encoded amino acid sequences . These studies demonstrate the importance of gammaherpesvirus purine-rich mRNA repeat sequences on antigenic epitope generation and evasion from T-cell mediated immune control , suggesting novel approaches to prevention and treatment of latent infection by this class of virus . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"medicine",
"infectious",
"diseases",
"antigen",
"processing",
"and",
"recognition",
"epstein-barr",
"virus",
"infectious",
"mononucleosis",
"immunology",
"biology",
"viral",
"diseases",
"immune",
"response"
] | 2012 | Messenger RNA Sequence Rather than Protein Sequence Determines the Level of Self-synthesis and Antigen Presentation of the EBV-encoded Antigen, EBNA1 |
Rapid bone destruction often leads to permanent joint dysfunction in patients with septic arthritis , which is mainly caused by Staphylococcus aureus ( S . aureus ) . Staphylococcal cell wall components are known to induce joint inflammation and bone destruction . Here , we show that a single intra-articular injection of S . aureus lipoproteins ( Lpps ) into mouse knee joints induced chronic destructive macroscopic arthritis through TLR2 . Arthritis was characterized by rapid infiltration of neutrophils and monocytes . The arthritogenic effect was mediated mainly by macrophages/monocytes and partially via TNF-α but not by neutrophils . Surprisingly , a S . aureus mutant lacking Lpp diacylglyceryl transferase ( lgt ) caused more severe joint inflammation , which coincided with higher bacterial loads of the lgt mutant in local joints than those of its parental strain . Coinjection of pathogenic S . aureus LS-1 with staphylococcal Lpps into mouse knee joints caused improved bacterial elimination and diminished bone erosion . The protective effect of the Lpps was mediated by their lipid moiety and was fully dependent on TLR2 and neutrophils . The blocking of CXCR2 on neutrophils resulted in total abrogation of the protective effect of the Lpps . Our data demonstrate that S . aureus Lpps elicit innate immune responses , resulting in a double-edged effect . On the one hand , staphylococcal Lpps boost septic arthritis . On the other hand , Lpps act as adjuvants and activate innate immunity , which could be useful for combating infections with multiple drug-resistant strains .
Despite colonizing more than half of the human population at some stage during their life [1] , Staphylococcus aureus ( S . aureus ) is a highly pathogenic microorganism responsible for a broad range of infections in humans [2] . Septic arthritis , considered to be one of the most aggressive joint diseases , is most commonly caused by S . aureus [3] . In a mouse model , S . aureus induced severe bone destruction 5 days postinfection [4] . In patients , even after initiating immediate treatment , the joint damage caused by septic arthritis is often irreversible [5] , leading to permanent joint dysfunction in up to half of the patients [6] . Furthermore , the emergence of methicillin-resistant S . aureus ( MRSA ) has severely reduced the available treatment options [2 , 7] . To limit the immune response and reduce the risk of permanent joint destruction , a combination treatment of antibiotics with immunomodulatory therapy has been proposed [8 , 9] . However , there are potential dangers associated with such combination therapies as long as the challenge of antibiotic resistance remains [10 , 11] . Therefore , the identification of the bacterial components responsible for joint inflammation and destruction is key for the development of new therapies . Antibiotic-killed S . aureus is known to induce destructive arthritis , and the bacterial cell wall components are the culprits in this context [12] . Among the S . aureus cell wall components , lipoproteins ( Lpps ) are Toll-like receptor 2 ( TLR2 ) agonists and the main immune stimulators , while lipoteichoic acids are much less important [13–15] . Lipidation of Lpps is known to be crucial for virulence in murine S . aureus systemic infection [16] . Depending on the degree of acylation in the lipid moiety of the Lpps , different TLR2 receptor combinations are activated: triacylated Lpps are agonists of TLR2/TLR1 heterodimers , while diacylated Lpps are agonists of TLR2/TLR6 heterodimers [17 , 18] . Staphylococcal species differ in the length of the fatty acid in the N-acyl group of the lipid moiety , which has drastic effects on innate and adaptive immune stimulation [19] . In the present study , we hypothesized that staphylococcal Lpps are the main inducer of synovitis and joint destruction in S . aureus-induced septic arthritis . Indeed , a single intra-articular injection of S . aureus Lpps induced macroscopic , chronic and destructive arthritis , which was mediated by monocytes/macrophages . However , the Δlgt strain , an Lpp diacylglyceryl transferase ( lgt ) deletion mutant , caused more severe joint inflammation than its parental strain . This increased severity in joint inflammation was due to the better survival of the Δlgt strain , suggesting that a lack of Lpps induces immune evasion . Importantly , coinjection of S . aureus with staphylococcal Lpps in mouse knee joints resulted in radical elimination of bacteria and diminished bone erosion . This protective effect was mediated by the lipid moiety of the Lpps and was fully dependent on TLR2 and the recruitment of neutrophils .
We injected the purified staphylococcal Lpp Lpl1 intra-articularly ( i . a . ) into mouse knee joints . Lpl1 is a model Lpp derived from the νSaα-specific lipoprotein-like cluster ( lpl ) that exists in highly pathogenic and epidemic S . aureus strains [20] . One single injection of Lpl1 at a dose of 10 μg/knee caused macroscopic joint inflammation after 24 hours , and the inflammation lasted for at least 21 days ( Fig 1A and 1B ) . The arthritogenic effect of Lpl1 was dose-dependent—even a 100-fold lower dose ( 0 . 1 μg/knee ) induced synovitis ( Fig 1C and 1D ) . Importantly , severe bone erosions were observed on day 7 , and all Lpl1-injected joints had erosions on day 10 , as verified by microcomputed tomography ( μCT ) scans ( Fig 1E–1G ) . Histologically , the highly inflamed synovium , pannus formation , and severe bone destruction ( Fig 1H ) exhibited characteristics of the typical histopathological picture of S . aureus septic arthritis [21] . The minimal bone destruction-inducing dose of Lpl1 was much higher ( 10 μg/knee ) than the inflammation-inducing dose ( 0 . 1 μg/knee ) . To further understand whether the lipid- or protein-moiety of Lpl1 was responsible for the arthritogenic effect , Lpp lacking the lipid moiety Lpl1 ( -sp ) was compared to the intact Lpp , Lpl1 ( +sp ) . Lpl1 ( -sp ) completely lacked the capacity to induce arthritis ( Fig 1I ) , suggesting that the lipid moiety of staphylococcal Lpps is fully responsible for their arthritogenic properties . Indeed , in vitro splenocyte proliferation was induced by both Lpl1 ( +sp ) and Pam3CSK4 ( a synthetic lipopeptide mimicking the N-terminal lipid portion of Lpps ) but not by Lpl1 ( -sp ) ( Fig 1K ) . To compare the arthritogenic capacity of Lpps with other S . aureus components , we injected mice i . a . with the staphylococcal superantigen toxic shock syndrome toxin-1 ( TSST-1 ) and with the peptidoglycan ( PGN ) purified from a mutant strain lacking Lpp diacylglyceryl transferase [15] . Only very mild and transient knee joint swelling was observed on day 1 , and the swelling disappeared by day 3 in the knee joints injected with 10 μg of PGN ( Fig 1J ) . Interestingly , heat-treated Lpl1 and Pam3CSK4 preserved their arthritogenic capacity ( S1 Fig ) , strongly suggesting that the Lpl1 is heat-insensitive . To understand the cellular mechanism behind Lpp-induced arthritis , we further analyzed the immune cells present in the local synovium using flow cytometry one day after Lpl1 injection . Synovial tissues from Lpl1-injected knee joints of wild-type mice demonstrated higher numbers and frequencies of CD11b+F4/80+ cells ( monocytes/macrophages ) and CD11b+Gr1+F4/80- cells ( neutrophils ) than those of phosphate-buffered saline ( PBS ) -injected knee joints . Significantly decreased numbers of infiltrating monocytes/macrophages and neutrophils were observed in TLR2-deficient ( TLR2-/- ) mice ( Fig 2A–2C ) . No difference with regard to B- and T-cells was observed . These results suggest that TLR2 is highly important for neutrophil and monocyte recruitment into synovial tissue following Lpl1 injection . Next , to elucidate which immune cells were responsible for the onset of arthritis , mice depleted of monocytes/macrophages , neutrophils or T-cells were i . a . injected with 0 . 33 μg Lpl1 , and the severity of the histopathological synovitis was examined on day 3 . The depletion of synovial macrophages and infiltrating monocytes by clodronate liposomes significantly reduced the severity of the synovitis ( Fig 3A ) , whereas neutrophil depletion by Ly6G antibodies had no effect on synovitis severity ( Fig 3B ) . To investigate the role of T-cells , mice were simultaneously depleted of CD4 and CD8 T-cells by intraperitoneal injection of anti-CD4 and anti-CD8 antibodies . No notable difference regarding the severity of synovitis was observed between the groups ( Fig 3C ) . Additionally , CTLA-Ig treatment ( abatacept ) that blocks T-cell activation had no effect on synovitis development , suggesting that T-cells were not essential for acute Lpp-induced joint inflammation . Fig 3D represents a typical picture of the knee joint from mice depleted of monocytes/macrophages showing the absence of leukocyte infiltration . To examine which cells were responsible for causing bone erosion , a higher dose of Lpl1 ( 10 μg/knee ) was i . a . injected . In line with the data described above , depletion of macrophages significantly attenuated macroscopic inflammation ( Fig 3E ) and more strikingly , completely protected the joints from bone damage caused by Lpl1 ( Fig 3F and 3H ) . No difference was found regarding bone erosions between the controls and neutrophil-depleted mice ( Fig 3G and 3H ) . TNF-α and IL-1 , released by monocytes/macrophages , are known to play a crucial role in septic arthritis [10 , 12 , 22] . To study the role of these cytokines in Lpl1-induced synovitis , both anti-TNF treatment ( etanercept ) and anti-IL1 treatment ( anakinra ) [10 , 11] were used . Anti-TNF , but not anti-IL1 , treatment significantly reduced the synovitis severity ( Fig 3I ) compared to the severity in the PBS-treated controls , indicating that Lpl1-induced arthritis is partially mediated by TNF-α . Finally , to study whether joint destruction caused by Lpp , a major ligand for TLR2 [23 , 24] , is mediated through this receptor , TLR2-/- mice were i . a . injected with Lpl1 . The TLR2-/- mice barely developed clinical signs of arthritis ( Fig 3J ) , and no joint destruction was observed compared to that of the wild-type mice , of which 100% presented joint erosions ( Fig 3K ) . To understand the role of Lpps in S . aureus septic arthritis , both heat-killed and live S . aureus SA113 strains , as well as a deletion mutant of the SA113 strain lacking lgt , were inoculated into the mouse knee joints . The heat-killed SA113Δlgt mutant strain resulted in less joint swelling than the SA113 parental strain on days 1 and 3 ( Fig 4A ) . Surprisingly , the reverse phenomenon was observed when live bacteria were used . The SA113Δlgt mutant strain caused a significantly higher degree of joint swelling on days 3 and 7 than the arthritis caused by its parental strain ( Fig 4B ) . The discrepancy between the heat-killed and live S . aureus SA113Δlgt mutant strains was also observed with regard to IL-6 levels in the knee homogenates on day 3 . Higher levels of IL-6 were induced by the live SA113Δlgt mutant , and a tendency towards lower levels of IL-6 was induced by the heat-killed SA113Δlgt mutant compared to the parental SA113 strain ( Fig 4C ) , whereas no differences were seen with regard to TNF-α levels ( Fig 4D ) . This unexpected increase in joint swelling and IL-6 levels could be explained by the higher bacterial load found in the knee joints of the SA113Δlgt mutant-inoculated mice on day 3 ( Fig 4E ) . To further elucidate the mechanism whether SA113Δlgt mutant survived better in joints than its parental stain , we analyzed the cytotoxicity of those strains towards mouse splenocytes ( S2 Fig ) . No significant difference was observed between those two strains , demonstrating that SA113Δlgt mutant does not have enhanced killing capacity of immune cells compared to its parental strain . We speculate that SA113Δlgt mutant is more resistant to immune-mediated bacterial killing than its parental stain . We reasoned that this might be due to a lack of Lpps that trigger the immune response . To test the above hypothesis , live S . aureus SA113Δlgt mutants mixed with various concentrations of Lpl1 ( +sp ) , Lpl1 ( -sp ) , or PBS were i . a . injected into mouse knee joints . Strikingly , the bacterial load in the knee joints was dose-dependently reduced when SA113Δlgt mutant was simultaneously administered with Lpl1 ( +sp ) . For the Lpp dose of 6 . 5 μg/knee , only 1 out of 5 knee joints was positive for bacterial culture with very low bacterial counts ( 101 colony forming units ( CFUs ) /knee ) , whereas in the PBS controls , all knee joints had high CFU counts ( 105 CFU/knee ) on day 3 postinfection ( Fig 4F ) . In contrast , a comparable dose ( 5 μg/knee ) of PGN purified from the SA113Δlgt mutant completely lacked the capacity to induce an eradicating effect on the bacteria ( Fig 4G ) . Importantly , the bacterial elimination effect was completely abolished when Lpl1 ( -sp ) , lacking the lipid moiety , was used ( Fig 4H ) . Additionally , synthetic lipopeptides ( Pam2CSK4 and Pam3CSK4 ) exhibited similar bacterial elimination effects as the Lpps , indicating that the lipid moiety of Lpl1 is important for bacterial elimination ( Fig 4I ) . To provide the evidence to support our hypothesis that SA113Δlgt mutant arouses immune responses to a lesser extent than its parental strain , we i . a inoculated both bacterial strains into mouse knee joints and compared the knee sizes , CFU counts in joints , and the levels of neutrophil attracting chemokines ( KC and MIP-2 ) on day 1 postinfection . The joints inoculated with SA113 strain were significantly more swollen ( Fig 5A ) and tended to have lower bacterial counts ( Fig 5B ) compared to SA113Δlgt injected knees . However , higher levels of KC were detected in SA113 injected knees than SA113Δlgt injected knees ( Fig 5C ) . Similar trend was also found in MIP-2 levels ( Fig 5D ) . These results demonstrate that SA113Δlgt is less potent to induce neutrophil attracting chemokines in S . aureus septic arthritis compared with its parental strain . We further analyzed the effect of addition of exogenous Lpl1 to SA113Δlgt in the similar setting . Strikingly , addition of Lpl1 to SA113Δlgt resulted in increased joint swelling ( Fig 5E ) , decreased CFU counts ( Fig 5F ) , and higher levels of KC and MIP-2 ( Fig 5G and 5H ) in the knees on day 1 postinfection . To understand the mechanism by which S . aureus Lpps mediate bacterial killing , we first studied whether the Lpps possess bactericidal capacity . Our data suggest that neither Lpp nor lipopeptides had direct bactericidal effect since in vitro incubation of SA113Δlgt with Lpl1 or Pam3CSK4 did not affect bacterial proliferation or survival ( S3 Fig ) . Since Lpps are specific ligands for TLR2 [23 , 24] , we hypothesized that the protective effect of Lpps is mediated by TLR2 . Indeed , the bacterial elimination induced by Lpl1 was completely abolished in TLR2-/- mice ( Fig 6A ) , strongly suggesting that TLR2 is essential for Lpp-induced bacterial killing . Next , we studied the importance of monocytes/macrophages and neutrophils in the knee joints after Lpl1 injection . The knee joints of monocyte/macrophage-depleted mice injected with a mixture of Lpl1 and SA113Δlgt mutant exhibited significantly higher bacterial counts than the knee joints of the non-depleted mice ( Fig 6B ) . However , macrophage depletion failed to fully abolish the effect of Lpps on bacterial elimination , suggesting that other cell types also contributed to the effect . Importantly , the protective effect of the Lpps completely disappeared in the neutrophil-depleted mice ( Fig 6C ) , suggesting that Lpl1 elicits neutrophils to kill bacteria . We further studied whether Lpps boost the phagocytic capacities of phagocytes in vitro . Mouse peritoneal macrophages stimulated with Lpl1 were incubated with green fluorescent protein ( GFP ) -expressing S . aureus . The S . aureus internalization rates in macrophages were analyzed by flow cytometry imaging . As expected , the opsonization of bacteria with mouse sera resulted in a 3–4 times higher rate of phagocytosis . However , no notable differences were observed between the Lpl1-stimulated and non-stimulated groups ( S4 Fig ) . Furthermore , the whole blood killing assay was thereafter performed to verify whether SA113Δlgt mutant survived better in whole blood than its parental strain . No tangible difference was observed between groups ( S5 Fig ) . As Lpps induced TNF-α release by macrophages ( S6 Fig ) and anti-TNF treatment attenuated the severity of Lpp-induced arthritis ( Fig 3I ) , we posed the question of whether the bacterial eliminating effect of Lpp was TNF-dependent . Coinjection of Lpl1 and S . aureus SA113Δlgt mutant reduced bacterial loads in a similar manner in mice treated with anti-TNF drug as in mice receiving PBS treatment ( Fig 6D ) , suggesting that TNF is not involved in Lpl1-mediated bacterial killing . S . aureus Lpps are known to induce nitric oxide production by macrophages [25] , and the expression of inducible nitric oxide synthase ( iNOS ) is associated with protective immunity against bacterial infections [26] . We used an iNOS inhibitor ( 1400W ) to block iNOS activity in mice receiving a mixture of Lpl1 and the SA113Δlgt strain . No notable difference was observed between the PBS-treated and iNOS inhibitor-treated animals ( Fig 6D ) . Since macrophages , but not neutrophils , are present in the synovium of healthy joints and monocytes/macrophages were responsible for the Lpp-induced joint inflammation ( Fig 2 and 3 ) , we hypothesized that Lpl1 stimulates macrophages via TLR2 , resulting in the release of large amounts of chemokines that in turn recruit neutrophils to kill bacteria . Indeed , peritoneal macrophages from wild-type mice stimulated with Lpl1 exhibited a quick dose-dependent release of neutrophil chemoattractant MIP-2 and KC 4 hours after stimulation and the monocyte-attracting chemokine MCP-1 24 hours after stimulation , whereas macrophages from TLR2-/- mice displayed no such chemokine release ( Fig 7A–7C ) . Mouse splenocytes , mainly composed of B- and T-lymphocytes , produced neither MIP-2 nor KC upon Lpl1 stimulation at 4 hours nor MCP-1 at 24 hours in wild-type and TLR2-/- mice . After 24 hours of stimulation , TNF-α levels were elevated in the supernatants of both peritoneal macrophages and splenocytes from wild-type mice stimulated with Lpl1 ( +sp ) and Pam3CSK4 but not those stimulated with Lpl1 ( -sp ) ( S6 Fig ) . As expected , increased TNF-α was observed in only LPS-stimulated cells from TLR2-/- mice . To further elucidate the importance of neutrophils attracting chemokine release in the Lpp-induced protective effect , CXCR2 blocking antibodies were used to inhibit in vivo neutrophil chemotaxis . Indeed , CXCR2 blocking antibodies efficiently reduced the total number of infiltrating neutrophils in Lpl1-injected knee joints to 13% on day 3 . Importantly , decreasing the infiltrating neutrophils by CXCR2 blocking antibodies almost fully abrogated the effect of bacterial elimination induced by Lpl1 ( Fig 7D ) , strongly suggesting that the release of neutrophil-attracting chemokines upon Lpp stimulation is the key mechanism of Lpp-induced bacterial killing . We next investigated whether enhanced bacterial elimination induced by Lpps leads to a better clinical outcome in septic arthritis . Mice were inoculated i . a . with a mixture of Lpl1 and the S . aureus LS-1 strain that was originally isolated from a mouse that spontaneously developed septic arthritis [21] . Mice injected with a mixture of Lpl1 and LS-1 had less swelling in their knee joints on days 7 and 10 postinfection ( Fig 8A and 8B ) and less bacterial load in their joints than mice in the control group that received a mixture of PBS and LS-1 ( Fig 8C ) . Furthermore , these mice exhibited significantly less pronounced bone erosions with lower frequency than the mice in the control group ( Fig 8D–8F ) , suggesting that the bacterial eradication effect elicited by Lpl1 applies not only to the S . aureus mutant strain deficient in lipidation of prelipoproteins but also to the wildtype S . aureus strains .
Our results demonstrate the dual role of staphylococcal Lpp in S . aureus-induced septic arthritis . On the one hand , Lpps induce inflammatory reactions and joint destruction mediated by monocytes/macrophages . On the other hand , Lpps cause the quick release of chemokines and consequent neutrophil recruitment , resulting in efficient bacterial killing . Importantly , both the lipid moiety of Lpps and TLR2 were identified as the molecular structures responsible for this outcome . Postinfectious complications in septic arthritis , such as joint deformation and deleterious contractures , remain a major medical challenge . Exaggerated immune responses have been proposed as the cause of such complications [9 , 12] . There are many bacterial components of S . aureus that possess the capacity to induce joint inflammation [12 , 27 , 28] . However , it is still unclear which component is most important in real-life infections . To address this uncertainty , not only the capacity of immune stimulation but also the quantity of components expressed in a single bacterium should be taken into consideration . A single intra-articular injection of Lpp ( 10 μg/knee ) induced chronic macroscopic arthritis lasting for at least 3 weeks as well as severe joint destruction verified by both histopathological and radiological examinations , demonstrating that Lpp is an arthritogenic molecule . Lpps were highly potent since they exerted a strong immunostimulatory effect even at the nanogram level in vivo . More importantly , the mice injected with the SA113Δlgt mutant strain lacking lipidation displayed less severe joint swelling than the mice injected with the SA113 parental strain at the early time points before the bacterial proliferation of Δlgt mutants exceeded that of the parental strains . This finding strongly suggests that Lpps are one of the major arthritogenic bacterial components . However , we have to keep in mind that the Δlgt strain that lacks lipidation maintained the capacity to induce joint inflammation , indicating that Lpps are not the only molecule in S . aureus that cause joint inflammation . Although intra-articular injection of PGN ( 10 μg/knee ) induced only transient and milder joint swelling , the importance of PGN in arthritis induction cannot be ruled out since PGN is the most abundant molecule in S . aureus . The synergistic effect of PGN and lipopeptides in immune activation has been reported previously [29] . Our data compellingly demonstrate that both neutrophils and monocytes rapidly migrated into the joints injected with Lpps , and monocytes/macrophages were fully responsible for joint destruction in Lpp-induced arthritis . Focal bone destruction in autoimmune arthritis is due to excessive bone resorption resulting from osteoclast activation that is mediated by local expression of receptor activator of nuclear factor kappa-B ( RANKL ) that is higher than that of its decoy receptor osteoprotegerin ( OPG ) [30] . Osteoclasts not only exist inside of the bone but also can be derived from mature monocytes and macrophages when a suitable microenvironment is provided by bone marrow-derived stromal cells [31] , which might be the case in this study . In fact , monocytes/macrophages were shown to mediate bone erosion in arthritis induced by other S . aureus components , such as bacterial DNA [27] and peptidoglycan [28] , as well as antibiotic-killed S . aureus [12] , suggesting that monocytes/macrophages might be the most important immune cells that determine the progression of septic arthritis . Indeed , mice depleted of monocytes developed less severe septic arthritis and joint lesions despite decreased bacterial clearance and higher mortality [32] . Macrophages are a major source of many cytokines involved in the immune response . In autoimmune arthritic diseases , proinflammatory cytokines play an important pathogenetic role . In particular , TNF release by macrophages , fibroblasts and T-cells in inflamed synovial tissue leads to joint swelling and subsequent joint destruction [33] . Furthermore , a previous study showed that the blockade of TNF , but not IL-1 , resulted in a reduction of bone erosions in a murine TNF-driven arthritis model [34] . Previously , we have shown that antibiotic-killed S . aureus induces arthritis through the TNF receptor . In the present study , TNF also played an important role in Lpp-induced arthritis since a ) Lpps induced TNF release in both macrophages and splenocytes and b ) TNF inhibition , but not IL-1 inhibition , significantly reduced synovitis severity . Unexpectedly , the live Δlgt strain gave rise to significantly more severe joint inflammation , although the heat-killed Δlgt strain caused less synovitis . This result is due to the fact that the Δlgt strain is more tolerant to host immune-mediated bacterial killing than the parental strain , as the bacterial load of the SA113 parental strain in the knee joints was significantly lower than that of the Δlgt strain . This finding is in line with the previous report that revealed that a lack of S . aureus Lpps causes bacterial immune evasion and lethal infections with disseminated abscess formation in mice with systemic S . aureus infection [35] . The importance of the lipid moiety in bacterial elimination was further confirmed by our experiments when Lpps were coinjected with S . aureus into mouse knee joints . This situation is a perfect example of the dual sides of host responsiveness to bacterial infections—on one hand , the host response protects the host against bacteria , but on the other hand , the response sometimes increases the infection severity when danger signals trigger exaggerated host responses . How is the bacterial killing effect mediated by Lpps ? Lpps themselves had no direct effect on bacterial proliferation . Rather , the enhanced local immune response triggered by Lpps was responsible for the bacterial killing effect . Macrophages , neutrophils and natural killer ( NK ) cells are the most important immune cell types in innate immunity . NK cells can be activated and exert their biological functions upon stimulation by synthetic lipopeptides with sequences from Lpps of S . aureus [36] . Macrophages seem to play a partial role , as the indirect bacterial killing effect was diminished in mice depleted of macrophages/monocytes . This finding suggests that Lpps have the capacity to trigger the activation of macrophages to better control and eliminate bacteria . Macrophages are phagocytes that play an important role in S . aureus septic arthritis [32] . S . aureus Lpps are the major inducers of inducible nitric oxide synthase ( iNOS ) and nitric oxide ( NO ) production in mouse macrophages [25] . The production of nitric oxide is important in controlling bacterial infections [26 , 37] . However , treatment with an iNOS inhibitor had no effect on the bacterial eradication induced by Lpps , suggesting that NO production is not critical for Lpp-induced bacterial killing . Notably , no complete abrogating effect was observed as a result of macrophage depletion , which indicates the involvement of other immune cells . Neutrophils were the most predominant cell type that infiltrated the local joints upon Lpp stimulation . Additionally , the total abrogation of the bacterial killing effect by Lpps in neutrophil-depleted mice strongly suggests that neutrophils are fully responsible for the protective effect of Lpps . In healthy joints , there are a very limited number of neutrophils . For the neutrophils to reach the joint cavity and exert their biological function , neutrophil chemoattractants are needed . KC and MIP-2 are known major chemokines responsible for recruiting neutrophils , and both bind to CXCR2 [38] . Resident tissue macrophages have been shown to be the major source of neutrophil chemokines [39] . In the current study , macrophages , but not T- or B-cells , quickly released large amounts of neutrophil-recruiting chemokines upon Lpp stimulation , and such chemokine release was fully controlled by lipid moiety-TLR2 signaling . Indeed , a positive feedback loop may exist in Lpp-injected knee joints , as monocytes are recruited to the local joints where they probably respond to Lpp stimulation with chemokine release that leads to an additional influx of monocytes and neutrophils to the local inflammation site . Actually , the bacterial killing effect of phagocytes might also be increased by Lpps , as Lpps can directly activate neutrophils by upregulating CD11b/CD18 and enhancing the production of reactive oxygen species [40] . In recent years , immune therapy targeting negative regulators of immune activation ( immune checkpoints ) has been an exciting research area in drug development , with promising results achieved in patients with a variety of cancers [41] . Similarly , in infectious diseases , the activation of the immune system to eliminate invading bacteria has always been an ultimate goal to overcome the challenges caused by the rapid emergence of antibiotic-resistant bacteria . The development of a vaccine that prevents S . aureus infections is of interest . However , thus far , all attempts to develop active or passive immunization against S . aureus have failed , which might be due to the lack of a well-defined , single virulence factor in S . aureus [42] . Purified Lpp acted in a protective fashion by means of less radiological bone destruction after treatment with a mixture containing a known pathogenic staphylococcal strain , i . e . , LS-1 . Our results suggest that Lpps/lipid moieties might be used as a potential candidate for immune therapy to combat local S . aureus infections , e . g . , septic arthritis and osteomyelitis . In this case , the dose of Lpps should be carefully controlled to keep the local tissue damage and efficient bacterial killing effect in balance . In addition , a similar protective immune response might also be induced by staphylococcal Lpps to eradicate other invading pathogens , such as gram-negative bacteria or even antibiotic-resistant bacteria . In conclusion , intra-articular injection of staphylococcal Lpps induced chronic joint inflammation and bone erosion . The observed effect was mediated by macrophages via TLR2 signaling and partially involved TNF . Surprisingly , purified S . aureus Lpps strengthened the immune response , consequently reducing bacterial burden and attenuating bone destruction in septic arthritis . Our findings may pave the way for the development of a novel strategy to address the challenge of antibiotic resistance .
Mouse studies were reviewed and approved by the Ethics Committee of Animal Research of Gothenburg ( Ethical number 58–2015 ) . Mouse experiments were conducted in accordance with recommendations listed in the Swedish Board of Agriculture's regulations and recommendations on animal experiments . Female NMRI mice and C57Bl/6 wild-type mice of both sexes , aged 6–8 weeks , were purchased from Envigo ( Venray , Netherlands ) and Charles River Laboratories ( Sulzfeld , Germany ) , respectively . Toll-like receptor 2-deficient B6 . 129-Tlr2tm1Kir/J ( TLR2-/- ) mice of both sexes were purchased from The Jackson laboratory ( Bar Harbor , Maine , USA ) . All mice were housed in the animal facility of the Department of Rheumatology and Inflammation Research , University of Gothenburg . Mice were kept under standard temperature and light conditions and were fed laboratory chow and water ad libitum . The SA113 , SA113Δlgt mutant[15] , and LS-1 [43] S . aureus strains were prepared as described . The strains were stored at -70°C until use . Before the experiments , the bacterial solutions were thawed , washed with sterile PBS , and adjusted to the required concentration . For experiments with heat-killed bacteria , the SA113 and SA113Δlgt mutant S . aureus strains were heat-killed at 95°C for 45 min and thereafter adjusted to a similar concentration using optical density at 600 nm ( OD600 ) . To ensure that no bacteria survived , the bacterial suspensions were plated and cultured for 24 hours . No bacterial growth was detected . The preparation and purification of the S . aureus lipoproteins Lpl1 ( +sp ) and Lpl1 ( -sp ) were performed by Dr . Nguyen ( Microbial Genetics , University of Tübingen , Germany ) , as previously described [20] . Lpl1 ( +sp ) was isolated from the membrane fraction of S . aureus SA113 ( pTX30::lpl1-his ) , whereas Lpl ( -sp ) was isolated from the cytoplasmic fraction of S . aureus SA113Δlgt ( pTX30::lpl1 ( -sp ) -his ) . Both of these Lpl1-his proteins were purified via Ni-NTA affinity chromatography . For the enhancement of protein expression , the clones were first cultivated aerobically at 37°C in the absence of xylose ( BO-medium ) until OD578 nm ≈ 0 . 5 was reached and were thereafter continuously cultivated for 4 hours in the presence of 0 . 5% xylose to induce Lpl1 expression . The bacterial cells were harvested and washed twice with Tris buffer ( 20 mM Tris , 100 mM HCl , pH 8 . 0 ) . Then , the pellet was resuspended in Tris buffer containing a protease inhibitor tablet ( Merck , Darmstadt , Germany ) and lysostaphin ( 30 μg/ml ) and was incubated at 37°C for 2 hours to disrupt the cell wall . After the first ultracentrifugation ( 235 , 000 x g for 45 min at 4°C ) , the supernatant containing the cytoplasmic proteins was collected for the next purification step . For membrane fraction isolation , the pellet was subsequently dissolved overnight at 6°C with Tris buffer containing 2% Triton-X100 . After the second ultracentrifugation step , the supernatant containing membrane fragments was collected . The purification step was carried with Ni-NTA Superflow beads ( Qiagen , Germany ) . The Ni-NTA beads capturing Lpl1 proteins were washed four times with the first washing buffer ( Tris buffer containing 0 . 25% Triton X-100 and 20 mM imidazole ) , followed by washing 2 times with the second washing buffer ( Tris buffer containing 0 . 25% Triton X-100 and 40 mM imidazole ) . Finally , Lpl1 was eluted with Tris buffer containing 500 mM imidazole . The Lpl1 were concentrated via a centrifugal ultrafilter unit with a molecular mass cut-off of 10 kDa ( Sartorius AG , Göttingen , Germany ) . The concentrated samples of Lpl1 were dialyzed overnight at 6°C with Dulbecco’s PBS ( DPBS ) buffer ( Life Technologies , Darmstadt ) by a MWCO 6–8 kDa tube dialyzer ( Merck , Darmstadt ) and were subsequently lyophilized overnight . A total of 2 μg of lyophilized samples were dissolved in water and subjected to sodium dodecyl sulfate-polyacrylamide gel electrophoresis ( SDS-PAGE ) to determine the purity and quantity of purified protein samples . The purified compounds of Lpl1 were stored at -70°C until use and were adjusted in PBS to the required concentration before each experiment . To study the arthritogenic properties of S . aureus Lpps , six sets of experiments were performed , and NMRI , C57Bl/6 wild-type or TLR2-/- mice were intra-articularly injected into the knee joint with one of the following compounds in 20 μl of PBS: 1 ) purified Lpl1 ( +sp ) or Lpl1 ( -sp ) S . aureus Lpps; 2 ) heat-killed SA113 or SA113Δlgt mutant S . aureus strains; 3 ) live SA113 or SA113Δlgt mutant S . aureus strains; 4 ) solutions containing mixtures of SA113Δlgt mutant with either Lpl1 ( +sp ) , Lpl1 ( -sp ) , PGN or two synthetic lipopeptides , Pam2CSK4 and Pam3CSK4 ( EMC , Tübingen , Germany ) ; or 5 ) a mixture of the LS-1 S . aureus strain with either Lpl1 ( +sp ) or LS-1 in PBS; or 6 ) heat-treated Lpl1 ( +sp ) or Pam3CSK4 or unheated Lpl1 ( +sp ) or Pam3CSK4 . For experiments with heat-treated purified compounds , Lpl1 ( +sp ) and Pam3CSK4 were heat-treated at 95°C for 45 min . The severity of the clinical arthritis was judged by measuring the difference between the diameters of the knee joints with a caliper every 2–3 days . Knee synovial tissue was collected from C57Bl/6 wild-type and TLR2-/- mice that received i . a . injection of Lpl1 ( +sp ) ( 5 μg/knee ) or PBS and was placed in RPMI medium ( Fisher Scientific ) . The tissue was resuspended in medium with DNase I ( Sigma-Aldrich ) and type IV collagenase ( Fisher Scientific ) and was incubated for 1 hour at 37°C . A single-cell suspension was obtained after the tissue was homogenized and passed through a 35 μm cell strainer ( Becton Dickinson ) . Synovial cells were then analyzed using the following antibodies: V450-conjugated anti-CD11b ( Becton Dickinson ) , allophycocyanin ( APC ) -conjugated anti-F4/80 ( BioLegend ) , PerCP-Cy5 . 5-conjugated anti–Gr-1 ( BioLegend ) , fluorescein isothiocyanate ( FITC ) -conjugated anti-CD3 ( BioLegend ) , and PerCP-conjugated anti-CD19 ( BioLegend ) . Cells were analyzed using a FACSverse flow cytometer ( Becton Dickinson ) . FlowJo version 10 . 1 software ( Tree Star , Ashland , USA ) was used to analyze the data . Clodronate liposomes ( Liposoma BV , Netherlands ) are known to function as selective eliminators of macrophages [44] . NMRI mice were i . a . injected in the knee joints with a volume of 20 μl of clodronate liposomes or PBS control liposomes ( Liposoma BV , Netherlands ) 1 day prior to challenge with Lpl1 or coinjections of SA113Δlgt mutant with either Lpl1 or PBS . The mice were also treated intravenously with 200 μl of clodronate liposomes or PBS control liposomes the day before the challenge and on days +1 , +3 , and +5 after the challenge . Anti-Ly6G ( clone 1A8; BioXCell ) , a specific monoclonal antibody ( mAb ) , is known to selectively deplete murine blood neutrophils [45] . NMRI mice were intraperitoneally ( i . p . ) injected with a dose of 400 μg of anti-Ly6G or isotype control ( clone 2A3; BioXCell ) in 200 μl of PBS/mouse the day before the challenge and on days +1 and , +4 after the challenge with Lpl1 or coinjections of SA113Δlgt mutant with either Lpl1 or PBS . CD4 and CD8 T-cells were depleted simultaneously using rat anti-mouse CD4 mAb ( clone GK1 . 5; BioXCell ) and rat anti-mouse CD8α ( clone 2 . 43; BioXCell ) mAb; a rat IgG2b isotype control ( clone LTF-2; BioXCell ) served as a control . NMRI mice were i . p . injected with a dose of 400 μg of each antibody in 200 μl of PBS/mouse the day before and the day after challenge with Lpl1 . The efficacy of cell depletion was verified by flow cytometry . The depletion was carried out as described 1 day prior to blood collection . Mouse blood was collected into heparin tubes , erythrocytes were depleted with lysis buffer ( 0 . 16 M NH4Cl , 0 . 13 mM EDTA , and 12 mM NaHCO3 ) , and samples were washed . The single-cell suspensions were then adjusted in FACS buffer to a density of 2x106 cells/mL and analyzed using the following antibodies: V450-conjugated anti-CD11b , PE-Cy7-conjugated anti-Ly6G , Per-CP-conjugated anti-CD4 and PE-conjugated anti-CD8 ( all from Becton Dickinson ) APC-conjugated anti-F4/80 ( BioLegend ) . The representative flow cytometry plots are presented in S7 Fig . For cell depletion protocols , 85 . 6% of monocytes ( CD11b+ , Ly6G- and F4/80+ white blood cells ) , 99 . 6% of neutrophils ( CD11b+ , Ly6G+ , F4/80- ) , 97 . 3% of CD11b- CD4+ T cells and 90% of CD11b- CD8+ T cells were depleted ( S7 Fig ) . Etanercept ( Enbrel; Wyeth Europa ) , a soluble TNF receptor , was used for the anti-TNF treatment because it fully inhibits the biological function of murine TNF [8 , 11] . Abatacept ( Orencia; Bristol-Myers Squibb ) , a fusion protein of CTLA4-Ig , was used to modulate the costimulation of T-cells in mice [11 , 46] . Anakinra ( Kineret; Amgen ) , an IL-1 receptor antagonist , was used to block murine IL-1 [10 , 47] . Etanercept ( 0 . 2 mg/mouse in 0 . 1 mL of PBS ) or abatacept ( 0 . 5 mg/mouse in 0 . 1 mL of PBS ) was given subcutaneously ( s . c . ) the day before and the day after challenge with Lpl1 . Anakinra ( 1 mg/mouse in 0 . 1 mL of PBS ) was given s . c . daily , starting from day -1 until day +2 after the challenge . PBS served as a control . A potent selective inhibitor of iNOS , 1400W , was used to block murine iNOS [48] . Etanercept ( 0 . 1 mg/mouse in 0 . 1 mL of PBS ) or 1400W ( 0 . 25 mg/mouse in 0 . 1 mL of PBS ) was s . c . injected into NMRI mice twice per day , starting on day -1 until day +2 after challenge with coinjections of SA113Δlgt mutant with either Lpl1 or PBS . PBS treatment served as a control for both groups . Monoclonal anti-mouse CXCR2 antibody ( clone 242216; R&D Systems ) was used to block murine CXCR2 [49] . NMRI mice were i . p . injected with either anti-CXCR2 or an isotype control antibody ( clone 2A3; BioXCell ) ( 75–95 μg/mouse in 0 . 2 mL of PBS ) the day before challenge with coinjections of SA113Δlgt mutant with either Lpl1 or PBS . Knee joints were homogenized with an Ultra Turrax T25 homogenizer ( IKA , Staufen , Germany ) . Then , the homogenate was diluted in PBS , spread on horse blood agar plates , and incubated for 24 hours at 37°C . Viable counts of bacteria were performed and quantified as CFUs . SA113Δlgt mutant bacteria ( 103 CFU/mL ) were incubated with 25 μg/mL of Lpl1 , 100 μg/mL of Pam3CSK4 , or PBS control in tryptic soy broth ( TSB ) medium . At specific time intervals ( 1 , 3 , 6 , and 24 hours ) , samples of the bacterial mixtures ( 100 μl ) were spread on horse blood agar plates . After incubation for 18 hours at 37°C , colonies were counted . The effect of exogenous Lpl1 and Pam3CSK4 on S . aureus growth was evaluated by comparing the number of CFUs in the PBS control and the Lpl1- or Pam3CSK4-treated staphylococcal cultures at the different time points . Blood samples from healthy NMRI mice ( n = 4 ) were collected into heparin-containing tubes . SA113 or SA113Δlgt mutant bacterial suspensions were prepared and added into the mouse blood to a final concentration of approximately 1x103 CFU/mL . The incubated mixtures were shaken at 300 rpm for 2 hours at 37°C . To determine bacterial viability in blood , aliquots were withdrawn after 0 , 30 , 60 and 120 minutes of incubation , and samples were plated onto horse blood agar plates . Bacterial survival was evaluated as a percentage of number of CFUs at different time points compared with the number of bacteria initially added to the whole blood . Splenocytes from healthy NMRI mice and peritoneal macrophages from C57Bl/6 and TLR2-/- mice were prepared under sterile conditions . To prepare the splenocyte culture , mouse spleens were aseptically removed and passed through a nylon mesh . To collect the peritoneal macrophages , peritoneal lavage was performed using 10 mL of ice cold PBS . Erythrocytes were depleted in both cultures by lysis in 0 . 83% ammonium chloride , and the remaining cells were washed in PBS . The single-cell suspensions were then adjusted in Iscove’s complete medium ( 10% fetal calf serum , 2 mM L-glutamine , 5x10-5 M mercaptoethanol and 50 μg/mL gentamicin ) to a density of 1 . 5x106 cells/mL for splenocytes and 5x105 cells/mL for macrophages . For the proliferation assay , the splenocytes were stimulated with purified Lpl1 ( +sp ) or Lpl1 ( -sp ) ( 20–200 ng/mL ) , Pam2CSK4 or Pam3CSK4 ( 40 ng/mL ) , and TSST-1 ( 100 ng/mL ) or Iscove's medium ( negative control ) . A total of 1 μl of Ci [3H]thymidine ( Amersham , Bucks , UK ) was added for incorporation 12 hours before the cells were harvested , and the proliferative response was read with a micro-β counter . For cytokine/chemokine analysis , the macrophages were stimulated with purified Lpl1 ( +sp ) or Lpl1 ( -sp ) ( 0 . 02–0 . 2 μg/mL ) , Pam3CSK4 ( 2–20 ng/mL ) , LPS ( 1 μg/mL ) , or Iscove's medium for 4 hours . The supernatants were saved for later analysis . Splenocytes ( 5x106 cells/mL ) from healthy NMRI mice ( n = 4 ) were incubated with either 5x106 CFU/mL ( multiplicity of infection [MOI] = 1 ) or 25x106 CFU/mL ( MOI = 5 ) of SA113 or SA113Δlgt mutant bacteria in Iscove’s complete medium for 6 hours at 37°C . Aliquots were collected at 0 . 5 , 1 , 3 , and 6 hours of incubation . LDH was measured in the supernatants with a cytotoxicity detection kit ( Roche Diagnostics GmbH , Mannheim , Germany ) , according to the manufacturer’s directions . Absorbance was measured at 490 nm , and the results show the percentage of maximal LDH release in relation to positive control ( splenocytes treated with Triton X-100 ) . Peritoneal macrophages from NMRI mice were adjusted in Kreb’s Ringers glucose ( KRG ) ( 1x106 cells/mL ) and stimulated with purified Lpl1 ( +sp ) ( 0 . 2 μg/mL ) or PBS at 37°C for 1 hour . To study whether Lpl1 impacts phagocytosis , green fluorescent protein ( GFP ) -expressing S . aureus in KRG was incubated with or without 20% mouse serum at 37°C for 30 min , mixed with the stimulated peritoneal macrophages , incubated at 37°C for 20 min and analyzed as previously described [50] . The levels of IL-6 , MIP-2 , KC and TNFα in the supernatants from the knee joint homogenates and the levels of KC , MIP-2 , MCP-1 and TNFα in the supernatants from peritoneal macrophage or splenocyte stimulation were quantified using DuoSet ELISA Kits ( R&D Systems , Abingdon , UK ) . The knee joints were fixed in 4% formaldehyde for 3 days and then transferred to PBS . Imaging of the knee joints was performed ex vivo with a Skyscan1176 μCT scanner ( Bruker , Antwerp , Belgium ) to detect bone destruction after the studies were completed . The μCT settings were adjusted to a voxel size of 18 μm , an aluminum filter of 0 . 2 mm , and an exposure time of 800 ms , and the scans were conducted at 45 kV/555 μA . The X-ray projections were obtained at 0 . 42° intervals with a scanning angular rotation of 180° . The projection images were reconstructed into three-dimensional images using NRECON software ( version 1 . 6 . 9 . 8; Bruker ) and analyzed with CT-Analyzer ( version 2 . 7 . 0; Bruker ) . After reconstruction , experienced observers ( M . M . and T . J . ) evaluated the extent of bone and cartilage destruction in a blinded manner on a grading scale from 0–3 as previously described [11] . After the μCT scanning , the joints were decalcified , embedded in paraffin and sectioned with a microtome . Tissue sections were thereafter stained with hematoxylin and eosin . All slides were coded and assessed under a microscope in a blinded manner by two observers ( T . J . and M . M . ) . The extent of synovitis and cartilage-bone destruction was judged as previously described [12] . Statistical significance was assessed using the Mann-Whitney U test and Fisher's exact test , as appropriate . The results are reported as the mean ± standard error of the mean ( SEM ) unless indicated otherwise . A p value <0 . 05 was considered statistically significant . Calculations were performed using GraphPad Prism version 7 . 0b software for Macintosh ( GraphPad software , La Jolla , CA , USA ) . | Rapid bone destruction often leads to permanent joint dysfunction in septic arthritis , which is mainly caused by S . aureus . Despite advances in the use of antibiotics , permanent reductions in joint function occur in up to 50% of patients , who may also need joint replacement surgery . Additional challenge is posed by increasing antibiotic resistance of S . aureus , causing significant clinical and economic consequences . Although the outcome is poor , the current treatments for septic arthritis remain unchanged since many decades . It is largely unknown which bacterial factors cause aggressive joint damage . Here , we show that a single intra-articular injection of S . aureus lipoproteins ( Lpps ) into mouse knee joints induced chronic destructive macroscopic arthritis , and the monocytes/macrophages were the main culprit . However , coinjection of pathogenic S . aureus with Lpps into mouse knee joints attenuated the disease . The protective effect of Lpps was mediated by their lipid moiety , TLR2 on the host cells , neutrophil chemokine release , and consequent neutrophil recruitment . Our finding might be used as a novel concept in the treatment of multidrug-resistant bacterial infections . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] | [
"blood",
"cells",
"rheumatology",
"knee",
"joints",
"medicine",
"and",
"health",
"sciences",
"immune",
"cells",
"diagnostic",
"radiology",
"pathology",
"and",
"laboratory",
"medicine",
"pathogens",
"immunology",
"microbiology",
"staphylococcus",
"aureus",
"magnetic",
"resonance",
"imaging",
"bacteria",
"neutrophils",
"bacterial",
"pathogens",
"research",
"and",
"analysis",
"methods",
"musculoskeletal",
"system",
"white",
"blood",
"cells",
"imaging",
"techniques",
"skeletal",
"joints",
"animal",
"cells",
"staphylococcus",
"medical",
"microbiology",
"proteins",
"arthritis",
"microbial",
"pathogens",
"lipoproteins",
"biochemistry",
"radiology",
"and",
"imaging",
"diagnostic",
"medicine",
"anatomy",
"cell",
"biology",
"biology",
"and",
"life",
"sciences",
"cellular",
"types",
"macrophages",
"organisms"
] | 2019 | The YIN and YANG of lipoproteins in developing and preventing infectious arthritis by Staphylococcus aureus |
Chagas disease , caused by Trypanosoma cruzi , is endemic in Latin America and an emerging infectious disease in the US and Europe . We have shown TcG1 , TcG2 , and TcG4 antigens elicit protective immunity to T . cruzi in mice and dogs . Herein , we investigated antigenicity of the recombinant proteins in humans to determine their potential utility for the development of next generation diagnostics for screening of T . cruzi infection and Chagas disease . Sera samples from inhabitants of the endemic areas of Argentina-Bolivia and Mexico-Guatemala were analyzed in 1st-phase for anti-T . cruzi antibody response by traditional serology tests; and in 2nd-phase for antibody response to the recombinant antigens ( individually or mixed ) by an ELISA . We noted similar antibody response to candidate antigens in sera samples from inhabitants of Argentina and Mexico ( n = 175 ) . The IgG antibodies to TcG1 , TcG2 , and TcG4 ( individually ) and TcGmix were present in 62–71% , 65–78% and 72–82% , and 89–93% of the subjects , respectively , identified to be seropositive by traditional serology . Recombinant TcG1- ( 93 . 6% ) , TcG2- ( 96% ) , TcG4- ( 94 . 6% ) and TcGmix- ( 98% ) based ELISA exhibited significantly higher specificity compared to that noted for T . cruzi trypomastigote-based ELISA ( 77 . 8% ) in diagnosing T . cruzi-infection and avoiding cross-reactivity to Leishmania spp . No significant correlation was noted in the sera levels of antibody response and clinical severity of Chagas disease in seropositive subjects . Three candidate antigens were recognized by antibody response in chagasic patients from two distinct study sites and expressed in diverse strains of the circulating parasites . A multiplex ELISA detecting antibody response to three antigens was highly sensitive and specific in diagnosing T . cruzi infection in humans , suggesting that a diagnostic kit based on TcG1 , TcG2 and TcG4 recombinant proteins will be useful in diverse situations .
The protozoan parasite Trypanosoma cruzi , transmitted by blood-sucking triatomines , causes Chagas disease , which is a health threat for an estimated 10 million people , living mostly in Latin America . More than 25 million people are at risk of the disease [1] . Increasing travel and immigration has also brought T . cruzi infection into non-endemic countries , e . g . , the U . S . , Spain and Australia , where natural transmission is absent or very low . The congenital and transfusion- or organ transplantation-related transmissions are becoming recognized as significant threats in recent decades [2] , [3] . Diagnosis and treatment of T . cruzi infection has remained difficult and challenging after 100 years of its identification . This is because the acute infection , in general produces mild clinical symptoms , e . g . , fever , dyspnea , local swelling at the site of infection , that are infrequently reported [4] . As a result , acute exposure when detection of blood parasitemia and treatment is possible , remain largely unnoticed . Only those who develop severe acute myocarditis or when an outbreak of T . cruzi infection occurs may receive early diagnosis and therapeutic treatment [5][6] . In >95% of human cases , T . cruzi infection remains undiagnosed until several years later when chronic evolution of progressive disease results in clinical symptoms associated with cardiac damage . A conclusive diagnosis of T . cruzi infection then often requires multiple serological tests , in combination with epidemiological data and clinical symptoms . Unfortunately , after complicated diagnosis , no vaccines or therapies are available to treat the chronically infected individuals . We have , previously , employed an unbiased computational/bioinformatics approach for screening the T . cruzi sequence database and identification of potential vaccine candidates [7] . A strategic analysis of the sequence database led to selection of 71 candidates that were unique to T . cruzi . Of these , eight candidates ( TcG1 , TcG2 , TcG3 , TcG4 , TcG5 , TcG6 , TcG7 , and TcG8 ) were selected for their ability to induce agglutinating antibody response in mice [7] . Further studies indicated TcG1 , TcG2 , and TcG4 were maximally immunogenic because these were recognized by IgGs in infected mice and dogs ( reviewed in [8] ) , by CD8+ T cells in infected mice [9] , and elicited type I cytokines ( e . g . IFN-γ ) in mice and dogs [10] . Immunization with candidate antigens as a DNA vaccine provided a degree of protective immunity ( TcG2 = TcG4>TcG1 ) that substantially controlled acute parasitemia after challenge infection in mice and dogs [10] , [11] . The delivery of candidate antigens as heterologous prime/boost vaccines further enhanced the protective efficacy , evidenced by >80% control of acute parasitemia and tissue parasite burden and associated inflammation in heart and skeletal tissue [9] , [12] . In this study , we investigated the antibody response to the three candidate antigens ( TcG1 , TcG2 , TcG4 ) in clinically characterized chagasic patients . Our objectives were to evaluate the antigenicity of the candidate antigens in humans and determine their utility in generation of improved diagnostics for screening of T . cruzi infection and disease . Our data demonstrate that the candidate antigens are recognized by antibody responses in chagasic patients from two distinct study sites where diverse strains of the circulating parasites were reported . Further , a multiplex assay consisting of the mixture of the three antigens was highly sensitive and specific in diagnosing T . cruzi infection in human patients .
T . cruzi trypomastigotes ( SylvioX10/4 , TCI lineage ) were maintained and propagated by continuous in vitro passage in monolayers of C2C12 cells . Amastigotes were obtained by incubation of the freshly harvested trypomastigotes in RPMI-10% FBS medium , pH 5 . 0 at 37°C , 5% CO2 for 2 h . Human sera samples used in this study were obtained from Salta Argentina ( located at the border of Bolivia ) and Chiapas Mexico ( located on the border of Guatemala ) that are known to be endemic for T . cruzi transmission and human infection . All procedures were approved by the Institutional Review Boards at the University of Texas Medical Branch ( UTMB ) , Universidad Nacional de Salta ( UNSa ) , Argentina , and the Universidad Autónoma de Chiapas , Mexico . All sera samples obtained for the study were analyzed for T . cruzi-specific antibodies by commercially available kits . Those positive by ≥ two tests were considered seropositive ( +ve ) and those negative for these tests were considered as seronegative ( −ve ) . All samples were decoded and de-identified before they were provided for research purposes . Written informed consent was obtained from all individuals . In Argentina , all individuals were clinically characterized , and venous blood samples were collected to obtain plasma or serum . Clinical data included medical history , physical examination and subjective complaint of frequency and severity of exertional dyspnea , electrocardiography ( 12-lead at rest and 3-lead with exercise ) to reveal cardiac rhythm and conduction abnormalities and transthoracic echocardiogram to obtain objective information regarding the left ventricular ( LV ) contractile function . Chest X-ray was used to assess cardiomegaly ( cardio-thoracic ratio>0 . 5 ) . Seronegative and physically healthy subjects exhibiting no history or clinical symptoms of heart disease were used as controls . Seropositive chagasic patients were classified according to clinical exam as follows: Group 0: no echocardiography abnormalities , no left ventricular dilatations , and ≥70% ejection fraction ( EF ) indicating preserved systolic function , Group I: negligible to minor EKG alterations , EF: 55–70% , no indication of heart involvement; Group II: a degree of heart involvement with systolic dysfunction ( EF: 40–55% ) ; and Group III: moderate to severe systolic dysfunction ( EF≤40% ) , left ventricular dilatation ( diastolic diameter ≥57 mm ) , and/or potential signs of congestive heart failure . In Mexico , human sera samples were collected within the framework of a research project on emerging zoonotic diseases conducted jointly by several institutions , including Chiapas State University ( UNACH ) , Mexican Social Security Institute ( IMSS ) , Chiapas Health Institute ( ISECH ) and UTMB at Galveston [13] . The seropositive subjects generally represented indeterminate/asymptomatic form of disease . Patients' detailed information is presented in Table 1 . All sera samples from Salta Argentina were analyzed for T . cruzi-specific antibodies by the personnel of the Clinical laboratories at the San Bernardo Hospital using a Wiener Chagatest-ELISA recombinant v . 4 . 0 kit ( cut-off absorbance at 450 nm: average of seronegative samples ( <0 . 1 ) +0 . 2 . , i . e . ≥0 . 3 ) . Serological tests were also done following the specifications of the commercial IHA test kit ( Wiener Chagatest-HAI , positive ≥1∶16 dilution ) . Those positive by both tests were considered seropositive , and exposed to T . cruzi infection . All sera samples from Chiapas Mexico were screened by epimastigote antigenic lysate-based ELISA , trypomastigote-based flow cytometry , and Chagas Stat-Pak immuno-chromatograpic test ( Chembio Diagnostic Systems , Medford NY ) . Those positive by at least two tests were considered seropositive [13] . Based upon the above analysis , we created sera pools for 2nd phase screening . Briefly , from Chiapas Mexico , a seropositive pool ( n = 65 ) and a seronegative pool ( n = 34 ) was included in the 2nd phase . From Salta Argentina , sera and plasma samples collected in year 2009 ( seropositive , n = 65; seronegative , n = 20 ) and year 2010 ( seropositive , n = 45; seronegative , n = 20 ) were included . All blood samples positive for T . cruzi-specific antibodies from Salta Argentina were analyzed for Leishmania infection by two PCR approaches . Briefly , total DNA was extracted from blood samples with a phenol-chloroform mixture , precipitated with ethanol and dissolved in 20 µl of TE buffer . A PCR reaction was performed using 1 µl isolated DNA with primer pairs ( 5′-GTGGGGGAGGGGCGT-TCT-3′ and 5′-ATTTTACACCAACCCCCAGTT-3 ) that specifically amplify 120 base pair product from the conserved region of Leishmania kDNA [14] . The polymorphism-specific PCR ( PS-PCR ) allows the identification of Leishmania species from Argentina , and was performed using the primer pairs as previously described [15] . To evaluate the specificity of the TcG1- , TcG2- , and TcG4-based diagnostic ELISA , sera samples from volunteer donors from Galveston TX , Buenos Aires Argentina , and Toluca Mexico with no history of residence in the endemic areas , and were healthy ( n = 42 , true negative controls ) or exhibited cardiomyopathy of other etiologies ( n = 20 ) were included in the study . Seronegative/cardiomyopathy patients were identified based clinical exam , and blood levels of NT-proBNP to be >2000 pg/ml ( normal<450 pg/ml ) . To examine the cross-reactivity of antigen-based assay , sera samples from subjects living in Salta Argentina , and diagnosed for Leishmania infection ( n = 35 ) or certain autoimmune diseases ( n = 15 ) were also included in the study . Patients were diagnosed for leishmaniasis ( cutaneous , mucocutaneous or visceral ) based upon three criteria ( smears and lesions , Montenegro reaction , epidemiology and clinical demonstration ) as described in [15] , and further identified to be positive for Leishmania infection by kDNA-specific PCR and PS-PCR approaches , as above . Briefly , parasitological analysis was done on smears of dermal scrapings stained with May-Grunewald Giemsa and examined under a microscope . For Montenegro skin test , promastigote protein lysates ( 4 µg/100 µl ) of Leishmania braziliensis were injected intradermally on the ventral forearm of the patients , and indurations ≥5 mm in diameter , observed 48 h after the injections were considered reactive . Clinical features included the presence of compatible tegumentary injuries with ulcerative , nodulous , papulous cutaneous or mucocutaneous lesions and a congruent epidemiological history . Patients' information is presented in Table 1 . Sera samples were analyzed for IgG antibody levels by using TcG1- , TcG2- , TcG4-based ELISA . Recombinant TcG1 , TcG2 and TcG4 proteins were purified from E . coli . The nucleotide sequences of TcG1 , TcG2 and TcG4 antigens have previously been submitted to GenBank under accession numbers AY727914 , AY727915 and AY727917 , respectively [7] . The cDNAs encoding TcG1 ( 1–166 amino acids ) , TcG2 ( 1–220 amino acids ) , and TcG4 ( 1–92 amino acids ) were cloned in pCR2 . 1 T/A vector , and then sub-cloned in to pET-22b plasmid ( Novagen , Gibbstown , NJ ) such that the encoded proteins were in-frame with a C-terminal polyhistidine tag . All cloned sequences were confirmed by restriction digestion and sequencing at the Recombinant DNA Core Facility at UTMB . The pET22b plasmids containing TcG1 , TcG2 or TcG4 were transformed in BL21 ( DE3 ) pLysS competent cells ( Invitrogen , Carlsbad CA ) and recombinant proteins purified using the polyhistidine fusion peptide metal chelation chromatography system ( Novagen ) [9] . After purification , proteins were dialyzed to remove contaminant particles , and endotoxins , and stored at −80°C till further use . Culture-derived parasites ( 70% trypomastigotes/30% amastigotes ) were lysed by repeated freeze-thaw in PBS ( 109/ml ) and used as a source of T . cruzi total antigen for positive control [10] . Flat bottom , high-binding , 96-well plates ( Greiner bio-one ) were coated overnight at 37°C with 100-µl/well of recombinant antigens ( 0 . 5 µg each antigen/well , individually or in combination ) or T . cruzi total lysate ( 5×105 parasite equivalents/well ) . Plates were blocked for 2 h at 37°C with 200-µl/well of 5% non-fat dry milk ( NFDM ) in PBS , washed with PBS-0 . 05%Tween-20 ( PBST ) twice , and PBS once , and then incubated for 2 h with test sera ( 100-µl/well ) added in triplicate in 2-fold dilutions . Plates were then washed and incubated at room temperature for 30 min with 100-µl/well of horseradish peroxidase-labeled goat anti-human IgG ( 1∶5000 dilution in PBST-1% NFDM ) , and color was developed for 5 min with 100 µl/well of Sure Blue TMB substrate ( Kirkegaard & Perry Laboratories ) . The colorimetric reaction was stopped with 2N H2SO4 , and absorbance measured at 450 nm using SpectraMax M5 Microplate Reader ( Molecular Devices ) . The sensitivity of the antigen-based ELISA ( 2nd phase ) was determined by calculating the percentage of chagasic samples that exhibited reactivity with recombinant proteins out of the total samples previously categorized as seropositive based on 1st phase screening with commercially available kits . The data from antigen-based ELISA analysis of sera samples from healthy individuals , other cardiomyopathy and leshmaniasis patients was utilized to calculate the specificity of the assay as follows: [ number of sera samples analyzed – number of sera samples that exhibited false positive reaction or cross-reactivity with TcG1 , TcG2 or TcG4/number of sera samples analyzed] ×100 . All samples were analyzed in duplicate and assayed at least twice for all experiments . Box plots and dot plots were made using Sigma Plot 12 . 0 and GraphPad Prism 5 software , respectively , and statistical analysis was conducted using SPSS v . 18 software . Results were analyzed using Student's t test for statistical evaluation of mean values for experimental and control samples , and the level of significance was taken at α<0 . 05 . Pearson's correlation analysis was performed to determine the relationship between predictive efficacies of the antibody response to selected antigens in diagnosis of disease severity . Seropositivity rates for anti-T . cruzi antibodies in different tests , and their confidence intervals [CIs] , were calculated using the mid-P 95% confidence interval ( 95% CI ) using Epi Info ( version 6 . 0 ) software .
To proceed with sample analysis , we optimized ELISA components by cross-titration , using a pool of known positive and negative controls ( 1∶20–1∶1600 dilutions ) . The optimal sera and HRP-conjugated secondary antibody dilutions providing maximum signal-to-noise ratio were determined to be 1∶50 and 1∶5 , 000 , respectively , and used in all further investigations . The variations in reactivity of negative and positive sera among different assays and plates of the same experiment ranged from 3–12% . We , first , monitored the antigenicity of TcG1 , TcG2 , and TcG4 using sera samples collected from volunteers enrolled in the study in Argentina in year 2010 . Samples were stored at −80°C just after collection , and thawed when utilized . The negative sera samples ( n = 20 ) from the endemic area near Argentina-Bolivia border exhibited low reactivity for TcG1 , TcG2 , and TcG4 , similar to that noted for confirmed negative controls ( n = 42 ) from non-endemic areas ( TcG1: 0 . 216±0 . 035 versus 0 . 211±0 . 048 , TcG2: 0 . 240±0 . 04 versus 0 . 230±0 . 044 , TcG4: 0 . 225±0 . 041 versus 0 . 252±0 . 038 , expressed as mean absorbance ± SD ) . In comparison , a 4-fold , 2 . 75-fold , and 2 . 65-fold increase in sera levels of antibody response to TcG1 , TcG2 , and TcG4 , respectively , was noted in previously characterized seropositive subjects ( n = 45 ) from Argentina-Bolivia border ( TcG1: 0 . 81±0 . 33 , TcG2: 0 . 66±0 . 20 , TcG4: 0 . 57±0 . 09 , expressed as mean absorbance ± SD , p<0 . 001 for all , Fig . 1 . A–C ) . The sera levels of antibodies to TcG1 , TcG2 , and TcG4 were above the meanseronegative level in 62 . 2% , 66 . 6% and 75 . 5% of the 1st-phase seropositive subjects . When analyzing plasma samples from the same individuals , we noted a 3 . 34-fold , 2 . 4-fold and 2 . 3-fold increase in plasma levels of antibodies to TcG1 , TcG2 and TcG4 in seropositive subjects as compared to seronegative controls ( TcG1: 0 . 77±0 . 21 versus 0 . 20±0 . 05 , TcG2: 0 . 65±0 . 12 versus 0 . 21±0 . 057 , TcG4: 0 . 67±0 . 09 versus 0 . 20±0 . 048 , expressed as mean absorbance ± SD , p<0 . 001 for all , Fig . 1 . D–F ) . The plasma levels of antibodies to TcG1 , TcG2 , and TcG4 were above the meanseronegative levels in 71 . 1% , 77 . 7% and 80% of the 1st-phase seropositive subjects . These data suggested that a ) TcG1 , TcG2 and TcG4 are recognized by antibody responses elicited in human patients infected by T . cruzi , and b ) both plasma and sera samples can be utilized to monitor the antibody response . We then analyzed the plasma samples that were collected in 2009 , and characterized as seropositive ( n = 65 ) and seronegative ( n = 20 ) by 1st-phase serology tests . These samples were subjected to two cycles of freezing/thawing during the two-year storage . Our data showed a 6 . 72-fold , 2 . 4-fold and 2 . 9-fold increase in plasma levels of antibodies to TcG1 , TcG2 and TcG4 in seropositive subjects as compared to seronegative controls ( TcG1: 1 . 66±0 . 55 versus 0 . 204±0 . 05 , TcG2: 0 . 69±0 . 13 versus 0 . 239±0 . 039 , TcG4: 0 . 75±0 . 20 versus 0 . 227±0 . 05 , expressed as mean absorbance ± SD , p<0 . 001 for all , Fig . 1 . G–I ) . The plasma levels of antibodies to TcG1 , TcG2 , and TcG4 were above the meanseronegative level in 61 . 5% , 64 . 6% and 81 . 5% of the seropositive subjects . These results suggest that antibody response to TcG1 , TcG2 , and TcG4 is stable , and field samples can be utilized to examine antigenicity of the selected candidates in large-scale population studies . Overall , the data presented in Fig . 1 also indicate that TcG1 , TcG2 and TcG4 are expressed by T . cruzi isolates circulating in the endemic areas at the Argentina-Bolivia border . It is important to know if antibody recognition of the three antigens can be expanded for diagnosis of T . cruzi infection in other countries where different isolates are suggested to be present in domestic and sylvatic cycle of parasite circulation . For example , in Argentina and neighboring countries in South America , TCII isolates are predominantly identified in peripheral blood of seropositive patients , though molecular studies have revealed the presence of TCI parasites also in heart biopsies of chronic chagasic patients [16] , [17] , [18] . In Mexico and Guatemala , TCI is dominantly found in epidemiological evaluation of infected triatomines as well as in blood samples from acute and chronic chagasic cardiomyopathy patients [19] , [20] . We , therefore , monitored the antigenicity of TcG1 , TcG2 , and TcG4 in human sera samples from Mexico-Guatemala border area that were characterized as seropositive ( n = 65 ) and seronegative ( n = 34 ) by 1st-phase serology tests . Our data showed 6 . 72-fold , 2 . 4-fold and 2 . 9-fold increase in sera levels of antibodies to TcG1 , TcG2 and TcG4 in seropositive samples as compared to that noted in seronegative healthy controls ( TcG1: 0 . 7±0 . 25 versus 0 . 211±0 . 047 , TcG2: 0 . 64±0 . 22 versus 0 . 25±0 . 05 , TcG4: 0 . 65±0 . 20 versus 0 . 212±0 . 048 , expressed as mean absorbance ± SD , p<0 . 001 for all , seropositive versus seronegative , Fig . 2 ) . The sera levels of antibodies to TcG1 , TcG2 , and TcG4 were above the meanseronegative level in 69 . 2% , 76 . 9% and 72 . 3% of the seropositive subjects from Mexico . These data demonstrate that TcG1 , TcG2 and TcG4 are antigenic , and recognized by antibody responses in chagasic patients from Mexico , and suggest that the three antigens are expressed by T . cruzi isolates circulating in Mexico-Guatemala border area . Next , we determined if the three candidate antigens can be utilized together to improve the diagnosis of exposure to T . cruzi . For this , we coated the 96-well plates with either the mixture of TcG1 , TcG2 and TcG4 ( 0 . 5 µg/well each ) or T . cruzi trypomastigote lysate ( TcTL , 2×105 parasite equivalent ) , and monitored the antibody response by ELISA under similar experimental conditions . When antibody response was captured using the TcGmix , the seronegative controls from the non-endemic areas exhibited a mean absorbance ± SD of 0 . 225±0 . 039 . Using the controls' mean absorbance+2SD , our data validated 40 of the 45 sera samples ( 88 . 8% ) from Argentina , characterized as seropositive in 1st-phase screening in the year 2010 , were seropositive for TcGmix-specific antibodies ( mean absorbance ± SD: 0 . 73±0 . 17 , maximum OD: 1 . 2 , Fig . 3A ) . One of the volunteer previously characterized as seronegative exhibited anti-TcGmix antibody response above the meanseronegative level . Similarly , the plasma detection of antibody response to TcGmix identified 102/110 of the seropositive subjects ( 92 . 7% ) identified in 1st-phase screening in the years 2009 and 2010 ( Fig . 3B , C ) . In comparison , when plates were coated with TcTL to capture anti-T . cruzi antibodies , the seronegative true controls from non-endemic areas , exhibited a mean absorbance ± SD value of 0 . 233±0 . 044 ( Fig . 3C , E ) . Using the mean absorbance for controls +2 SD as a cut off , our data validated 97 . 7–100% sera and plasma ( year 2010 ) and 96 . 9% plasma ( year 2009 ) samples characterized as seropositive in 1st-phase screening in Argentina were also positive for TcTL-specific antibodies; the mean absorbance ± SD for the positive population was 1 . 1±0 . 6 ( year 2009 ) and 0 . 73±0 . 08 ( year 2010 ) with the highest value being 2 . 5 and 0 . 98 , respectively ( Fig . 3B , D , F ) . One of the volunteer previously characterized seronegative exhibited anti-TcTL antibody response . To validate that diagnostic potential of the TcGmix based ELISA is not restricted to samples from Argentina , we monitored the antibody response using sera samples collected in Mexico . Of the 65 samples characterized as seropositive in 1st-phase screening , 58 ( 89 . 2% ) and 63 ( 96 . 9% ) exhibited reactivity when plates were coated with TcGmix and TcTL antigens , respectively ( Fig . 4 ) . The mean absorbance ± SD for the antibody response to TcGmix and TcTL in the positive population positive population was 0 . 75±0 . 14 ( max: 1 . 2 ) and 0 . 87±0 . 4 ( max: 2 . 6 ) , respectively , the difference between the two values being observed non-significant . No significant difference was observed when either plasma or sera were used as the source of antibodies . Five of the 110 seropositive/chagasic subjects from Argentina exhibited reactivity for Leishmaina-specific antibodies and were likely infected with both pathogens . To examine if the antigen-based ELISA is specific for T . cruzi detection , we performed recombinant antigens-specific ELISA using the sera samples from non-chagasic individuals ( total 112 ) , including leishmaniasis patients ( n = 35 ) , volunteer donors with cardiomyopathy ( n = 20 ) and autoimmune diseases ( n = 15 ) of other etiologies , and healthy donors from non-endemic areas ( n = 42 ) , and calculated the specificity of the antigen-based ELISA . Sera samples from leishmaniasis patients ( n = 35 ) exhibited very low reactivity for TcG1 , TcG2 , and TcG4 , similar to that noted for sera samples from confirmed negative controls from non-endemic areas ( TcG1: 0 . 18±0 . 04 versus 0 . 21±0 . 048 , TcG2: 0 . 15±0 . 06 versus 0 . 230±0 . 044 , TcG4: 0 . 232±0 . 05 versus 0 . 252±0 . 038 , expressed as mean absorbance ± SD ) . Likewise , sera samples from patients exhibiting symptoms of cardiomyopathy of non-chagasic etiology or autoimmune diseases showed reactivity to TcG1 , TcG2 , and TcG4 below the cut-off threshold values derived from healthy/seronegative controls . The specificity for TcGmix was highest ( 98% ) , followed by TcG2 ( 96% ) , TcG4 ( 94 . 6% ) , TcG1 ( 93 . 6% ) and TcTL ( 77 . 8% ) , determined by detection of false positive signal for 2 , 4 , 6 , 7 and 25 of the samples , respectively , out of the total 112 samples from non-chagasic individuals that were submitted for antigen- and TcTL-based ELISA . These data indicated that the recombinant antigens were highly specific for the detection of anti-T . cruzi antibodies than the whole parasite ( trypomastigote ) lysate , and exhibited no cross-reactivity to Leishmania-specific antibodies . Pearson correlation analysis was employed to identify correlation between antigen-specific antibody response and disease state including the data derived from seronegative/healthy controls and seropositive/chagasic subjects . For this , seronegative/healthy subjects were labeled as 0 , and patients classified as 0 , I , II and III ( Materials and Methods ) were labeled as 1 , 2 , 3 , and 4 , respectively . Antibody response was titrated using 2-fold sera dilutions ( 1∶50–1∶1600 ) . We observed no significant correlation between the anti-TcG2 , anti-TcG4 , anti-TcGmix and anti-TcTL antibody titration curves and clinical disease category in any of the patient population ( data not shown ) . The representative correlation data from sera levels of anti-TcGmix and anti-TcTL antibody response ( 1∶50 dilutions ) and clinical disease category for the clinically characterized Argentine patients enrolled in the study is shown in Fig . 5 . It is worth noting that TcGmix-specific antibodies exhibited a clear downward trend with patients' disease severity , indicating that presence of antibodies for TcG1 , TcG2 , and TcG4 is protective during progressive Chagas disease ( Fig . 5A ) . No clear trend or correlation was observed for TcTL specific antibody response and disease severity in any of the patient population ( Fig . 5B ) .
Serological diagnosis of Chagas disease is frequently based on tests such as enzyme-linked immunoassays ( EIAs ) , indirect immunofluorescence assays , and indirect haemagglutination assays ( IHAs ) . Most of these serological tests , including one recently licensed by the United States Food and Drug Administration for use as a blood screening test in the U . S . [21] , use crude or semi-purified T . cruzi epimastigote forms as the antigen , the stage that is present in insects but not in infected humans . Overall , the current diagnostics fail to provide a high degree of sensitivity and specificity , requiring use of multiple tests for diagnosis of T . cruzi infection [22] . Further , most of the currently available kits produce questionable results when used for donors with low titers [23] . The absence of a true gold standard makes it difficult for the medical practitioners to provide proper treatment as not all cases are properly identified and treatment response cannot be accurately monitored . Accordingly , World Health Organization has emphasized the need to employ defined antigens for improved serodiagnosis of T . cruzi infection [24] . However , before an antigen can be used for diagnostics , several criteria should be met: ( i ) the selected candidate antigen ( s ) should be expressed in isolates circulating in different areas of endemicity , ( ii ) antigens should be highly immunogenic in populations with different genetic background , ( iii ) they should be absent from other pathogens to prevent cross-reactivity , and ( iv ) easily expressed and purified from traditional protein expression systems ( e . g . E . coli ) to ensure reproducibility and quality control [25] . Because of the complexity involved in antigen selection and testing , limited progress has been made towards the development of antigen-based kits for the diagnosis of T . cruzi infection . Keeping the above guidelines in mind , we believe that TcG1 , TcG2 and TcG4 are ideal candidates for the development of diagnostic assays . One , these antigens were previously shown to be expressed in infective and intracellular forms of clinically relevant multiple isolates of T . cruzi [7] . Two , the small size of the three antigens ( TcG1: 18 . 4 kDa , TcG2: 24 kDa , TcG4: 10 kDa ) allows reproducible high-yield purification of the proteins from E . coli ( 0 . 5–1 . 0 g/L ) and is amenable to large-scale production . Three , the presence of multiple 12-mer B cell epitopes of high specificity ( Table 2 ) suggests that these antigens will be immunogenic , recognized by B cells of the immune system , and elicit antibody response . Indeed , we have found that TcG1 , TcG2 and TcG4 are recognized by antibodies elicited in infected dogs and mice [9] , [11] . The diagnostic potential of the three antigens in humans was evident from our analysis of a panel of sera or plasma from chronic chagasic patients from Argentina-Bolivia and Mexico-Guatemala border areas . T . cruzi strains of lineages TCV ( IId ) , TCII ( IIb ) and TCI were identified in seropositive chagasic patients from Argentina/Bolivia ( Patricio Diosque and Monge Rumi , personal communication ) , and of lineage TCI in seropositive subjects from Mexico/Guatemala ( unpublished results ) . Others have documented the predominance of TCII isolates in peripheral blood of seropositive patients from South America [16] , [17] , [18] and TCI in Mexico and Guatemala [19] , [20] . Thus , our data presented in Figs . 1 , 2 , 3 , 4 suggest the diagnostic kit developed using TcG1 , TcG2 and TcG4 will be useful in identifying T . cruzi circulation and transmission in South , Central and North America . It is important to note that TcG1 , TcG2 and TcG4 , when used individually , exhibited 93 . 6–96% specificity in detecting anti-T . cruzi antibody response that was in the range for other recombinant proteins ( e . g . TSSA , CP1 and CP2 ) proposed for Chagas disease immunodiagnosis [26] , [27] . However , 98% specificity of the TcGmix-based ELISA was much higher than that observed for single antigens in this study or other recombinant antigens in other reports [26] , [28] . Further , TcGmix specificity at 98% was significantly better than that observed with T . cruzi trypomastigote-based ELISA ( 77 . 8% ) that is not desirable because culturing of human-infective form of the pathogen requires special facilities , technology , and expertise . Another major concern that is generally not taken into consideration when one is using serological tests for Chagas disease is the potential frequency of cross-reactivity . In some areas of endemicity in Central America and Brazil , where T . cruzi and the nonpathogenic protozoan Trypanosoma rangeli can be found infecting the same vectors and vertebrate hosts , cross-reactivity has been proposed to contribute to miscalculated higher percentage of T . cruzi transmission and misdiagnosis of patients may have severe health-related and economic consequences [29] , [30] . Others have documented the crude antigen-based serodiagnostic kits exhibit cross-reactivity between sera of patients infected with T . cruzi and sera of patients infected with Leishmania spp . [31] , [32] , [33] . In our study , homology searches suggested that only TcG1 is similar to a ribosomal protein of other trypanosomes ( >90% identical ) while TcG2 and TcG4 exhibited no clear paralog in the public databases ( Fig . S1 ) . Thus , to rule out the possibility of cross-reactivity with Leishmania , we have screened sera samples from confirmed leishmaniasis patients for antibody cross-reactivity to TcG1 , TcG2 and TcG4 . Our data clearly showed that the three antigens exhibited no cross-reactivity to antibodies generated in cutaneous , mucocutaneous and visceral leishmaniasis patients , providing evidence that the antibody response to TcG1 , TcG2 and TcG4 was solely directed by exposure to T . cruzi infection . Finally , TcGmix-ELISA performed at a higher potency in discriminating weakly positive samples from background , demonstrated statistically by calculating the quotient of measured optical density values divided by the cutoff values . Only 12 of all the seropositive samples had a quotient smaller than 2 . 0 in the TcGmix-ELISA whereas TcG1- , TcG2- TcG4 and TcTL-based ELISA resulted in 18 , 16 , 14 and 16 samples below the mean value . This potency will facilitate diagnosis because weakly positive results routinely need to be confirmed by alternative assays . Thus , based upon high sensitivity ( 93% ) and specificity ( 98% ) to sera from chagasic patients from different endemic countries , we surmise that TcGmix-based ELISA can serve as a single assay to determine the T . cruzi status of a given blood sample , and diagnose Chagas disease . We plan to further standardize the assay for large-scale screening and establishing the prototype assay for commercial use of the selected antigens for diagnostic kits . The immune interactions necessary to eradicate T . cruzi parasites are extremely complex and require both humoral and cell-mediated components of the immune system . Previous experimental studies and a few studies in humans indicate that antibodies are able to kill T . cruzi in the presence of phagocytic cells , such as macrophages . Others have shown in B cell knockout mice or mice depleted of B cell function , the pivotal role of antibody molecules in attaining resistance to T . cruzi infection and Chagas disease ( reviewed in [34] ) . In this study , the mean antibody response to TcG1 , TcG2 and TcG4 ( individually or in combination ) exhibited a downward trend in correlation to clinical disease severity suggesting that antibody response to the candidate antigens was protective , though further studies with larger patient population will be required to validate the significance of this observation . The experimental data indicating the up regulation of IgG1 antibodies specific to TcG1 , TcG2 and TcG4 [9] , [11] suggest that isotypes related to up regulation of opsonization , cell dependent cytotoxicity and activation of classical complement pathway are elicited by the candidate antigens , and might be present in significant levels in seropositive/chagasic individuals . Considering that the three antigens have been shown to be located on plasma membrane of trypomastigote/amastigote stages of T . cruzi by flow cytometry [7] , we predict that these will be available as a target to antibody-dependent cell cytotoxicity effector mechanisms mediated by IgG1 . Indeed , TcG1 , TcG2 and TcG4 have been shown to elicit potent trypanolytic antibodies in accordance with the intensity of the surface expression of these antigens in infective and intracellular stages ( as compared to eight others tested in similar experiments ) [7] . In addition to IgG1 , candidate antigens elicited IgG2b known to drive the type 1 adaptive immunity in experimental mice and dogs [9] , [11] . High levels of candidate antigens-induced IgG2b in vaccinated mice were linked to complement-dependent trypanolytic activity [9] . Patients in the indeterminate phase display higher levels of lytic antibodies compared with patients with chagasic heart disease indicating an association between the presence of lytic antibodies and a protective response in chronic patients [35] , [36] . This was also evidenced by absence of lytic antibodies in patients treated with anti-parasite drugs that displayed negative hemocultures for over ten years [37] . Based upon these observations , we consider that activation of IgG2a ( along with IgG1 ) will be an important attribute that contributes to host protection against T . cruzi infection and Chagas disease when TcG1 , TcG2 and TcG4 will be tested as vaccine candidates in humans . In summary , this is the first report identifying the antigenicity of TcG1 , TcG2 and TcG4 in humans . We have concluded that the three candidate antigens are recognized by IgGs in sera samples of seropositive/chagasic patients . Further , the candidate antigens were represented in diverse strains of T . cruzi as the sera samples from two different endemic areas from the South , Central and North America recognized the antigens at the same rate . Our data show the serology test developed using the TcGmix is a significantly better alternative to epimastigote extracts currently used in T . cruzi serodiagnosis or the trypomastigote lysate used in this study for comparison purposes . TcGmix ELISA was >98% specific and 93% sensitive , was not discriminatory of sex , age or geographic location of the individuals , performed at a higher potency in identifying weakly positive samples , and , thus , has a potential to serve as a single assay for the diagnosis of T . cruzi infection . Future studies with large cohorts of patients will be required to determine if immunological responsiveness to the three antigens detects with high confidence the chronicity of infection , severity of disease , and effectiveness of treatment etc . | Chagas disease is the most common cause of congestive heart failure related deaths among young adults in the endemic areas of South and Central America and Mexico . Diagnosis and treatment of T . cruzi infection has remained difficult and challenging after 100 years of its identification . In >95% of human cases , T . cruzi infection remains undiagnosed until several years later when chronic evolution of progressive disease results in clinical symptoms associated with cardiac damage . Diagnosis generally depends on the measurement of T . cruzi–specific antibodies that can result in false positives . A conclusive diagnosis of T . cruzi infection thus often requires multiple serological tests , in combination with epidemiological data and clinical symptoms . In this study , we investigated the antibody response to TcG1 , TcG2 , and TcG4 in clinically characterized chagasic patients . These antigens were identified as vaccine candidates and shown to elicit protective immunity to T . cruzi and Chagas disease in experimental animals . Our data show the serology test developed using the TcGmix ( multiplex ELISA ) is a significantly better alternative to epimastigote extracts currently used in T . cruzi serodiagnosis or the trypomastigote lysate used in this study for comparison purposes . | [
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] | [
"medicine",
"infectious",
"diseases",
"public",
"health",
"and",
"epidemiology",
"biomarker",
"epidemiology",
"clinical",
"laboratory",
"sciences",
"diagnostic",
"medicine",
"chagas",
"disease",
"epidemiology",
"neglected",
"tropical",
"diseases",
"epidemiological",
"methods"
] | 2013 | Antigenicity and Diagnostic Potential of Vaccine Candidates in Human Chagas Disease |
An important follow-up step after genetic markers are found to be associated with a disease outcome is a more detailed analysis investigating how the implicated gene or chromosomal region and an established environment risk factor interact to influence the disease risk . The standard approach to this study of gene–environment interaction considers one genetic marker at a time and therefore could misrepresent and underestimate the genetic contribution to the joint effect when one or more functional loci , some of which might not be genotyped , exist in the region and interact with the environment risk factor in a complex way . We develop a more global approach based on a Bayesian model that uses a latent genetic profile variable to capture all of the genetic variation in the entire targeted region and allows the environment effect to vary across different genetic profile categories . We also propose a resampling-based test derived from the developed Bayesian model for the detection of gene–environment interaction . Using data collected in the Environment and Genetics in Lung Cancer Etiology ( EAGLE ) study , we apply the Bayesian model to evaluate the joint effect of smoking intensity and genetic variants in the 15q25 . 1 region , which contains a cluster of nicotinic acetylcholine receptor genes and has been shown to be associated with both lung cancer and smoking behavior . We find evidence for gene–environment interaction ( P-value = 0 . 016 ) , with the smoking effect appearing to be stronger in subjects with a genetic profile associated with a higher lung cancer risk; the conventional test of gene–environment interaction based on the single-marker approach is far from significant .
Genome-wide association studies that focus on detecting the main effect from individual single nucleotide polymorphisms ( SNPs ) have successfully identified more than 4 , 000 SNPs associated with different diseases [1] . To achieve a better understanding of the mechanisms underlying disease development , it is of great interest to follow up those genetic findings with more detailed analyses investigating how the gene and environment interact in their influence on disease risk . One popular approach aims at detecting SNP-environment interaction between individual SNPs and established environmental risk factors [2] , [3] , [4] . One of the few successes for this approach is the interaction detected between cigarette smoking and two genetic variants , a NAT2 tagging SNP and a GSTM1 deletion , in a multi-stage genome-wide association study ( GWAS ) of bladder cancer [3] . The standard approach to the study of gene–environment joint effect inspects one marker at a time , assuming that a single marker is the functional unit in the gene and environment interplay . This single-marker approach could misrepresent and underestimate the genetic contribution to the joint effect when one or more functional loci , some of which might not be genotyped , exist in the region , and interact with the environment risk factor in a complex way . A more global approach that simultaneously considers all genetic markers might capture more of the genetic variation within the entire targeted region , and provides a better opportunity to reveal complicated gene–environment interactions [5] . The global approach would be more informative if it has the capability showing how an environmental effect varies according to a subject's genetic profile . We provide a flexible Bayesian modeling framework for the study of gene–environment joint effects . We consider a case-control study with genotypes G at a set of SNPs within a given region and a measurement for the environment exposure E available for each subject . We seek to identify a latent genetic profile variable L that classifies the multilocus genotype G into different categories ( clusters ) such that subjects with their genotype assigned to the same genetic profile category share the same disease risk model , which is a standard logistic regression model with its own intercept term and slope . The intercept term represents the baseline log odds , common for subjects sharing the same genetic profile . The slope represents the effect ( i . e . , log odds ratio ) of the environment risk factor for subjects with the given genetic profile . The model that we try to build and make inferences from is essentially the logistic regression model consisting of L and E as main effects and their product as an interaction term; the unusual aspect is that the definition of the latent genetic profile L is a priori unknown . To account for the uncertainty in the cluster assignment underlying the definition of L , we adopt an idea from the hidden Markov model originally developed for modeling the spatial heterogeneity of the disease event rate , observed on a predefined set of areas [6] . In this Bayesian model approach , Green and Richardson tried to allocate areas into a number of clusters and assumed a common disease rate for areas assigned to the same cluster . The mechanism for the area allocation was modeled through the Potts model [7] , which favors probabilistically those allocation patterns where “neighboring” areas are assigned to the same cluster . Note that the spatial dependence assumption is generally appropriate in situations where the event rate is expected to take on similar values in neighboring areas . To draw the connection , we can think of each type of observed multi-locus genotype G as an “area” . We would like to use the Potts model to guide the cluster assignment through a certain level of “spatial” dependence , i . e . , similar genotypes ( nearby areas ) tend to be assigned to the same cluster , as in other applications in genetics studies , including the study of haplotype association [8] , [9] . We use the Markov chain Monte Carlo ( MCMC ) sampling method ( e . g . , [10] , [11] ) to fit the proposed model , incorporating several recent advances in the MCMC methodology . We adopt a recently developed algorithm [12] to update the regulating parameter in the Potts model , which has an intractable normalizing constant , and cannot be handled by the standard Metropolis Hastings algorithm . This algorithm allows us to consider the parameter of interest on its original continuous scale and obviates the need for a finite number of selected grids with their normalizing constants pre-calculated , a strategy taken by Green and Richardson [6] . To identify the optimal genetic profile assignment , we use an ensemble averaging method that aggregates different cluster assignments generated by the MCMC samplers into a consensual one . We find that this cluster algorithm works quite well in simulation studies . A similar idea has been used by Liang [13] and Molitor et al [14] in different contexts . We also propose a resampling-based test based on the fitted Bayesian model that can be used to formally test for the existence of gene–environment interaction . We apply the proposed method to study the joint effect of cigarette smoking intensity and genetic variants in chromosome region 15q25 . 1 using data from EAGLE , a population-based case-control study conducted in Italy [15] . Cigarette smoking is an established major risk factor for lung cancer . Besides environmental exposures , recent GWAS identified a few chromosome regions ( e . g . , chromosomes 15q25 . 1 , 5p15 , and 6p21 ) harboring genetic variants underlying a susceptibility for lung cancer [15] , [16] , [17] . In particular , the chromosome 15q25 . 1 region , which includes the CHRNA5-CHRNA3-CHRNB4 cluster of cholinergic nicotinic receptor subunit genes , has been shown to be associated with both lung cancer and smoking behaviors , such as cigarette smoking intensity [18] , [19] , [20] , [21] , [22] . Although there is no evidence suggesting the existence of multiple loci in this region independently contributing to lung cancer susceptibility in populations of European ancestry [16] , it does appear that there are multiple independent loci within 15q25 . 1 affecting smoking intensity [19] . The main goal of our analysis is to evaluate whether the effect of smoking intensity varies with genetic variants in 15q25 . 1 . Our analysis finds evidence for gene–environment interaction , with the relative risk for smoking appearing to be stronger in subjects with a genetic profile associated with a higher lung cancer risk . The proposed resampling-based test derived from the fitted Bayesian model also detects significant gene–environment interaction ( P-value = 0 . 016 ) . On the other hand , the standard single-marker approach that aims at detecting the interaction between a SNP and smoking intensity fails to reveal any evidence of interaction , with the smallest observed nominal P-value being 0 . 021 among the 32 testing SNPs , and the adjusted P-value based on the permutation test being 0 . 29 .
Assume we have data collected from a case-control study , with cases , controls . Let be the total number of subjects in the study . For the ith subject , we denote its observation by , where for a case , 0 for a control; is the observed exposure for the environment risk factor of interest; represents measures on a set of covariates; and represents multilocus genotypes observed on a set of SNPs in a pre-specified region . In the following discussion , we use the term genotype to refer to the multilocus genotype observed on all considered SNPs within the targeted region . We intend to develop a model for the G-E joint effect that permits G-E interaction . More specifically , we assume the true underlying risk model has the following form: ( 1 ) where clusters 1 to K represent a partition of the genotype space; is the intercept term representing the common baseline log odds for subjects with their genotypes in cluster k; is the effect of E ( in term of log odds ratio ) in the disease model for cluster k , ; and is the vector of coefficients for the set of covariates X and is constant regardless of a subject's genotype . Notice that if the partition of the joint genotype space is known a priori , we can derive the corresponding K-category genetic profile variable L based on the cluster assignment . The above model ( 1 ) is then essentially the standard logistic regression model consisting of L and E as main effects and their product as the interaction term , with adjustment for X , and has the following form:Thus it is clear that there is no G-E interaction if , and the interaction exists if otherwise . In real applications , we do not know a priori the partition of the genotype space . If G consists of just one SNP , the goal can be achieved easily by using a saturated logistic regression including both E and G ( as a three-level categorical variable ) as the main effects and their product as the interaction term . For situations where G consists of multiple SNPs ( e . g . , more than 10 ) , as in the case of the EAGLE lung cancer study , we propose the following Bayesian model that simultaneously searches for the optimal partition of the genotype space and estimates the unknown parameters in the corresponding risk model ( 1 ) . The Bayesian model is built up in a hierarchic framework . We first describe our model by assuming K , the total number of clusters , is known . We will describe how to choose K later . Suppose there are H types of genotype configurations observed in the sample , labeled as genotype 1 , 2 , … , H . We define the latent genotype allocation vector , with , being the cluster assignment for genotype h , . For subject i , we denote its genotype id by . Given the allocation vector and the set of coefficients for the disease model ( 1 ) , the probability of subject i having the disease outcome is ( 2 ) In the above model specification , we use the prospective likelihood function ( 2 ) for observed case-control data , which were collected under a retrospective sampling scheme given the disease outcome . The use of the prospective likelihood function can be partially justified by the general results from Staicu [23] and Seaman and Richardson [24] . They showed the equivalence of prospective and retrospective analysis in the Bayesian framework in the sense that both approaches could yield the identical marginal posterior distribution of the log odds ratio under analyses with properly specified priors . In model ( 2 ) , the effect of E varies with G . Thus we call it the Bayesian risk model allowing for G-E interaction . As a comparison in the analysis , we also consider a model assuming a homogeneous effect from E , which is defined as ( 3 ) We call this model the Bayesian risk model without G-E interaction . In what follows , we will describe methods for fitting model ( 2 ) , the one allowing for G-E interaction . Similar procedures can be applied to model ( 3 ) . To model the distribution of the allocation vector z , we first choose a similarity metric to define the spatial contiguity between any two genotypes . Let J denote the total number of considered SNPs within the region , with the genotype at a given SNP being coded as 0 , 1 , or 2 according to the number of copies of the minor allele . Let genotypes h and have the configurations and , where is the genotype at the jth SNP for the multilocus genotype h . We first define the distance between them aswhere is the variance for the genotype at SNP j observed in the sample , with being the genotype at SNP j for subject i , and . Then we define if is among the 4 ( distinctive ) genotypes closest to genotype h , and is among the 4 genotypes closest to genotype ; if ( or h ) is among the 4 genotypes closest to genotype h ( or ) but this is not true in both cases; and for all other cases . We model z with the Potts model , which has a regulating parameter governing the level of spatial dependency in the cluster assignment . The Potts model has the following form:where , with being the indictor function , i . e . , if and 0 otherwise , and whereis the log normalizing constant . Under the Potts model with , the cluster assignments are allocated independently for different genotypes . When , the cluster assignments for two neighboring genotypes h and ( i . e . , two genotypes with ) are correlated . The level of correlation ( spatial dependence ) increases with . For example , under the genotype configuration observed in the EAGLE study and , the average probability that any two neighboring genotypes are allocated to the same cluster is 0 . 5 when . It increases to 0 . 83 , and 0 . 97 for and 1 . 2 , respectively . More discussions of the Potts model can be found in [6] . We need to specify our prior models for and . In this paper , we consider the normal distribution with a mean of 0 and a variance of 4 or the uniform distribution on the interval of as the prior for each parameter in . We describe the appropriateness of those priors for the prospective likelihood model in the Discussion Section . Both priors are very uninformative and generate similar conclusions on the EAGLE study and simulated datasets . Therefore we present only results based on the normal prior in the following discussions . Following Green and Richardson [6] , the prior distribution for is set to be a uniform distribution on the interval , which covers an appropriate region of such that the resulting class of Potts models are flexible enough to capture a wide range of spatial dependence . We note that cannot be too large . If is over a critical point , the corresponding Potts model would essentially force almost all elements into the same cluster , a well known phenomenon for the Potts model called phase transition property [25] , and in this situation , the MCMC simulation tends to get stuck . We did some experiments to explore the setting of for the Potts model based on the neighborhood configuration observed in the EAGLE study . We found the value induces a high level of spatial dependence , with the average probability that any two neighboring genotypes are allocated to the same cluster being 0 . 97 at ; and when , the average probability goes to 0 . 99 , which indicates an extremely high level of spatial dependence for the Potts model . Based on these observations , we decided to set in our EAGLE study application , as well as in simulation studies that assume the same neighborhood structure as the EAGLE study . We consider only a uniform prior for since in practice we usually do not know which level of spatial dependence is more likely than the others . But the algorithm described below can certainly be used with other prior functions if necessary . Putting all the foregoing models together , we can express the joint distribution of all variables aswhere . The inference ( for a fixed total number of clusters K ) on , and z can be based on the following MCMC algorithm . In our simulation studies and the real data application , we find the MCMC algorithm generally converges after 100 . 000 iterations . Below we describe a procedure for determining the number of clusters , and an ensemble averaging method for the identification of the cluster assignment based on the MCMC samples . It is usually desirable to have a formal statistical test or decision rule for inference regarding the presence of an interaction . Here we propose a resampling-based test for this purpose . First we fit model ( 2 ) , the Bayesian risk model allowing for G-E interaction , under various numbers of clusters . Then we use the +1 rule to identify , the optimal number of clusters that is not less than 2 , and the corresponding consensual cluster assignment L . We require for this interaction test because the interaction test is not defined for . We use the maximum likelihood estimate ( MLE ) to establish the following logistic regression model , ( 9 ) where is the cluster assignment for subject , given by the consensual cluster assignment L . This model includes the main effects of L and E , as well as their interactions . We can conduct a likelihood ratio test comparing model ( 9 ) with the similar model without the interaction terms and obtain the corresponding “P-value” , denoted by , based on the Chi-squared distribution with degrees of freedom ( df ) . Clearly , this “P-value” tends to overestimate the significance level of the interaction , as the variable is data-driven , but a small value for provides evidence against the null . We can use as the test statistic and apply the following resampling-based procedure to evaluate its significance level . In Steps 1 and 2 we establish the Bayesian risk model under the null hypothesis that there is no G-E interaction and the corresponding logistic regression model . We use the fitted logistic regression model ( 10 ) to generate multiple null datasets in Step 3 based on the parametric bootstrap procedure [31] . In Step 4 , for the bth generated null dataset , we first apply the MCMC procedure to establish the Bayesian model given by ( 2 ) and next identify the optimal number of clusters with the +1 rule , as well as the corresponding consensual cluster assignment . Then we fit the corresponding logistic regression model with G-E interaction and obtain the test statistic from the likelihood ratio test .
We used data generated by the lung cancer GWAS in the EAGLE study [15] with 1920 lung cancer cases and 1979 population controls as the basis for our simulation studies and real data applications . We focused on the chromosome region 15q25 . 1 between 76 . 5 Mb and 76 . 72 Mb , with the boundary defined by loci where the recombination rate is relatively high . This region covers all replicated loci relating to smoking behavior or lung cancer risk . We have genotypes on 32 SNPs in the region that have a minor allele frequency ( MAF ) larger than 4% ( estimated in 1979 EAGLE control samples ) . We removed 17 redundant SNPs , leaving a minimal set of 15 SNPs where the pairwise was always less than 0 . 8 . We used genotypes on these 15 tagging SNPs to represent each subject's genetic variation pattern in the region . The reason for removing redundant SNPs is to ensure that the similarity measure between any two types of multilocus genotypes is not dominated by a set of SNPs in high linkage disequilibrium . The summary of the 15 chosen tagging SNPs is given in Table 1 . We conducted simulation studies to evaluate the performance of the proposed method for fitting the Bayesian model allowing for G-E interaction . In the simulation study we were interested in studying the interaction between a binary environment risk factor ( or 1 ) and genetic variants ( G ) within a candidate region . We used genotypes at 15 tagging SNPs ( Table 1 ) in 15q25 . 1 observed in the EAGLE study to represent the joint genotype distribution for the simulated population , which consisted of 766 distinct multilocus genotypes . We chose the 2nd , 6th , and 10th SNPs listed in the Table 1 as the functional SNPs , and divided the genotype space into the following three regions according to the total number of risk alleles ( assuming the minor allele to be the high-risk allele ) among the 3 functional SNPs: region I , consisting of genotypes with ; regions II , consisting of genotypes with ; and region III , including genotypes with . We conducted a principal component ( PC ) analysis on subjects from the EAGLE study with genotypes at the 15 SNPs as their coordinates . Figure 1 shows how genotypes ( subjects ) in each of the three regions were distributed in the first 2-PC space , with regions I , II , and III in green , blue , and red , respectively . The disease risk models we considered had the form given by ( 1 ) . Their definitions are given in Table 2 . Under Model there was no genetic effect and no interaction between G and E , and thus there was no risk heterogeneity in the genotype space . Under and , coefficients and had the same clustering pattern . Under models , the risk heterogeneity patterns for and were not matched , unlike those under model and . In model , the two clusters defined by were region I , and regions II and III combined , while the two clusters defined by the effect of were regions I and II combined , and region III . We assumed that the environmental exposure status E ( 0 or 1 ) and G were correlated in the general population . The distribution of E depended on G in the following way: for a subject with genotype in region I , the probabilities of being unexposed ( ) or exposed ( ) were 0 . 7 and 0 . 3; for a subject with genotype in one of the other two regions , those probabilities were 0 . 4 and 0 . 6 for and . Thus the distribution of E was quite different for subjects with different genotypes . Under each model , we simulated 50 datasets representing a case-control study with 1500 cases and 1500 controls . We ran the MCMC algorithm with 2 , 000 , 000 iterations with the first 1 , 000 , 000 iterations being discarded . We used an algorithm similar to that described in [32] to simulate the case-control study . Note that under the case-control sampling scheme , we do not need to specify a value for . Instead , we just need to know the values of , , in order to simulate datasets from a case-control study . For each simulated dataset , we applied our method with auxiliary samples , with the number of clusters K ranging from 1 to 8 . We used the +1 rule defined by ( 8 ) to identify , the optimal number of clusters . Table 3 provides a summary of the number of clusters identified over 50 simulated datasets under each risk model . We can see from the table that the +1 rule performs quite well in identifying the right number of clusters , even in situations where there is no clustering structure ( i . e . , the true number of clusters , , is 1 ) . We evaluated the performance of the algorithm for cluster assignment by comparing the cluster assignment estimated at with the true underlying cluster assignment chosen by the simulation design . For model , the clustering patterns for and were not matched . In this case we treated the finer partitioning ( given by Figure 1 ) that accommodates the clustering patterns of both and as the true one . The accuracy of the estimated cluster assignment was measured as the proportion of subjects being assigned to the same cluster by both assignments ( the estimated one and the true one ) . The accuracy summary over 50 replications under various considered models ( except , the model with no clustering structure ) is given in Table 4 . It indicates that the cluster assignment algorithm appears to be able to partition the subjects ( and genotypes ) into the proper subgroups , provided that the correct number of clusters can be identified . We also evaluated the accuracy of the estimated coefficients ( and ) . Based on the true risk model ( 1 ) , subject i with genotype was assigned to one of the risk models . We considered coefficients and in that risk model to be the true coefficient values for this subject . Thus , subjects with their genotypes belonging to the same cluster would share the same true coefficient values . We used , the posterior median of assigned to subject i based on MCMC samples generated under , as the estimates for the underlying coefficients . The number of clusters was estimated by the +1 rule , as described previously . Since the odds for the genetic effect is not identifiable under the case-control design , we were interested only in the difference in between two groups . To present the result for each experiment , we shifted the value of , the posterior median of for subject i , by a constant value , which was chosen as , the median of among subjects in true cluster 1 . As a result , the median level of the shifted posterior median ( we still represent it as ) among subjects in cluster 1 is 0 . In Figure 2 , Figure 3 , Figure 4 , and Figure 5 , we present summaries of and for each of the 50 experiments under models and . Summarized results for model are given in Figures S1 and S2 . Each boxplot is a summary of or among subjects in a true underlying cluster . From those figures , we can see that the estimates align with their true values quite well . Notice that these estimates were obtained under the model with the number of clusters estimated by the +1 rule . We inspected the algorithm's convergence using the Gelman and Rubin's diagnostic plot [33] , as implemented in the CODA R package [34] . For each model , we checked the convergence on the first 5 simulated datasets used in the above simulation studies , with 5 independent runs on each dataset . We found that the proposed algorithm can converge within 100 , 000 iterations , with the estimated shrinkage factor falling below the recommended threshold of 1 . 1 . We also show in Figures S3 and S4 the posterior distributions for , resulting from each of 5 independent runs on the first simulated dataset under models , and . It is evident from these plots that we can obtain very consistent posterior distributions for parameters of interest among different runs on the same data . We conducted a simulation study to evaluate whether the proposed resampling-based test can maintain the proper type I error rate . We considered a disease risk model that had the main effects from G ( with OR = 4 for genotypes falling into regions II and III vs . those in region I ) and E ( with a common OR of 4 for vs . ) , with no interaction between G and E . Regions are defined in Figure 1 . We assumed a study sample size of 600 cases and 600 controls , and simulated 1000 datasets under the considered risk model as did before . For each dataset , we ran the resampling-based test with 1000 bootstrap steps for the estimation of the P-value , allowing the number of clusters to vary from 2 to 5 . To reduce the computing time further , we ran the MCMC algorithm for 300 , 000 iterations with the burn-in period consisting of the first 200 , 000 iterations for each bootstrapped sample ( as 200 , 000 iterations appear to be enough to ensure the convergence of the MCMC algorithm ) . We found that the proposed resampling-based test can maintain the proper type I error in the considered scenario , with estimated false positive rates of 0 . 055 and 0 . 097 under nominal levels of 0 . 05 and 0 . 10 , respectively . We compared the power of the proposed resampling-based test with two other standard interaction tests , the minP-SNP and minP-PC tests . Both test statistics are based on the minimum P-value observed on a set of univariate G-E interaction tests , with their significant levels being evaluated through a resampling-based procedure . The minP-SNP test is based on the set of single SNP-environment interaction tests , with each SNP-environment interaction test statistic being derived from the standard likelihood ratio test comparing two logistic regression models with and without the SNP-environment interaction term . The SNP effect is modeled with a categorical variable with three levels so each SNP-environment interaction test considered in the minP-SNP test is a 2 df test . The minP-PC is based on a set of tests that evaluate the interaction between a single principal component ( PC ) and the environment variable . PCs are derived from the principal component analysis of genotypes on all considered SNPs . Similar to the minP-SNP test , each PC-environment interaction test is derived from the likelihood ratio test comparing two logistic regression models with and without the interaction term . The PC effect is model as a continuous variable . Both minP-SNP and minP-PC were based on 15 univariate tests in the simulation study as there were a total of 15 SNPs in the considered chromosome region . We evaluated the power under six different disease risk models , including , , and defined in Table 2 , and the three additional models , , and . Model and had just one functional SNP ( the 10th SNP in Table 1 ) . Model had 2 clusters , with coefficients in the formula ( 1 ) being for genotypes satisfying the condition ( cluster 1 ) , and for , or 2 ( cluster 2 ) . Model had 3 clusters , with coefficients for ( cluster 1 ) , for ( cluster 2 ) , and for ( cluster 3 ) . Model adopted a 2-cluster pattern observed in the analysis of the EAGLE study described later , with clusters 1 and 2 consisting of genotypes in red and blue colors , respectively ( Figure 6 ) , and with for cluster 1 , and for cluster 2 . The correlation between E and G was defined similarly as before . For a subject with genotype in cluster 1 , the probabilities of being unexposed ( ) or exposed ( ) were 0 . 7 and 0 . 3; for a subject with genotype not in cluster 1 , those probabilities were 0 . 4 and 0 . 6 . Under each disease model , we simulated 500 datasets , with each consisting of 600 cases and 600 controls . The summary for the power comparison results is given in Table 5 . It can be seen from the table that the proposed test has a clear advantage over two other standard interaction tests , especially when the underlying clustering pattern in the disease risk cannot be properly approximated by a single SNP or PC ( e . g . , under the model ) . Even under the model where the single SNP-environment interaction test based on the 10th SNP is most optimal , due to the multiple comparison adjustment , the minP-SNP test is only slightly more powerful than the proposed test . Under the model where the functional SNP ( the 10th SNP ) has a dominant effect in its interaction with E , the minP-SNP test compares less favorably with the proposed test since each of single SNP-environment interaction test considered in the minP-SNP global test spends one more df than necessary ( as there are only two cluster in the model ) . The minP-PC test has the worst overall performance as it is very sensitive to its underlying assumption that the genetic effect is linearly correlated with one of the PC direction . We applied the proposed method to study the joint effect of cigarette smoking intensity ( number of packs per day ) and genetic variants in chromosome region 15q25 . 1 on lung cancer risk , using data generated by the EAGLE study . We focused on former and current smokers who had been genotyped on the 15 tagging SNPs . We also removed , as outliers , 8 subjects who had smoked more than 3 packs of cigarette per day . The final dataset for our analysis consisted of 1326 controls and 1720 cases . In the analysis we treated smoking intensity as a continuous variable and adjusted for the effects of gender and of age at diagnosis ( categorized as: < = 60 , 61–70 , >70 ) . We used a Bayesian model that allowed for G-E interaction , unless specified otherwise . To determine the number of clusters , we ran the MCMC algorithm 20 times with different random seeds for each K , , in order to estimate the Monte Carlo standard error for DIC . Figure 7 shows the DIC values for each K over 20 replications . Based on the 1 SE rule given by ( 7 ) , the optimal number of clusters was found to be 2 , with its averaged DIC value being 3810 . 5 . The partitioning of subjects into 2 clusters based on our proposed clustering algorithm is very consistent among 20 replications . The discrepancy rate between assignments from any two runs , defined as the proportion of subjects being assigned to two different clusters , is less than 1 . 4% under . Below we present the posterior summary based on a single run of our algorithm . To present the summary result , we first conducted a PC analysis on the case-control samples using genotypes at the 15 tagging SNPs as coordinates . In Figure 6 , we plotted subjects by their first 2 PCs , with different colors representing their cluster assignments under . The cluster assignment was performed with the ensemble averaging method described above . Since subjects with the same genotype were represented by one point in the first 2-PC space , we can think of each point as either a unique genotype or a group of subjects sharing that genotype . There are 2240 subjects with 410 unique genotypes grouped into one cluster ( shown in red in Figure 6 ) and 806 subjects with 252 unique genotypes grouped into another cluster ( shown in blue in Figure 6 ) . Notice that the two clusters are defined in term of estimated risk coefficient values ( ) , but not in term of genotypes distribution in the PC space . That is why these two clusters do not appear to be well separated in the PC space . To summarize the effect of smoking on a subject with genotype h , , we focused on , the posterior median of , with being the coefficient for smoking in the risk model assigned for a subject with genotype h . We can interpret as the posterior median of the OR associated with one more pack of cigarettes per day for a subject with genotype h . To summarize the genetic effect of genotype h , we used , the ratio of the posterior median of versus the posterior median of , with being the intercept for the risk model assigned for a subject with genotype h and being the chosen reference genotype . We chose the reference genotype as the one having the lowest posterior median of . We can interpret as the posterior median OR between genotype h and the reference genotype . In Figure 8 , we show a smoothed surface plot for the smoking effect measured by , and the genetic effect measured by for each genotype in the first 2-PC space , based on models run under . The smooth surface was estimated by the kriging method with each genotype's top 5 PCs ( which account for over 85% of the total variation ) as predictors . The plots were generated using the functions provided in the R package called “fields” [35] . It is evident from Figure 8 that neither the smoking effect nor the genetic effect is uniformly distributed over the genotype space . The smoking effect on a subject depends on his or her genotype . It is considerably lower on subjects who have their genotypes projected on the lower part of the PC space than on subjects with their genotypes projected elsewhere . Some understanding of the 2nd PC is helpful for interpreting the patterns observed in Figure 8 . From Table 1 , we can see that the 2nd PC is driven mainly by the 8 SNPs with absolute loading values larger than 0 . 18 , with the remaining having loading values less than 0 . 07 . These 8 SNPs also turn out to be the ones that are most significantly associated with lung cancer risk ( Table 1 ) . We point out the fact that the loading value for each of the 8 SNPs is negative if the SNP's major allele is the high-risk allele , positive if its minor allele is the high-risk allele . As a result , a genotype's 2nd PC coordinate is positively correlated with its total number of risk alleles among the 8 SNPs ( see Figure S5 ) . Genotypes with smaller 2nd PC coordinates tend to have fewer high-risk alleles and are expected to have smaller ORs than genotypes having larger 2nd PC coordinates . As a comparison , we also fit model ( 3 ) , the Bayesian model without G-E interaction . The optimal model based on the 1 SE rule was again achieved at , with its averaged DIC value being 3817 . 5 over 20 runs ( Figure S6 ) . The DIC is noticeably higher than that obtained under the Bayesian model allowing for G-E interaction ( DIC = 3810 . 5 ) . This suggests that the model allowing for G-E interaction fits the data better than the model without the G-E interaction . Finally , to demonstrate the existence of G-E interaction further , we applied the resampling-based test described in the Methods section . The observed test statistic was . We applied the resampling-based test by allowing the number of clusters to vary from 2 to 5 . The estimated P-value based on 2000 bootstrap steps was 0 . 016 , suggesting that there is a significant G-E interaction . On the other hand , for each of the 32 relatively common SNPs ( MAF>0 . 04 ) in this considered 15q25 . 1 region , we conducted the standard SNP-smoking interaction test ( 2 df ) based on the logistic regression model by treating the genotype as a three-level categorical variable . The smallest nominal P-value we observed was 0 . 021 . The global minP-SNP test had a P-value of 0 . 29 , which was well above the 0 . 05 level . We also conducted the PC-smoking interaction test by modeling each PC as a continuous variable . The smallest nominal P-value was 0 . 058 . The P-value from the global minP-PC test was 0 . 62 .
Our new method can evaluate gene–environment interaction at the gene/region level by integrating information observed on multiple SNPs in the considered gene/region with measures of environmental exposure . This method reduces the impact of loss of efficiency and bias from the misclassification error inherent in the single-marker approach that studies the environmental risk factor and one SNP at a time . The method provides a coherent inference framework that allows us to evaluate the environmental effect on different strata defined by the multi-locus genotype . A heterogeneous environmental effect across different strata would signal the presence of gene–environment interaction . We also propose a resampling-based test to formally test for the existence of gene–environment interaction . Genetic variations within the 15q25 . 1 region have been shown to be associated with both lung cancer risk and smoking behaviors , such as the smoking intensity . Our analysis based on the EAGLE case-control study indicates that the smoking effect varies according to the subject's genetic makeup in the 15q25 . 1 region . The proposed resampling-based test also supports the existence of gene–environment interaction ( P-value = 0 . 016 ) . On the other hand , two conventional tests of gene–environment interaction based on the single-marker and single-PC approaches are far from significant . This highlights the advantage of our proposed method over standard approaches . Accurate assessment of the environment risk exposure in the evaluation of gene–environment interaction is as important as identification of functional genetic variants or their proper surrogates [36] . In the EAGLE population-based case-control study , the information on smoking collected near the time of diagnosis is likely to provide a more accurate measure of risk exposure than information collected in other prospective cohort studies , such as the Prostate , Lung , Colorectal , and Ovarian ( PLCO ) Screening Trial [37] , which does not reflect subsequent changes in smoking behavior like quitting . For example , we observed a much larger OR for smoking one more pack of cigarette per day ( 3 . 7 , z statistic = 15 . 58 ) in the EAGLE study than in a lung cancer case-control study nested within the PLCO cohort ( 1 . 84 , z statistic = 8 . 87 ) , which includes 1390 lung cancer cases and 1924 controls . We also could not find evidence for smoking-15q25 . 1 interaction in this PLCO nested case-control study by using our proposed method . The difference in the smoking OR estimates and the absence of gene–environment interaction evidence using our method in the PLCO study may be a consequence of greater misclassification error in the smoking information assessment in the cohort study ( PLCO ) than in the case-control study ( the EAGLE study ) . In our method , we adopted the Potts model for the latent allocation vector for cluster assignment , as did Green and Richardson [6] . We used the MCMH algorithm [12] for simulating the regulating parameter of the Potts model . The MCMH algorithm overcomes the intractable normalizing constant problem that cannot be handled by the standard MH algorithm , while ensuring the consistency of the Monte Carlo estimates . Furthermore , this MCMH algorithm can readily be used for Potts models with certain restrictions on the sampling space by modifying the MH step to generate allocation vectors accordingly . We proposed to use the +1 SE rule ( or the +1 rule ) based on DIC to identify the optimal number of clusters . We found through simulation studies that this approach works quite well . An alternative approach would be to treat the number of clusters as a random variable and integrate it into a Bayesian model [6] . A reversible jump MCMC algorithm [38] could be used to facilitate the move between sampling spaces with different dimensions . It would be interesting to compare the performance of these approaches , especially in term of their abilities to identify the proper number of clusters . The proposed procedure relies upon a user-specified similarity metric to define the neighborhood among different genotypes in the Potts model . This neighborhood structure is used to induce the spatial dependency in the cluster assignment . In this paper , for a given genotype , we chose its 4 nearest genotypes as its neighbors . We found that the analysis result was not very sensitive to how the neighborhood is defined as long as the chosen Markov structure can generate an appropriate spatial dependence . For example , we reanalyzed the EAGLE study with two other types of Markov structures: one using the 3 nearest genotypes as neighbors , and the other one using the 5 nearest genotypes as neighbors . We show in Figure S7 the comparison of the posterior medians of the genetic effect ( ) and the smoking effect ( ) estimated for each subject between each of the new runs and the original runs under . It is clear that results from these three analyses are quite similar . We used the prospective likelihood model in the Bayesian framework for case-control studies , even though the data were collected retrospectively according to a subject's disease status . According to [23] , [39] , given certain priors for parameters in the retrospective model , we can derive corresponding priors for the prospective model parameters that yield the same marginal posterior distributions as their retrospective counterparts . In this paper we consider both normal and uniform distributions as priors for the prospective model parameters . Although we cannot derive explicitly their corresponding priors for the retrospective model , our simulation studies demonstrated that the proposed prospective approach indeed had the desired performance when applying to data generated from case-control studies . The normal prior has also been used with the prospective likelihood model on case-control studies in other contexts ( e . g . , [40] , [41] ) . We have created an R package called BaDGE ( Bayesian model for Detecting Gene Environment interaction ) implementing the proposed Bayesian model and the associated post-processing procedures . The package is freely available from the website http://dceg . cancer . gov/bb/tools/badge . Currently , we consider only binary or continuous environmental exposure variable . It is straightforward to expand the algorithm to deal with a categorical ( with more than 2 levels ) environmental variable . To use the program , the user needs to specify priors ( normal or uniform distribution ) for parameters in the risk model and a prior ( a uniform distribution ) for the regulating parameter in the Potts model . The program will be expanded further to incorporate other prior functions . The running time for 200 , 000 iterations using 50 auxiliary samples in the MCMH algorithm on a dataset of 1000 cases and 1000 controls , with approximate 450 unique genotypes , is about 14 minutes on a Linux machine with the 2 . 8 GHz AMD Opteron processor . For a dataset with a large number of genotypes ( e . g . , over 1000 ) , we can reduce the computing time by first dividing the whole genotype space into a few hundreds of subgroups through the PAM clustering algorithm [30] and then treating subgroups as genotypes in the proposed MCMC procedure . For example , the running time on the same testing example mentioned above decreases to 8 minutes if we regroup the genotypes into 250 unique subgroups . Another way to reduce the total number of genotypes is to limit tagging SNPs to those with a relatively large minor allele frequency . The resampling-based test could be computationally intensive for a dataset like the EAGLE study . We are still investigating whether it is possible to replace the computationally intensive resampling-based procedure with a one-step multiple comparison adjustment approach , similar to one used in [42] , for the assessment of the statistical significance level . Comparing to the standard single-marker or principal component based approaches , our proposed method is more computationally intensive , but it has several important advantages . First , it offers a more flexible way to model gene–environment interaction , especially complicated ones that cannot be depicted properly by the single-marker or PC based approaches , such as in situations where genetic variants ( might or might not be directly genotyped ) in multiple loci within the considered region interplay with the environment risk factor . Second , it provides an estimate of the environmental effect on subjects with a given joint genotype profile . This could be potentially useful to generate new hypotheses on how the gene and environment risk factor interacts . Third , as shown in the simulation studies and real application , the proposed resampling-based test derived from the Bayesian model has a more robust performance than the standard single-marker , or PC based testing procedures . For example , in situation where the single marker test is most appropriate , i . e . , there is only one functional locus in the considered region , the proposed test is only slightly less powerful than the single-marker test . But it has a considerable power advantage over the standard tests when the underlying disease risk pattern cannot be properly approximated by a single SNP or PC . Although our method is described in the context of gene–environment interaction detection , it is in fact quite general . It provides a general strategy for studying the interaction between an observed risk factor and a latent categorical variable not directly observed or clearly defined , but one that can be derived from a set of observed relevant covariates . For example , our method can be used with some minor modifications to evaluate the interaction between smoking behavior ( e . g . , smoking intensity ) and a latent dietary pattern that can be derived from food frequency questionnaires . Also , it is possible to extend our method to study gene-gene interaction by introducing two latent factors to capture the effect of both genes , as was done in [43] . | Many common diseases result from a complex interplay of genetic and environmental risk factors . It is important to study the potential genetic and environmental risk factors jointly in order to achieve a better understanding of the mechanisms underlying disease development . The standard single-marker approach that studies the environmental risk factor and one genetic marker at a time could misrepresent the gene–environment interaction , as the single genetic marker might not be an appropriate surrogate for the underlying genetic functioning polymorphisms . We propose a method to look at gene–environment interaction at the gene/region level by integrating information observed on multiple genetic markers within the selected gene/region with measures of environmental exposure . Using data collected in the Environment and Genetics in Lung Cancer Etiology ( EAGLE ) study , we apply the proposed model to evaluate the joint effect of smoking intensity and genetic variants in the 15q25 . 1 region and find evidence for gene–environment interaction ( P-value = 0 . 016 ) , with the smoking effect varying according to a subject's genetic profile . | [
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] | [
"mathematics",
"computer",
"science"
] | 2012 | A Flexible Bayesian Model for Studying Gene–Environment Interaction |
Meningococcal meningitis is a major health problem in the “African Meningitis Belt” where recurrent epidemics occur during the hot , dry season . In Niger , a central country belonging to the Meningitis Belt , reported meningitis cases varied between 1 , 000 and 13 , 000 from 2003 to 2009 , with a case-fatality rate of 5–15% . In order to gain insight in the epidemiology of meningococcal meningitis in Niger and to improve control strategies , the emergence of the epidemics and their diffusion patterns at a fine spatial scale have been investigated . A statistical analysis of the spatio-temporal distribution of confirmed meningococcal meningitis cases was performed between 2002 and 2009 , based on health centre catchment areas ( HCCAs ) as spatial units . Anselin's local Moran's I test for spatial autocorrelation and Kulldorff's spatial scan statistic were used to identify spatial and spatio-temporal clusters of cases . Spatial clusters were detected every year and most frequently occurred within nine southern districts . Clusters most often encompassed few HCCAs within a district , without expanding to the entire district . Besides , strong intra-district heterogeneity and inter-annual variability in the spatio-temporal epidemic patterns were observed . To further investigate the benefit of using a finer spatial scale for surveillance and disease control , we compared timeliness of epidemic detection at the HCCA level versus district level and showed that a decision based on threshold estimated at the HCCA level may lead to earlier detection of outbreaks . Our findings provide an evidence-based approach to improve control of meningitis in sub-Saharan Africa . First , they can assist public health authorities in Niger to better adjust allocation of resources ( antibiotics , rapid diagnostic tests and medical staff ) . Then , this spatio-temporal analysis showed that surveillance at a finer spatial scale ( HCCA ) would be more efficient for public health response: outbreaks would be detected earlier and reactive vaccination would be better targeted .
Meningococcal meningitis ( MM ) , caused by the bacterium Neisseria meningitidis ( Nm ) , is a major health problem in sub-Saharan Africa . The highest incidences of the disease are observed in the so-called “African Meningitis Belt” where annual recurrent epidemics occur during the very hot , dry season [1] . In Niger , reported meningitis cases varied between 1 , 000 and 13 , 000 from 2003 to 2009 , with case-fatality rates of 5–15% . The factors involved in the spatio-temporal occurrence of meningococcal epidemics are only suspected and still poorly understood . Surveillance and reactive vaccination are the predominant strategies for managing meningococcal meningitis outbreaks in the Belt , recently completed with a conjugate vaccine to prevent the carriage of Nm serogroup A . In Niger like in most sub-Saharan countries , surveillance is performed at the district level . Quantitative morbidity and mortality data on meningitis are collected within a reporting network managed by the Direction for Statistics , Surveillance and Response to Epidemics ( DSSRE ) from the Ministry of Public Health . Data from all health care facilities covering the entire Niger population are collected on a weekly basis by the district health authorities , which aggregate and forward their data to the regions and subsequently to the DSSRE . These reported data include all suspected and probable cases , according to the standard clinical definition of meningococcal meningitis [2]: A suspected case is any person with sudden onset of fever ( >38 . 5°C rectal or 38 . 0°C axillary ) and one or more of the following signs: stiff neck , altered consciousness or other meningeal sign; in patients under one year of age , a suspected case occurs when fever is accompanied by a bulging fontanelle . A probable case is defined as a suspected case with turbid CSF or Gram stain showing Gram-negative diplococcus or petechial/purpural rash or ongoing epidemic . Laboratory confirmation of meningitis is not required to report a case . In parallel to this epidemiologic surveillance and in close collaboration with DSSRE , the Centre de Recherche Médicale et Sanitaire ( CERMES ) is in charge of the national microbiological surveillance of meningitis . The CERMES collects the cerebrospinal fluid ( CSF ) samples taken from suspected cases of meningitis by health care workers or physicians and carries out the etiological diagnosis ( see Methods ) . Based on this national surveillance system , the strategy applied in Niger to respond to meningitis outbreaks with limited amounts of available vaccines , is to initiate reactive vaccination in a district once weekly incidence exceeds the epidemic threshold defined by WHO ( see [3] and Methods for definitions ) . Thus , early detection of epidemics is essential for an effective operational response . Analysing spatio-temporal patterns of epidemics at a fine geographic scale could lead to a better understanding of the underlying causes of the disease and potential future prediction of outbreaks [4] . One of the techniques to uncover spatial patterns of disease is cluster detection . In epidemiology , a cluster is a number of health events situated close together in space and/or time [5] . Identifying spatial and spatio-temporal clusters of cases could help: ( i ) to generate new information for further etiologic studies; ( ii ) to identify risk areas where to focus the surveillance and allocate the resources ( antibiotics , rapid diagnostic tests… ) ; ( iii ) to develop cost-efficient vaccination strategies . In sub-Saharan Africa , data from national disease notification have already been used at country , regional or district levels to study the geographical and temporal dynamics of epidemics and their correlation with environmental factors [6]–[12] . However , little is known about MM emergence and distribution at a sub-district level . Using a finer spatial scale such as health centre catchment areas ( HCCAs ) would have several advantages: ( i ) it would capture heterogeneity in MM incidence at sub-district level; ( ii ) epidemic thresholds would be studied at a more accurate scale , allowing for a more rapid and targeted public health response; ( iii ) monitoring of the impact of the intervention would be performed at the same level as the intervention itself . Therefore , we aimed to investigate the spatio-temporal distribution of MM epidemics in Niger at the health centre catchment area level , to identify the most frequently affected HCCAs requiring a particular attention from public health authorities . The national microbiological surveillance database was used to perform two cluster detection methods in order to uncover spatial and spatio-temporal clustering of MM incidence from July 2002 to June 2009 . Then , as a preliminary analysis to a more thorough etiologic study , we searched for ecologic correlation of MM incidence with human density and roads at the HCCA level . Finally , to further investigate the benefit of using a finer spatial scale for surveillance and disease control , we compared timeliness of epidemic detection at the HCCA level versus district level . This paper provides new insights into the spatio-temporal dynamics of MM epidemics and discusses the potential implications of our findings for meningitis control in sub-Saharan Africa .
The CERMES is the national laboratory in charge of the microbiological surveillance of meningitis in Niger . This surveillance has been reinforced since 2002 [13] , [14] by its extension to the whole country ( it was only effective in the capital city before 2002 ) and by the inclusion of a Polymerase Chain Reaction ( PCR ) assay for etiological diagnosis of meningitis to the DSSRE routine surveillance . CSF samples were collected by health care workers or physicians from suspected cases of acute meningitis . Each CSF was documented with an epidemiological form that included date of sample collection , clinical information and general characteristics about the patient ( age , sex , geographic origin such as region , district , HCCA and village ) . The samples were kept either refrigerated or frozen in health facilities , or inoculated into a trans-Isolate ( TI ) medium . The more remote health centres sent CSF samples ( frozen in a cool box ) on a voluntary basis to CERMES by mandated transport companies . Additionally , CERMES carried out active collection of samples twice a day in Niamey , so that the samples remained suitable for culture , and every month within a radius of about 300 kilometres around Niamey , in the regions of Tillabery and Dosso . Etiological diagnosis of MM was carried out by PCR for all CSF as described in [13] and [14] and by culture [15] for suitable CSF received promptly at CERMES ( fresh CSF and CSF inoculated into TI medium ) . Questionnaire data and microbiological results were entered in a database managed by CERMES . The data were used for a retrospective study on meningococcal meningitis cases between July 1 , 2002 and June 30 , 2009 . All data were collected through the national routine surveillance system . Therefore , written consent was not asked and approval from the national ethics committee was not needed . However , patients were informed of the reason why their cerebrospinal fluid was sampled and confidentiality on patients' identity was guaranteed . In 2008 , in order to create a digitized National Health Map of Niger , CERMES mapped the country's HCCAs , each of which included all villages served by the same health centre . As projected data were required for the spatial statistics , all analyses were carried out with a projected version of the National Health Map , using the WGS84 – UTM32N projection . The number of inhabitants per village was extracted from the 2001 census database of the Institut National de la Statistique ( INS ) and an annual population growth rate of 3% was applied . A shapefile of primary roads was retrieved from the HealthMapper application of the World Health Organization ( WHO ) .
From July 1 , 2002 to June 30 , 2009 , a total of 15 801 CSF specimens from meningitis suspected cases were analysed at the CERMES laboratory ( table 1 ) . 112 CSF ( 0 . 7% ) could not be tested ( depleted , broken tubes… ) and 79 ( 0 . 5% ) did not give conclusive results because of contamination . Overall , biological specimens originated from 416 ( 61% ) of the 682 HCCAs mapped in 2009 . Among these CSF , 6556 ( 41 . 5% ) were confirmed as bacterial meningitis cases , 82 . 2% of which were positive for Neisseria meningitidis . Serogroup A was the predominant serogroup every year , except in 2006 . The mean ( SD ) age of the MM cases was 9 . 6 ( 7 . 5 ) years and 58 . 8% were male . Over the study period , MM cases were detected in 349 HCCAs ( 51 . 2% ) in all regions of Niger ( figure 1 ) , with contrasting incidence rates within districts . The highest incidence rates were found in HCCAs of Niamey , Tillabery , Dosso , Tahoua , Maradi and Zinder regions . As for the temporal distribution , 82 . 5% of the MM cases occurred from February to April . Figure 2 depicts for each year the Kulldorff's spatial scan statistic results overlain on the Anselin's Local Moran's I results . Over the seven years , the Local Moran's I method identified 140 high-risk HCCAs ( 130 high-high and 10 high-low ) , with an annual number ranging from 11 ( in 2003 and 2007 ) to 31 ( in 2008 and 2009 ) . The spatial scan method identified 58 significant spatial clusters altogether , with an annual number ranging from 3 ( in 2003 ) to 16 ( in 2009 ) . The median number of HCCAs per cluster was 2 ( IQ range = 1–5 ) and the median annual incidence rate of the clusters was 34 . 9 ( IQ range = 20 . 5–72 . 3 ) cases per 100 , 000 . Almost 80% of the high-risk HCCAs identified with the Local Moran's I were included in clusters detected by SaTScan and 62% of the SaTScan clusters encompassed high-high or high-low HCCAs . Spatial clusters generally occurred in different HCCAs from year to year over the study period , as shown by the low frequencies observed at the HCCA level ( figure 3 ) . Among the HCCAs contributing to a cluster at least once over the study period , the median frequency was 1 ( range = 1–4 ) for clusters detected by at least one method , and 1 ( range = 1–3 ) for clusters detected by both methods . Only four HCCAs were detected three or more times by at least one method and two or more times by both methods . They were: Chare Zamna ( in Zinder urban community ) , Gazaoua , Doumega and Loudou . Spatial clusters most frequently occurred within nine districts out of 42 , containing three or more times a cluster detected by at least one method , and two or more times a cluster detected by both methods . These districts were: Tera and Say ( bordering Burkina Faso ) , Keita , Zinder and five districts bordering Nigeria , Doutchi , Madaoua , Guidan Roumji , Madarounfa and Aguie . The median time interval between two clusters occurring in the same district was one year . When a district contained a cluster detected by at least one method , only 13 . 3% ( median ) of its HCCAs contributed to that cluster , and 9 . 7% when a district contained a cluster detected by both methods . No systematic spatio-temporal pattern for cluster emergence and epidemic spread was observed within the seven years of the study period . Figure 4 shows the 66 significant spatio-temporal clusters detected with the SaTScan space-time scan ( except a 2009 northeast cluster in Dirkou , Bilma district , which is outside the displayed zone ) and the incidence rate observed for each HCCA of a spatio-temporal cluster during the time period associated to that cluster . They essentially occurred between February and April , with an additional few at the beginning ( November–January ) and the end ( May ) of the epidemics . In 2003 , the epidemic could be summarized in two western and eastern poles , with the western pole occurring before the eastern one . In 2004 , the first cluster was detected in the west; then all clusters appeared in the east , ending with the northernmost one in Tanout district . In 2005 , clusters were detected only in the eastern part . In 2006 , two spatio-temporal poles were clearly distinguished , first in the east and then in the west . In 2007 , the first three clusters were detected in the west , followed by one in the east , still another one in the west and a final northernmost one in Keita district . In 2008 , between the eastern clusters at the beginning and the end of the epidemic season , other clusters essentially appeared in the centre ( Tahoua region ) and the west ( Tillabery and Dosso regions ) without a clear order , concluding again the northernmost cluster in Keita district . In 2009 , from the first cluster in the east to the final one in the west , clusters appeared in all regions in between , but followed no clear geographical direction . No significant correlation was found between MM incidence at the HCCA level and human density ( r = 0 . 02 ) or distance to primary roads ( r = −0 . 07 ) . Between 2003 and 2009 , 88 districts crossed the alert threshold . For 42 ( 47 . 7% ) of them , the alert threshold was crossed earlier ( 4 weeks early in median ) in at least one HCCA of these districts . Between 2003 and 2009 , 46 districts crossed the epidemic threshold . For 15 ( 32 . 6% ) of them , the epidemic threshold was crossed earlier ( 3 weeks early in median ) in at least one HCCA of these districts .
To our knowledge , this is the first study using health centre catchment areas as spatial units for the spatio-temporal analysis of MM over a whole sub-Saharan country . The study's first finding was the more frequent detection of spatial clusters within nine southern districts , mainly on the southern border with Nigeria . Second , clusters most often encompassed only a few HCCAs within a district , without expanding to the entire district . In addition , no consistent annual spatio-temporal pattern for cluster emergence and epidemic spread could be observed , thus precluding the capacity to predict where the next epidemic would break out , and what geographical direction it would follow . These findings rely on laboratory-based data and have important public health implications as discussed hereafter . The first asset of this study was the quality of the microbiological data . We used laboratory-confirmed N . meningitidis cases data , coming from a surveillance system managed by CERMES and DSSRE throughout the country . Most other spatio-temporal studies on meningitis epidemics in sub-Saharan Africa [6]–[12] , [19] are based instead on suspected cases reported in the framework of the national surveillance systems . In our dataset , none of the three typical bacterial aetiologies ( N . meningitidis , S . pneumoniae and H . influenzae ) could be identified in almost 60% of the CSF analysed by CERMES over the study period ( see Table 1 ) . Relying only on suspected cases would therefore introduce a large number of misclassified cases . However , our system may suffer from underreporting from areas where performing a lumbar puncture and shipping the samples to CERMES may represent logistical difficulties . Further analyses ( not shown here ) have documented that indeed the districts the most remote from CERMES ( in Maradi and Zinder regions ) were sending less CSF specimens than the closer ones , for a similar number of suspected cases notified to DSSRE . However , the proportion of negative cases among the received CSF specimens was fairly similar among the healthcare centres ( outside the capital Niamey ) . This suggested that the decision to take or not a CSF sample from a patient based on clinical criteria had no significant spatial variability . Moreover , our cluster analyses enabled us to detect the importance of remote regions in the epidemic dynamics according to the recurrent clusters identified there . Like in many other settings , the surveillance system may not cover the entire population of Niger affected by meningitis . However , we can reasonably assume that most meningitis cases , because of their severity , end up reaching the healthcare centres , with or without prior self-treatment or consultation of a tradi-practitioner . Moreover , free healthcare offered to all people suffering from meningitis in Niger probably reduces social and spatial disparities in care-seeking behaviours . Thus , for all the reasons above , we are confident that the surveillance system is representative enough and that underreporting did not substantially affect the validity of our results , which are more likely to reflect the dynamism peculiar to meningitis than the spatial disparities in the surveillance system efficiency . Incidence estimates were based on the 2001 census and constant population growth rates were applied . We could not take into account possible variations of population growth rate over time and space , due to the difficulty in quantifying population migrations . The second asset of this study was the use of HCCAs as spatial units for the spatio-temporal analysis of MM . They represent a more accurate spatial unit of analysis than the district level on which reactive vaccination strategies and spatio-temporal studies are usually based [3] , [6] , [7] , [19] . Analysing data at the HCCA level has greater relevance for understanding the epidemic dynamics , for making decisions in response to starting epidemics and for assessing control strategies . Indeed , this study has shown that clusters most often included only a few HCCAs within a district . This finding , previously suggested by [20] , is important for understanding meningitis epidemics and should encourage surveillance at the health centre level . Clusters occurred in different HCCAs within the same districts in consecutive years , demonstrating strong intra-district heterogeneity and year-to-year variability of the affected HCCAs . This could result from outbreaks limited to HCCAs without exceeding the threshold at the district level: the district is not vaccinated and may be affected by a large outbreak the following year . Besides , waiting for the threshold to be reached at the district level to initiate reactive vaccination may incur unnecessary delays: we showed that a decision based on threshold estimated at the health centre level might lead to earlier detection of outbreaks , so more reactive and possibly more cost-effective vaccination strategies . Thus , adding HCCA-level surveillance to the current district-level surveillance would improve the timeliness of epidemic detection . With the introduction of a new meningococcal A conjugate vaccine ( MenAfriVac™ ) in the meningitis belt over the next few years , the use of the health centre catchment areas as spatial units can also help to monitor more accurately the vaccine supply at a finer spatial scale , saving doses that could be given inadequately , and to evaluate its impact and protective efficacy in the population ( herd immunity ) at the same level . Although this vaccine brings new hope to the control of meningitis epidemics , reactive vaccination with polysaccharide vaccines and research to improve control strategies will still be needed in the coming years , since it will take several years to immunize against the A the vulnerable population across the belt and since other serogroups like W135 may replace meningitis A as the dominant serogroup [21] . New decision criteria will have to be found for reactive vaccination . With the additional use of a finer spatial scale like the HCCAs , an interesting strategy would be real-time cluster detection , with prospective space-time scan statistic [22] or other existing methods [23] . In the context of a resource-limited country , this study can also assist public health authorities in their decision-making regarding resource allocation . The spatial clusters detected in our study were located in different HCCAs from year to year , but nine of the 42 districts were more recurrently affected by clustering of MM cases . Thus , these findings provide approaches to better adjust allocation of resources , including a ready supply of antibiotics and rapid diagnostic tests [24] , [25] , as well as additional health care personnel . In order to reduce the reaction time of the vaccination , one may consider allocating vaccines to these districts' hospitals prior to the meningitis season , provided the cold chain can be maintained . Given cost and organizational constraints , further cost-effectiveness and feasibility analyses are needed to evaluate this strategy , before any policy recommendation . Clusters were more often found in nine districts , including five bordering Nigeria within a 500 km distance between Doutchi and Aguie , most likely because of intense mobility of border populations [26] . However , no consistent annual spatio-temporal pattern could be found over the study period; hence , no spread in a systematic geographical direction from a fixed source could be identified . This is contrary to a study carried out in Mali , which highlighted a potential south-north spread , with Bamako and Mopti as probable sources [12] . Instead , our results suggest the emergence of scattered sources , likely from a pool of carriers when conditions are favorable to the occurrence of the invasive disease . Favorable conditions may include climatic conditions occurring during the dry season ( low absolute humidity and dust-laden Harmattan wind ) , which would damage the nasopharyngeal mucous membrane and increase the risk of bloodstream invasion by a colonizing meningococcus [27] . In this study , we observed that the latest spatio-temporal clusters during the epidemic season were often the northernmost ones , which could be correlated with the northward advance of the Intertropical Front preceding the arrival of rains from the south , thus raising relative humidity . However , climatic factors do not entirely explain these spatio-temporal epidemic patterns . As suggested by Mueller's hypothetical explanatory model [20] , their role may be limited to the hyperendemic increase during the dry season , while transition from a hyperendemic state to highly localized epidemics may be due to increased transmission , possibly caused by viral respiratory co-infections . Moreover , in equivalent climatic conditions , an area in which the proportion of susceptible individuals is higher due to waning immunity ( acquired by infection or vaccination ) would be more prone to outbreaks [28] . Recently , Irving et al [29] suggested that population immunity may be a key factor in causing the unusual epidemiology of meningitis in the Belt . Although density and distance to primary roads were not individually correlated with MM incidence at the HCCA level , other socio-demographic factors ( poverty , overcrowded housing , migrations , markets… ) may also have an influence on local transmission of the bacteria and carriage and contribute to the risk of micro-epidemics of co-infections [20] . Of note , one spatio-temporal cluster of four adult cases was detected in February 2009 in Bilma district ( see figure 1 ) , in the oasis town of Dirkou , located on an important south-north route of trans-Saharan trade and transit migration . Meningococcal strain variations most likely play a role in the occurrence of epidemic waves [20] , [30] , [31] . In this study , the spatio-temporal distribution of all N . meningitidis cases was analysed irrespective of the serogroups . A subsequent analysis will differentiate serogroups of meningococci as their spatio-temporal patterns may significantly vary [32] , [33] . Further etiologic studies are needed to explore causality of the spatio-temporal patterns highlighted in this paper . Finally , our findings provide an evidence-based approach to reflect on public health policies and indicate a promising strategy to improve prevention and control of meningitis in sub-Saharan Africa . They can serve as an example for other meningitis belt countries , illustrating what finer scale surveillance and spatial analyses can offer for prevention and control of meningitis . Research efforts should now focus on investigating the role of dust , socio-demographic factors , co-infections and vaccination strategies on cluster occurrence at the HCCA level , and on developing an operational decision support tool to respond better to meningitis outbreaks with the introduction of the new conjugate vaccine . | Meningococcal meningitis ( MM ) is an infection of the meninges caused by a bacterium , Neisseria meningitidis , transmitted through respiratory and throat secretions . It can cause brain damage and results in death in 5–15% of cases . Large epidemics of MM occur almost every year in sub-Saharan Africa during the hot , dry season . Understanding how epidemics emerge and spread in time and space would help public health authorities to develop more efficient strategies for the prevention and the control of meningitis . We studied the spatio-temporal distribution of MM cases in Niger from 2002 to 2009 at the scale of the health centre catchment areas ( HCCAs ) . We found that spatial clusters of cases most frequently occurred within nine districts out of 42 , which can assist public health authorities to better adjust allocation of resources such as antibiotics or rapid diagnostic tests . We also showed that the epidemics break out in different HCCAs from year to year and did not follow a systematic geographical direction . Finally , this analysis showed that surveillance at a finer spatial scale ( health centre catchment area rather than district ) would be more efficient for public health response: outbreaks would be detected earlier and reactive vaccination would be better targeted . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [
"medicine",
"bacterial",
"diseases",
"infectious",
"diseases",
"geography",
"public",
"health",
"and",
"epidemiology",
"computer",
"science",
"epidemiology",
"earth",
"sciences",
"public",
"health",
"geoinformatics",
"human",
"geography"
] | 2012 | Analysing Spatio-Temporal Clustering of Meningococcal Meningitis Outbreaks in Niger Reveals Opportunities for Improved Disease Control |
Eastern Equine Encephalitis ( EEE ) ( Togaviridae , Alphavirus ) is a highly pathogenic mosquito-borne arbovirus that circulates in an enzootic cycle involving Culiseta melanura mosquitoes and wild Passeriformes birds in freshwater swamp habitats . Recently , the northeastern United States has experienced an intensification of virus activity with increased human involvement and northward expansion into new regions . In addition to its principal role in enzootic transmission of EEE virus among avian hosts , recent studies on the blood-feeding behavior of Cs . melanura throughout its geographic range suggest that this mosquito may also be involved in epizootic / epidemic transmission to equines and humans in certain locales . Variations in blood feeding behavior may be a function of host availability , environmental factors , and/or underlying genetic differences among regional populations . Despite the importance of Cs . melanura in transmission and maintenance of EEE virus , the genetics of this species remains largely unexplored . To investigate the occurrence of genetic variation in Cs . melanura , the genome of this mosquito vector was sequenced resulting in a draft genome assembly of 1 . 28 gigabases with a contig N50 of 93 . 36 kilobases . Populations of Cs . melanura from 10 EEE virus foci in the eastern North America were genotyped with double-digest RAD-seq . Following alignment of reads to the reference genome , variant calling , and filtering , 40 , 384 SNPs were retained for downstream analyses . Subsequent analyses revealed genetic differentiation between northern and southern populations of this mosquito species . Moreover , limited fine-scale population structure was detected throughout northeastern North America , suggesting local differentiation of populations but also a history of ancestral polymorphism or contemporary gene flow . Additionally , a genetically distinct cluster was identified predominantly at two northern sites . This study elucidates the first evidence of fine-scale population structure in Cs . melanura throughout its eastern range and detects evidence of gene flow between populations in northeastern North America . This investigation provides the groundwork for examining the consequences of genetic variations in the populations of this mosquito species that could influence vector-host interactions and the risk of human and equine infection with EEE virus .
Eastern equine encephalitis ( EEE ) virus ( Alphavirus , Togaviridae ) causes severe disease in humans and equines with high case mortality and persistent neurologic impairment in survivors [1] . Historically , outbreaks of this relatively rare but highly pathogenic arthropod-borne virus occurred intermittently in the eastern United States , predominantly in the mid-Atlantic and Gulf Coast states such as Florida , as well as in isolated foci in the northeast [2–6] . However , since the early 21st century , this region has experienced a recurring seasonal intensification of EEE virus activity [5 , 7] , and a northward geographic expansion [8–11] . In the northeastern United States , EEE virus is maintained in an enzootic transmission cycle in freshwater swamp foci involving the ornithophilic mosquito Culiseta melanura ( Coquillett ) ( Diptera: Culicidae ) and passerine birds [12–15] . Human and equine disease cases occur predominantly in close proximity to the freshwater swamp habitats in which Cs . melanura breeds [2 , 12] . Culiseta melanura is distributed throughout eastern North America [16] and is widely considered the principal enzootic vector of EEE virus . This mosquito species exhibits considerable variability in avian host choice across geographic regions , favoring Passeriformes birds [6 , 15 , 17–20] . In addition to avian hosts , recent studies indicate that between 1 and 11% of Cs . melanura bloodmeals originate from mammalian hosts including humans [6 , 17–20] , suggesting the involvement of this mosquito species in epidemic/epizootic transmission of EEE virus to humans and equines [17 , 21] . Moreover , in a survey of thirty-five mosquito species from the northeastern United States , Cs . melanura was the only species with the consistently high titers needed for transmission of the virus [21] , and this species is among the predominant sources of virus isolations from field-collected mosquitoes in this region [21] . The apparent flexibility in host choice exhibited by Cs . melanura in various geographic regions [6 , 15 , 17–19] may be a function of environmental factors and/or underlying genetic variation that could influence vector-host interactions and potentially vectorial capacity . Despite the importance of Cs . melanura in transmission and maintenance of EEE virus , the genetics of this species remain largely unexplored . The present study was undertaken to gain insight into possible genetic variation in Cs . melanura populations that may contribute to its involvement in epidemic as well as enzootic transmission of EEE virus in certain locales . The specific objectives of the study were to: 1 ) characterize the genetic diversity of Cs . melanura in EEE virus foci across eastern North America , 2 ) investigate the occurrence of genetic structure among populations of Cs . melanura , and 3 ) examine patterns of gene flow among these populations .
Sequencing of forty SMRT Cells yielded a total of 4 . 2 million raw reads with an average read length of 9 kilobases ( kb ) . These raw reads were assembled to produce an unpolished draft assembly of 1 . 278 gigabases ( Gb ) across 25 , 500 contigs , with a contig N50 of 93 . 23 kb . Subsequent polishing provided modest increase to assembly size and contig N50 . The Quiver-polished assembly was 1 . 279 Gb , with a contig N50 of 93 . 29 kb , while the Arrow-polished assembly was a total of 1 . 281 Gb , with a contig N50 of 93 . 36 kb . Although polishing provided only minor improvements to assembly size and N50 , improvement to the results of BUSCO analyses were more substantial . The unpolished genome assembly had evidence of 61 . 3% of expected complete single copy orthologs , with 22 . 4% missing entirely . More than 70% of complete single copy orthologs were detected in both polished genomes , with the Arrow-polished assembly performing the best in BUSCO analyses; BUSCO results for this assembly found more than 75% of expected single copy orthologs , and in total , at least partial evidence for 88 . 7% of expected single copy orthologs ( Full BUSCO Results: Complete , Single Copy: 75 . 4%; Complete , Duplicate: 4 . 5%; Fragmented: 8 . 8%; Missing: 11 . 3%; Fig B in S1 File ) . This Arrow-polished assembly performed similarity to other publicly available genomes on VectorBase ( Fig 3 ) . Polymerase chain reaction duplicates comprised 5 . 29% ( SD = 4 . 0% ) of reads across libraries . Of these , 71 . 9% ( SD = 10 . 7% ) were single duplications , although some reads were duplicated hundreds of times ( Fig C in S1 File ) . Following removal of PCR duplicates and quality filtering , libraries resulted in 16 . 4 ( SD = 4 . 3 ) million reads per mosquito , which aligned to the draft reference genome at a mean rate of 93 . 2% ( SD = 3 . 8% ) . One mosquito from Canada had only 337 , 113 total reads and was removed from all downstream analyses , resulting in a total of 119 individuals from 10 geographic locations . Mean read depth across loci was 20 . 4X ( SD = 5 . 8X ) for the reference-aligned dataset , and 14 . 9X ( SD = 4 . 3X ) when Stacks was used for de novo assembly of loci ( Table B in S1 File ) . In total , Stacks identified 3 . 9 million variant loci across all samples for the reference-aligned dataset , and 8 . 7 million variant loci for the de novo dataset . Datasets varied in number of SNPs and genotyping rates ( Table 2 ) . Datasets in which p was set contained the largest number of SNPs , with more overall SNPs recovered as p declined , but with a lower genotyping rate . Fewer SNPs were recovered with Stacks’ de novo assembly of loci . The results presented below are for the variants called from the reference-aligned dataset , and without a specification of the population parameter p , which contained a total of 40 , 384 SNPs following the filtering of minor alleles and linked loci; specification of the population parameter to different values resulted in qualitatively similar results in downstream analyses , as did the de novo-assembled dataset ( see Supporting Information for examples ) .
To facilitate the study of genetic variation in Cs . melanura , a draft genome was generated . This genome is comparable in presence of single copy orthologs to many other currently available assemblies of Dipteran vector genomes . The relatively large size of the Cs . melanura draft assembly , at over 1 . 24 Gb , is consistent with a previous estimate of the genome size of this species based on cytophotometry at 1 . 2 Gb [24] . Moreover , this is also consistent with the genome sizes of other mosquitoes in the subfamily Culicinae [24 , 60 , 61] , such as Aedes aegypti where current genome assembly , AaegL5 , is just under 1 . 28 Gb [28 , 61] . By comparison , the vectors of human malaria in the genus Anopheles have significantly smaller genomes: The current Anopheles gambiae assembly , AgamP4 , is under 0 . 27 Gb . The draft assembly presented here was constructed principally as a reference for use in population studies of Cs . melanura , and further investigations are underway to improve the genome through additional sequencing and transcriptomics , a necessary first step to annotating the genome beyond the well-curated catalog of Dipteran single copy orthologs used in BUSCO analyses . As Cs . melanura is only the fourth non-Anopheline mosquito genome to be sequenced [60–62] , further development of this genome could provide important insight into the ecology , evolution , and control of this and other vector species , as the availability of genomic data for vector species has enabled the characterization of pathways important to host seeking behavior ( e . g . odorant receptors [63 , 64] ) , description of components of mosquito immunity [65–67] , and the design of novel vector control methods [68–70] . The draft genome assembled in this study is a useful tool for population studies , as demonstrated here . Reads aligned to the reference at a mean rate greater than 93% , indicating the relative completeness as a reference . Utilization of the draft genome allowed the capture of a larger number of variants than under de novo variant calling with Stacks ( Table 2 ) , and with significantly less computation time . Thus , the draft assembly holds promise as a tool in future population studies , either as a reference for additional genotyping-by-sequencing , or for discovery of microsatellite markers , a low-cost alternative for the processing of larger sample sizes . The SNP dataset in this study recovered more than forty thousand markers passing filtering that were present in at least 75% of individuals . This dataset enabled the fine-scale differentiation of Cs . melanura populations ( see Results , and below ) , despite the relatively small sample size . This is consistent with previous studies with empirical [71] and simulated [72] SNP data sets indicating that even sample sizes between two and ten [71–73] individuals per population can recover small degrees of significant genetic differentiation among populations , as long as at least 1500 SNPs are used [71] . Thus , the sample sizes used in this study should be adequate to evaluate even fine-scale genetic differentiation given the recovery of more than forty thousand SNPs . Moreover , this dataset provides an important framework for future studies that could expand on the sampling here , as RAD-seq datasets can be combined across laboratories with a high overlap of markers recovered , assuming identical library construction methods [74] . This study identified both regional differences between northern and southern populations of Cs . melanura , as well as fine scale population structure in this species . Pairwise Fst , AMOVA analyses , SNMF clusters , and PCA results all supported that populations within geographic regions of north ( 1-CAN , 2-ME , 3-NH , 4-VT , 5-NY , 6-MA , 7-CT ) and south ( 9-VA and10-FL ) were more akin to one another than to populations in other regions ( Fig 4 , Tables D and E in S1 File ) . These results suggest a degree of genetic differentiation at a geographic scale , although future studies are needed to confirm this differentiation and investigate its consequences with additional sampling , particularly in the southern United States , where our sampling was confined to two locations . FineRADstructure identified finer-scale population structure than did the estimation of individual ancestry coefficients via the least squares-based method SNMF , consistent with results in other systems ( e . g . with human SNP data [48] ) , as well as with simulations [47 , 48] , indicating that the MCMC algorithm used by fineSTRUCTURE and fineRADstructure can be more sensitive than other clustering algorithms . The fine-scale population structure detected by fineRADstructure is consistent with the significant molecular variation and pairwise Fst values among populations , as well as the ability of DAPC to assign most individuals back to their correct population . Although fineRADstructure was unable to cluster individuals from three populations into single population clusters ( 3-NH , 4-VT , 7-CT ) , DAPC successfully assigned many of these individuals to their respective populations at a rate similar to several other northern sites , suggesting underlying genetic variation differentiating these populations . The genetic variation among populations identified here could , among other factors , influence behavioral characteristics exhibited by Cs . melanura such as heterogeneities in host choice across regions [6 , 15 , 18 , 19] , but correlating the fine-scale population structure reported here with previously observed host feeding patterns requires further investigations . Of note , human-derived blood meals from Cs . melanura have been identified in populations across the eastern United States , including from mosquitoes trapped near several locations utilized in this study , albeit at low frequencies ( e . g . near 4-VT [17] , 6-MA [19] , 8-NJ [75] , and 10-FL [18] ) . However , these populations belong to multiple genetic clusters identified in the present study , indicating that a propensity to feed on humans might be widespread across genetic clusters . Given the importance of this mosquito vector , the influence of population structure warrants further investigation , such as through the simultaneous sampling of engorged mosquitoes and SNP data across populations . Moreover , our results are consistent with a post glacial recolonization of the northeast from southern populations following the retreat of the Laurentide ice sheet twelve to eighteen thousand years ago [76 , 77] , and the melting of permafrost from the mid-Atlantic at this time period ( e . g . near southern New Jersey [78] ) . The genetic cluster associated with southern populations ( 9-VA and 10-FL in Cluster B , Figs 4 and 5 ) is basal to Clusters C and D containing northern populations ( 1-CAN , 2-ME , 3-NH , 4-VT , 5-NY , 6-MA , 7-CT ) and the mid-Atlantic population ( 8-NJ ) in the MAP tree ( Fig 5 ) . Although the MAP tree is only an approximate guide to relatedness of populations [47 , 48] , MAP trees perform well in simulations at inferring ancestral history between populations [48] . Moreover , southern populations ( 9-VA , 10-FL ) had more private alleles than northern populations ( 1-CAN , 2-ME , 3-NH , 4-VT , 5-NY , 6-MA , 7-CT ) , indicating a higher genetic diversity ( Table 3 ) . The decline in private alleles south to north , as well as limited genetic variation among northern populations ( e . g . the limited population structure , low pairwise Fst values , and lower number of private alleles ) , would be consistent with colonization northward and subsequent loss of genetic diversity through founder effects and/or bottlenecking events , the occurrence of which could be investigated in future studies . Thus , genomic evidence is consistent with the northward expansion of Cs . melanura from southern source populations . To confirm the pattern observed here , future studies with additional sampling throughout the range of this mosquito species could infer evolutionary history of these populations through Bayesian analyses estimating divergence time and biogeographic/demographic history [79 , 80] . Substantial evidence of shared ancestry among populations was detected , likely indicating the presence of gene flow and/or shared ancestral polymorphism , particularly at the regional scales of northern ( 1-CAN , 2-ME , 3-NH , 4-VT , 5-NY , 6-MA , 7-CT ) and southern ( 9-VA , 10-FL ) populations . A lack of isolation by distance ( Table C in S1 File ) , as well as lower pairwise Fst values among populations within a region compared to pairwise Fst values between northern and southern regions suggest not only a degree of differentiation between regions but could also indicate contemporary gene flow or a history of connectivity among populations [81 , 82] . At least one individual from northern sampling sites ( 1-CAN , 2-ME , 3-NH , 4-VT , 5-NY , 6-MA , 7-CT ) had high ancestry with a different genetic cluster from the majority of those mosquitoes sampled at the same site , suggesting a degree of gene flow among genetic clusters ( Fig 4 ) . Moreover , the coancestry matrix and associated MAP tree provides additional evidence of gene flow among populations , as several individuals had higher coancestry with alternative genetic clusters and geographically distant populations in the north compared with other individuals from the same sampling location ( e . g . two individuals from 1-CAN with higher coancestry to mosquitoes from 5-NY in Cluster C , Fig 5 ) . It is unlikely that this signal of gene flow is the result of direct dispersal among populations sampled in this study , as the mean flight range of Cs . melanura , estimated via mark and recapture , ranged from 4 to 9 kilometers [83] . Instead , these results suggest Cs . melanura in swamp habitats are not isolated at a regional scale , and there exists a large degree of connectivity among populations in northeastern North America . In addition to the aforementioned fine-scale population structure , an unexpected genetic cluster was identified ( Fig 4 , Figs D , F , G in S1 File , Table E in S1 File ) . This cluster was comprised primarily of a small number of individuals from 2-ME and 3-NH; however , it also contributed substantial ancestry to individuals in other northern populations ( particularly 1-CAN , 5-NY; Fig E in S1 File ) . As 2-ME and 3-NH were sequenced in separate sequencing lanes ( Table 1 ) and had similar coverage and rates of alignment to the reference genome ( Table B in S1 File ) it seems unlikely this cluster is a sequencing artifact . Subsequent sequencing of ITS2 confirmed previous morphological identification that individuals belonging to Cluster A were indeed Cs . melanura , as pairwise identity of ITS2 sequences was >99% between individuals belonging to this cluster and all other Cs . melanura sequenced . Although the similarity in pairwise ITS2 sequence identity between members of Cluster A and other Cs . melanura may seem contrary to genomic results , this discrepancy is likely due to the greater number of variable sites in the SNP dataset , compared to only 297 nucleotide positions in the ITS2 alignment . Mosquitoes belonging to Cluster A were relatively differentiated compared to mosquitoes belonging to other genetic clusters , with pairwise Fst values greatly exceeding those among other populations ( Table E in S1 File ) . In addition , this cluster had far more private alleles than any population ( Table 3 ) . These results are consistent with evidence of an influx of genetic variation from a source not sampled in this study , such as admixture between northern populations of Cs . melanura and an unsampled and highly diverged population of this species , or interspecific hybridization with another closely related mosquito . However , due to the small number of individuals in this cluster , the lack of any previous evidence of hybridization between this species and others , and the paucity of sequencing information for other Culiseta species , it is not feasible with the available data to assess hypotheses regarding this cluster . Further studies investigating northeastern North American Cs . melanura populations with a larger sampling of mosquitoes , particularly targeting populations near 2-ME and 3-NH , may elucidate the origins of this cluster and any potential impacts it may have on behavioral characteristics of this mosquito species e . g . host feeding pattern and the risk of human infection with EEE virus . The results presented here provide the first insight into genetic variation in Cs . melanura . Although the existence of fine-scale structure and gene flow among populations was detected in these analyses , correlation between our findings and observed human and equine disease risk is difficult at this time . Future studies integrating sampling of populations along with mosquito bloodmeal analysis and observations of host diversity and abundance would be more informative to answering whether the genetic structure identified here is correlated with heterogeneities in blood host choice and/or disease risk across the region . By utilizing the new genomic resources available from this study , future investigations may build on the framework this study represents and provide greater insight into the vector-host interactions in Cs . melanura and its associated disease risk to humans and equines . | Eastern equine encephalitis ( EEE ) is a highly pathogenic mosquito-borne virus responsible for outbreaks of severe disease in humans and equines , causing high mortality and neurological impairment in most survivors . In the past , human disease outbreaks in the northeastern United States occurred intermittently; however , recently , this region has experienced a recurring seasonal intensification of EEE virus activity with expansion into more northerly locales . Eastern equine encephalitis virus is maintained in a transmission cycle involving the ornithophilic mosquito , Culiseta melanura , and wild passerine birds in freshwater swamp foci . Recent studies on the blood feeding behavior of Cs . melanura suggest this mosquito species could also be involved in transmission of EEE virus to humans and equines . Since variation in host feeding of this mosquito vector may be a function of environmental factors and/or genetic differences among regional populations , we examined the population genetics of Cs . melanura in order to: 1 ) characterize the genetic diversity of Cs . melanura at EEE virus foci across eastern North America , 2 ) investigate the occurrence of genetic structure among populations of Cs . melanura , and 3 ) examine patterns of gene flow among these populations . We generated a draft genome of this species as a reference for population studies , identified evidence of significant genetic differentiation and fine-scale genetic structure among populations , and found evidence of gene flow among northeastern populations . This study provides the molecular basis for future investigations on the causes and consequences of the genomic variation in Cs . melanura on the risk of human and equine infection with EEE virus . | [
"Abstract",
"Introduction",
"Results",
"Discussion"
] | [
"invertebrates",
"medicine",
"and",
"health",
"sciences",
"population",
"genetics",
"genomic",
"library",
"construction",
"animals",
"invertebrate",
"genomics",
"genome",
"analysis",
"dna",
"construction",
"molecular",
"biology",
"techniques",
"population",
"biology",
"insect",
"vectors",
"research",
"and",
"analysis",
"methods",
"infectious",
"diseases",
"genetic",
"polymorphism",
"comparative",
"genomics",
"molecular",
"biology",
"disease",
"vectors",
"insects",
"animal",
"genomics",
"arthropoda",
"mosquitoes",
"eukaryota",
"dna",
"library",
"construction",
"heredity",
"genetics",
"biology",
"and",
"life",
"sciences",
"species",
"interactions",
"genomics",
"evolutionary",
"biology",
"gene",
"flow",
"computational",
"biology",
"organisms"
] | 2018 | Population genomics of Culiseta melanura, the principal vector of Eastern equine encephalitis virus in the United States |
The type 2 immune response is the central mechanism of disease progression in schistosomiasis , but the signals that induce it after infection remain elusive . Aberrant microRNA ( miRNA ) expression is a hallmark of human diseases including schistosomiasis , and targeting the deregulated miRNA can mitigate disease outcomes . Here , we demonstrate that efficient and sustained elevation of miR-203-3p in liver tissues , using the highly hepatotropic recombinant adeno-associated virus serotype 8 ( rAAV8 ) , protects mice against lethal schistosome infection by alleviating hepatic fibrosis . We show that miR-203-3p targets interleukin-33 ( IL-33 ) , an inducer of type 2 immunity , in hepatic stellate cells to regulate the expansion and IL-13 production of hepatic group 2 innate lymphoid cells during infection . Our study highlights the potential of rAAV8-mediated miR-203-3p elevation as a therapeutic intervention for fibrotic diseases .
Schistosomiasis is a serious but neglected tropical infectious disease , affecting more than 230 million people worldwide [1] . Hepatic granuloma and secondary fibrosis caused by lodged eggs from the parasite are the primary cause of morbidity and mortality from this chronic and debilitating disease . Elucidating the mechanisms that initiate hepatic schistosomiasis has been a major research objective for decades , and it is now well-established that hepatic schistosomiasis is an immune pathological disease [2 , 3] . A major breakthrough was the identification of type 2 immune response , characterized by the T helper 2 cell associated cytokines such as interleukin 4 ( IL-4 ) and IL-13 , as a central regulator of disease progression in schistosomiasis [2 , 3] . However , the signals that induce type 2 immune response after infection remain elusive . Quiescent hepatic stellate cells ( HSCs ) are located in the subendothelial space , between the anti-luminal side of sinusoidal endothelial cells and the basolateral surface of hepatocytes , and are characterized by their cytoplasmic vitamin droplets [4] . When liver injury occurs , quiescent HSCs are activated to become proliferative , contractile , and fibrogenic myofibroblasts [5] . Activated HSCs produce excessive extracellular matrix ( ECM ) that is deposited in the liver , and are the main effector cells in various types of hepatic fibrosis , including fibrosis induced by schistosome infection [6] . In addition , more recent studies have uncovered the fundamental role of HSCs in hepatic inflammation and immunity [7 , 8] . MicroRNAs ( miRNAs ) are endogenous , small noncoding RNAs which control the activity of more than 30% of protein-coding genes through target mRNA degradation or translational inhibition [9 , 10] . Increasing evidence has demonstrated that miRNAs are involved in regulating almost every cellular process , and aberrant miRNA expression is a hallmark of many human disorders , including infectious diseases [11 , 12] . Several studies by ours and other groups have shown that miRNAs play a crucial role in the pathogenesis of schistosomiasis and may serve as useful therapeutic targets [13–15] . In particular , one of our previous studies has shown that depletion of a single miRNA , miR-21 , in the liver protects hosts from lethal infection through attenuation of hepatic fibrosis [15] . In this study , we used a murine model of Schistosoma japonicum ( S . japonicum ) to investigate the role of miR-203-3p , a miRNA down-regulated following infection in the progression of hepatic schistosomiasis [15] . We found that recombinant adeno-associated virus 8 ( rAAV8 ) mediated elevation of miR-203-3p in the liver protected mice from lethal infection through alleviating type 2 pathology . Importantly , our data indicate that miR-203-3p targets IL-33 , an inducer of type 2 immunity [16 , 17] , in HSCs to regulate the expression of IL-13 in hepatic group 2 innate lymphoid cells ( ILC2s ) during infection .
We previously identified more than thirty deregulated host miRNAs by expression profiling during the progression of hepatic schistosomiasis . This includes miR-203-3p as the most down-regulated [15] . To examine the role of miR-203-3p in schistosomiasis in vivo , mice were first challenged with a lethal dose of S . japonicum cercaria and then intravenously injected with either rAAV8-pri-miR-203-3p vector sustainedly expressing the miRNA , control vector , or PBS at day 10 post-infection . We found that a single dose of rAAV8-pri-miR-203-3p protected infected mice from the lethal effect of schistosomiasis . Six of ten mice receiving rAAV8-pri-miR-203-3p survived to the end of the study ( i . e . 80 days; Fig 1A ) . In contrast , the majority of mice receiving control vector ( n = 10 ) or PBS ( n = 10 ) died within 9 weeks post-infection ( Fig 1A ) . Hepatic granuloma and fibrosis , induced by host type 2 immune response resulting from liver-trapped parasite eggs , are the primary cause of morbidity and mortality from this disease [2 , 3] . Thus , we next investigated if the rAAV8-pri-miR-203-3p-mediated intervention was indeed through effective elevation of miR-203-3p activity that in turn attenuated the type 2 pathology . To this end , mice were exposed to a mild dose of parasites and then treated with vectors or PBS . Our data revealed that the level of miR-203-3p in the rAAV8-pri-miR-203-3p treated group was significantly higher than in the control groups ( Fig 1B ) . Excessive ECM deposition is the main feature of hepatic fibrosis . By 6 weeks post-infection , mice receiving rAAV8-pri-miR-203-3p displayed a significant reduction in ECM deposition as shown by hydroxyproline quantification ( Fig 1C ) and Masson’s trichrome staining ( Fig 1D and 1F ) , whereas the size of hepatic granulomas in all groups was similar as shown by H&E staining ( Fig 1E and 1F ) . Reduction of fibrosis was further confirmed by qPCR-based quantification of fibrosis associated gene expression in the livers of infected mice . Amounts of Col1α1 , Col3α1 , and α-Sma mRNA were dramatically reduced in livers of mice treated with rAAV8-pri-miR-203-3p ( Fig 1G , 1H and 1I ) . In addition , we detected the alteration of the cytokines that are associated with type 2 immune response in liver tissues after elevation of miR-203-3p . Consistent with the antenuated fibrotic phenotype , a strong reduction in mRNA levels of Il13 and Tgf-β1 was detected in livers of mice treated with rAAV8-pri-miR-203-3p ( Fig 1J and 1K ) . However , levels of other cytokines , including interferon-γ ( Ifn-γ ) , tumor necrosis factor-α ( Tnf-α ) , Il4 , and Il5 , were not significantly altered ( S1A Fig ) . Consistent with our previous study , the virus vector did not affect the survival and egg production of parasites in the hosts ( S1B Fig ) , and virus delivery in the liver showed no significant differences between groups ( S1C and S1D Fig ) . HSCs are the predominant cellular source of ECM during hepatic fibrosis . Thus , we investigated whether the anti-fibrotic effect of rAAV8-pri-miR-203-3p intervention directly modulated the activity of HSCs . To this end , we isolated primary HSCs from infected mice after administration with vectors to quantify mRNA levels of miR-203-3p , Col1α1 , Col3α1 , and α-Sma . Our data showed that significantly decreased miR-203-3p expression and increased collagen and α-Sma expression were observed in the infected mice without rAAV8-pri-miR-203-3p treatment compared with uninfected mice ( S2 Fig ) . As expected , miR-203-3p expression was clearly elevated , while collagen and α-Sma expressions were distinctly reduced in HSCs after rAAV8-pri-miR-203-3p intervention ( S2 Fig ) , suggesting that miR-203-3p could modulate the activation of HSCs in vivo . To investigate whether elevation of miR-203-3p in the liver can reverse the parasite egg-induced hepatic fibrosis , mice were infected with a mild dose of parasites . At 42 days after infection , when hepatic fibrosis was clearly manifest , mice were treated with praziquantel to kill the parasite , then injected with either rAAV8 vectors or PBS , and necropsied at 70 days post-infection ( Fig 2A ) . Again , the expression of miR-203-3p was significantly increased in the rAAV8-pri-miR-203-3p treated mice ( Fig 2B ) . Importantly , hydroxyproline quantification and Masson’s trichrome staining revealed that fibrosis in rAAV8-pri-miR-203-3p treated mice was markedly reduced compared with controls ( Fig 2C , 2D and 2F ) , but the size of hepatic granulomas in all groups was similar as shown by H&E staining ( Fig 2E and 2F ) . This was confirmed by reduced expression of Col1α1 , Col3α1 and α-Sma in these mice ( Fig 2G , 2H and 2I ) . Of cytokines tested , only Il13 mRNA was reduced ( Fig 2J and 2K ) , and livers showed no significant change in egg burden ( Fig 2L ) . Considering that elevation of miR-203-3p attenuates type 2 pathology , we speculated that miR-203-3p could regulate the initiation of type 2 immunity after infection . IL-33 , an IL-1-related cytokine , is an inducer of type 2 immunity in several organs [18] , and is a potential target of miR-203-3p , predicted by TargetScan database . To analyze the relationship between miR-203-3p and IL-33 , we evaluated their expression during the progression of hepatic schistosomiasis . We found that expression of miR-203-3p began to decrease in the liver at day 32 post-infection , reaching its lowest levels at day 42 and 52 ( Fig 3A ) ; in contrast , the level of Il33 mRNA was maintained during the early stage of infection , then significantly elevated by day 42 post-infection ( Fig 3B ) . In addition , we investigated the expression of miR-203-3p and Il33 mRNA in different hepatic cell compartments , including hepatocytes , HSCs , and Kupffer cells ( KCs ) . Our data showed that , similar to the expression pattern in whole liver , the expression of miR-203-3p in hepatocytes and HSCs began to decrease at day 42 post-infection , while the level of Il33 mRNA was elevated at the same time ( Fig 3C and 3D ) . However , the expression of both miR-203-3p and Il33 mRNA in KCs was unchanged during the observed time ( Fig 3C and 3D ) . We also characterized the relative abundance of miR-203-3p and Il33 mRNA in different hepatic cell compartments , and found that , in both the uninfected and infected livers , miR-203-3p was selectively expressed in hepatocytes and HSCs ( Fig 3E ) , whereas Il33 was primarily expressed in HSCs ( Fig 3F and 3G ) . To further validate that activated HSCs could be a source of IL-33 in infected livers , we carried out immunochemistry staining for IL-33 and α-SMA , and we observed that both factors were mainly located in the periphery of egg granulomas ( S3 Fig ) . Double staining using immunofluorescence displayed a co-localization of IL-33 and α-SMA staining ( Fig 3H ) , indicating that activated HSCs express IL-33 in vivo . In addition , we detected the expression of miR-203-3p and Il33 during the progression of HSC activation in vitro . Resting HSCs will be automatically activated when cultured on a plastic surface [4] . We found that , when primary HSCs from uninfected mice were cultured on plastic plates , expression of miR-203-3p in these cells was significantly reduced , while the level of Il33 mRNA was significantly elevated , accompanied by a dramatic increase in collagen expression ( S4 Fig ) . Taken together , these results suggest that the activated HSCs in infected livers are a source of IL-33 , and that IL-33 is a potential target of miR-203-3p in HSCs . To test whether IL-33 is a direct target of miR-203-3p , we first generated a reporter construct that contains the firefly luciferase gene fused to the 3’ UTR from Il33 mRNA containing a putative miR-203-3p target site ( Fig 4A ) . This construct was transiently transfected into 293T cells together with miR-203-3p mimics or a negative control miRNA . We observed a marked reduction in luciferase activity in cells transfected with miR-203-3p mimics together with Il33-UTR ( Fig 4B ) . In contrast , mutation of 5 nt in the miR-203-3p seed sequence led to a complete abrogation of reporter inhibition ( Fig 4B ) . We transfected miR-203-3p mimics or inhibitors into primary HSCs from uninfected mice , and quantified the level of IL-33 by qPCR or western blot . Our data revealed that , at both the mRNA and protein levels , elevation of miR-203-3p distinctly reduced the expression of IL-33 , while depletion of miR-203-3p significantly increased the expression of IL-33 ( Fig 4C and 4D ) . Finally , we analyzed IL-33 levels in primary HSCs from infected livers treated with rAAV8-pri-miR-203-3p , and found that IL-33 expression was markedly reduced ( Fig 4E and 4F ) . Taken together , these data indicate that IL-33 is a direct target of miR-203-3p in HSCs . In addition , we noticed that the target site of miR-203-3p in the 3’UTR of Il33 gene is not conserved from mouse to human . However , we provided evidence that human IL-33 is also a direct target of miR-203-3p in the HSCs ( S5 Fig ) . Though a number of cell types were suggested as sources of IL-13 in response to IL-33 stimulation , a recent study has proved that ILC2s , instead of lymphocytes , basophils , or mast cells , are the predominant source of IL-13 in the liver after IL-33 stimulation [19] . Having found that IL-33 is a target of miR-203-3p and that elevation of miR-203-3p leads to a reduction of IL-13 in the liver , we hypothesized that down-regulation of miR-203-3p during the progression of hepatic schistosomiasis could lead to higher expression of IL-13 in ILC2s via increased levels of IL-33 in HSCs . To address this , we first analyzed the number of hepatic ILC2s and production of IL-13 by hepatic ILC2s during infection using flow cytometry ( S6 Fig ) . Our data indicated that ILC2s were a primary source of IL-13 production in the infected livers ( S7 Fig ) , and both the number of ILC2s and production of IL-13 in these cells were increased by day 32 post-infection , peaking at day 42 post-infection ( S8A Fig ) . This initial elevation occurred prior to the elevation of IL-33 in HSCs ( Fig 3D and S8B Fig ) . STAT6 phosphorylation , a marker of IL-13 pathway activation , also began to increase in HSCs , peaking at the same time point , and the production of collagen in HSCs was dramatically elevated at day 42 post-infection ( S8B Fig ) . Moreover , we observed that , following elevation of miR-203-3p in HSCs using rAAV8-pri-miR-203-3p vectors , the number of ILC2s and production of IL-13 in these cells were markedly reduced ( Fig 5A and 5B ) . The expression of IL-33 ( Fig 4E and 4F ) , the phosphorylation of STAT6 ( Fig 5C ) , and production of collagen ( S2 Fig ) in HSCs were all significantly decreased . Finally , we found that purified primary HSCs , stimulated with recombinant IL-13 ex vivo , responded by phosphorylation of STAT6 and production of collagen in a time-dependent manner ( Fig 5D and 5E ) . Taken together , these data indicate that miR-203-3p regulates the expression of IL-13 in ILC2s by targeting IL-33 in HSCs , thus modulating the expression of ECM by HSCs , during the progression of hepatic schistosomiasis .
In this study , we demonstrate that efficient and sustained elevation of miR-203-3p in liver tissues , using the highly hepatotropic rAAV8 , protects mice against lethal schistosome infection by alleviating hepatic fibrosis . Importantly , we show that miR-203-3p targets IL-33 , an inducer of type 2 immunity , in HSCs to regulate the expansion and IL-13 production of hepatic ILC2s during infection . Type 2 immune response , featured by elevation of IL-4 and IL-13 levels , plays a crucial role in host protection as well as pathological tissue fibrosis after helminth infection , including schistosome infection , but the signals that induce type 2 immunity are poorly understood . Recently , numerous studies have highlighted that tissue damage , which induces the release of cytokine alarmins such as IL-33 , is a potent mechanism driving type 2 immunity , particularly in the context of helminth infection [20 , 21] . The role of IL-33 in schistosomiasis has also been intensively studied , but published reports have been inconsistent . Mchedlidze et al . reported that , in the animal model of Schistosoma mansoni ( S . mansoni ) infection , IL-33 was critical for inducing the development of IL-13-dependent hepatic fibrosis [19] . This phenomenon was also observed in the animal model of S . japonicum infection [22] . However , a recent study showed that IL-33 needed to synergize with two other cytokine alarmins , thymic stromal lymphopoietin ( TSLP ) and IL-25 , in the regulation of IL-4/IL-13-dependent inflammation or fibrosis after S . mansoni infection [23] . The inconsistency might be due to the difference of intervention time: when IL-33 is depleted in the embryo stage , the function of IL-33 might be compensated by other factors; but when IL-33 is inhibited during the progression of diseases , the role of IL-33 in disease progression become obvious . In this study , our data also indicated that IL-33 is crucial for inducing the progression of type 2 pathology after infection , as significant reductions in hepatic fibrosis and IL-13-producing ILC2s were observed upon down-regulation of IL-33 in the liver , which resulted from rAAV8-mediated elevation of miR-203-3p . These studies , including our current study , have established that IL-33 is a crucial mediator in the maintenance of type 2 pathology induced by schistosome infection . Importantly , our study revealed other important aspects of the role of IL-33 in the regulation of type 2 pathology after infection . Although IL-33 is involved in the progression of many human diseases , little is known about the regulation of its expression . Our study , for the first time , revealed that IL-33 is regulated by miRNAs . IL-33 is a direct target of miR-203-3p in HSCs , and downregulation of miR-203-3p leads to elevated levels of IL-33 in HSCs , initiating type 2 pathology after infection . Previous studies have demonstrated that miR-155 can regulate the expansion and IL-13 production of ILC2 in the context of IL-33 [24] , and miR-29a can regulate IL-33 effector function via targeting its decoy receptor sST2 [25] . Thus , miRNA could be an important regulator in the initiation of type 2 pathology . Activated HSCs are a major source of IL-33 in infected livers . Our findings confirmed that IL-33-producing cells are located in the periphery of egg granulomas where activated HSCs produce excess collagen , and that expression of IL-33 in primary HSCs is significantly elevated after infection . These findings are consistent with previous studies , which demonstrated that activated HSCs and pancreatic stellate cells ( PSCs ) are major sources of IL-33 in the fibrotic liver and pancreas , respectively , of both mice and human [26–28] . These studies suggest that HSCs or PSCs might be the sentinel cells in the organs , which detect the injury signals and promote wound healing response by releasing damage-associated molecular pattern such as IL-33 . Our data showed that the initial elevation of IL-13 in hepatic ILC2 cells ( day 32 post-infection ) occurs prior to the elevation of IL-33 in whole livers and HSCs ( day 42 post-infection ) . It is reported that cell death by necrosis or active necroptosis , instead of active secretion , might be the dominant mechanism by which IL-33 reaches the extracellular milieu [29] . Therefore , we speculate that the source of IL-33 that activates hepatic ILC2s at day 32 after infection derives from the necrosis or active necroptosis of HSCs . Here , we created a schematic diagram showing the molecular mechanism underpinning the regulation of type 2 pathology after infection by miR-203-3p ( Fig 6 ) : The toxic challenge derived from parasite eggs trapped in the liver tissue induces the down-regulation of miR-203-3p in HSCs , which relieves the inhibition to IL-33 . Sequential elevation of IL-33 is released into the liver tissue and stimulates the proliferation and IL-13 production of hepatic ILC2s . IL-13 then activates HSCs to produce excessive ECMs through activation of STAT6 pathway . Thus , our study highlights the crucial role of miR-203-3p and its target IL-33 in the initiation of type 2 pathology during schistosome infection . It is noteworthy that IL-33 expresison in HSCs begins to elevate at day 42 after infection , thus , this mechanism mainly exerts its role in the progression of Th2 pathology after 42 days . In addition , IL-33 is reported to promote the development of fibrosis in many organs , including liver [19] , lung [30] , kidney [31] , heart [32] , skin [33] , and other organs [34] . Therefore , miR-203-3p might serve as a useful target in the treatment of these fibrotic diseases .
All animal experiments were performed in strict accordance with the Guide for the Care and Use of Laboratory Animals of the National Institutes of Health , and were approved by the Animal Ethics Committee of Second Military Medical University ( laboratory animal usage number FYXK ( Shanghai ) 2014–0003 ) . To minimize pain and discomfort , all animal surgeries were undertaken under sodium pentobarbital anaesthesia . Male BALB/c mice ( 6 week ) were obtained from the experimental animal center of Second Military Medicine University , and were housed under specific pathogen-free conditions and fed with autoclaved food and water as needed . To establish the animal model of schistosomiasis , mice were exposed percutaneously to 16 or 30 S . japonicum cercariae . For parasite perfusion , the portal vein was dissected at the root , then the thoracic cavity of the mouse was opened and the circulatory system was perfused via the aorta with sterile PBS . Parasites were collected and counted in a sterile petri dish containing medium . Subsequently , the liver was removed and snap-frozen in liquid nitrogen . For egg counting , part of the liver was digested overnight with 4% potassium hydroxide , then the total number of schistosome eggs was counted , and the liver egg burdens were defined as 104 eggs per gram of liver tissue . The size of hepatic granuloma was measured from Mayer’s H&E staining of sections using a calibrated measuring eyepiece , and the extent of fibrosis was analysed by Masson’s trichrome staining of sections as described previously [15] . All granulomas within each section were scored for blue density on a scale of 1–4 , and a second measurement of area involved was also determined using the same scale . The total fibrosis score was determined by multiplying the density and area for each granuloma ( a score of 16 would be the maximum ) . The hydroxyproline content in the liver was detected using a colourmetric assay kit according to the manufacturer’s instructions ( Nanjing Jiancheng Bioengineering Institute , Nanjing , China ) . The procedure was performed as described previously [15] . In short , HSCs were first isolated by density-gradient centrifugation and then further purified using negative selection with magnetic CD11b antibody beads ( MACS , Miltenyi , Auburn , CA ) ; Kupffer cells were first isolated by density-gradient centrifugation and then further purified using positive selection with magnetic CD11b antibody beads . Primary HSCs were cultured on plastic dishes in DMEM supplemented with 10% fetal bovine serum ( Hyclone ) , 4 mmol/L L-glutamine , penicillin ( 100 IU/ml ) , and streptomycin ( 100 mg/ml ) . Cells were maintained at 37°C , 5% CO2 in a humidified atmosphere . For transfection , HSCs were transfected with 40 nM miR-203-3p mimics ( Qiagen ) , miR-203-3p inhibitors ( Qiagen ) , or negative controls at day 3 after seeding using Lipofectamine 3000 ( Invitrogen ) . Total RNA was isolated using Trizol reagent ( Invitrogen ) according to the manufacturer’s protocol . Real-time PCR was performed as described previously [35] . The levels of miR-203-3p , Col1α1 , Col3α1 , α-Sma , Ifn-γ , Tnf-α , Tgf-β1 , Il4 , Il5 , Il10 , Il13 , and Il33 were detected using the SYBR Green Master Mix kit ( Roche ) . U6 snRNA or Gapdh was used as an internal control , and the fold change was calculated by the 2-ΔΔCt method . Sequences of primers used in this study are shown in S1 Table . Total cell protein was extracted on ice using RIPA lysis buffer in the presence of freshly added protease and phosphatase inhibitors ( Thermo ) , then quantified by the BCA method ( Pierce ) . A total of 30 μg protein extract per lane was loaded onto a 14% SDS-polyacrylamide gel and transferred to nitrocellulose membranes ( Pierce ) . Nonspecific binding was blocked with 5% nonfat milk in PBS . The membrane was incubated with rat anti-IL-33 ( R&D ) or rabbit anti-phospho-STAT6 ( Cell signaling ) antibody overnight at 4°C . IRDye 800CW goat anti-rabbit IgG or goat anti-rat IgG ( LI-COR ) was used as secondary antibody , and rabbit anti-GAPDH antibody ( Abcam ) was used as an internal standard . Immunohistochemistry was performed on formaldehyde-fixed , paraffin-embedded mouse livers . After hydration , liver sections were incubated with rat anti-IL-33 ( R&D ) or rabbit anti-α-SMA ( Abcam ) antibody for 1 hour at 37°C , and HRP or fluorescence conjugated secondary antibody ( Abcam ) was used to display the signals . Nonparenchymal cells isolated from liver were stimulated with PMA ( 50 ng/mL ) , Ionomycin ( 1 μg/mL ) , and BFA ( 3 μg/mL ) for 4 hours . Cells were surface stained with FITC conjugated lineage cocktail ( CD3 / Gr-1 / CD11b / CD45R / Ter-119 / Siglec-f / CD11c / NK1 . 1 ) ( Biolegend ) , PerCP/Cy5 . 5 conjugated ICOS ( Biolegend ) and APC conjugated ST2 ( Biolegend ) , permeabilized with 0 . 1% spaonin buffer for 15 minutes , and further stained with PE conjugated IL-13 ( eBioscience ) before acquiring with FACS Calibur . ILC2s are defined as Lineage ( - ) ST2 ( + ) ICOS ( + ) cells in this study . Data were analyzed in Flowjo . The human or mouse wild-type or mutant 3’ UTRs of IL-33 containing the predicted miR-203-3p binding sites were synthesized ( South Gene Technology , China ) and cloned into the pGL3 . 0-control vectors according to the manufacturer’s instructions ( Promega ) . 293T cells were seeded in 24-well plates , then transfected with 40 nM miR-203-3p or a negative control ( Qiagen ) and co-transfected with 0 . 8 μg per well wild-type IL-33 3’ UTR-luc or mutant IL-33 3’UTR-luc , using Lipofectamine 3000 ( Invitrogen ) according to the manufacturer’s instructions . pRL-TK vectors ( 0 . 1 μg per well ) were co-transfected as endogenous controls for luciferase activity . After 24 h , cells were lysed , and luciferase activities were measured using a dual-luciferase assay kit ( Promega ) . To express miR-203-3p , pri-miR-203-3p fragment was amplified by PCR from C57/B6 mouse genomic DNA using primer pri-miR-203-3pF ( 5´AACAGGTCCTCGCACAGAGTGCAGCCCGGC 3´ ) and pri-miR-203-3pR ( 5´AACAGGTCCTCCACCCCCGCGCCCCTCTCA3´ ) , then cloned into the PpuMI restriction site in the intron of pscAAVCBPI GLuc plasmid [36] . The identity of pri-miR-203-3p was verified by sequencing . rAAV8 vectors used in this study were generated , purified , and tittered as described [37] . All analyses were carried out with the SPSS 19 . 0 software . Data were shown as mean ± s . d . The significance of difference between two groups was identified using a Student’s t-test . Multiple comparisons were performed by one-way ANOVA , and followed by Bonferroni post test for comparison between two groups . Survival between different groups was compared by Kaplan–Meier survival curves with log-rank test . P values less than 0 . 05 were considered significant . | Schistosomiasis is a serious but neglected tropical infectious disease . Hepatic fibrosis caused by lodged eggs from the parasite is the primary cause of morbidity and mortality from this disease . Type 2 immune response , featured by the T helper 2 cell associated cytokines such as IL-4 and IL-13 , is the central regulator of disease progression in schistosomiasis , but the signals that induce it after infection remain elusive . Aberrant expression of miRNAs underlies a spectrum of human diseases , including infectious diseases . In this study , using a well-studied murine model of human schistosomiasis , we show that Schistosoma infection down-regulates the miR-203-3p expression , and that IL-33 , an inducer of type 2 immunity , is a direct target of miR-203-3p in hepatic stellate cells . The reduced miR-203-3p leads to elevated levels of IL-33 , promoting the expansion and IL-13 production of hepatic group 2 innate lymphoid cells and thus initiating type 2 pathology . Importantly , rAAV8-mediated elevation of miR-203-3p in liver tissues protects mice against lethal schistosome infection by alleviating hepatic fibrosis . Thus , our study highlights the crucial role of miR-203-3p in the initiation of type 2 pathology during schistosome infection , and suggests miR-203-3p as a potential target for fibrotic diseases . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] | [
"schistosoma",
"invertebrates",
"medicine",
"and",
"health",
"sciences",
"helminths",
"gene",
"regulation",
"tropical",
"diseases",
"fibrosis",
"parasitic",
"diseases",
"animals",
"liver",
"diseases",
"collagens",
"animal",
"models",
"developmental",
"biology",
"micrornas",
"model",
"organisms",
"gastroenterology",
"and",
"hepatology",
"experimental",
"organism",
"systems",
"neglected",
"tropical",
"diseases",
"research",
"and",
"analysis",
"methods",
"proteins",
"liver",
"fibrosis",
"gene",
"expression",
"mouse",
"models",
"biochemistry",
"rna",
"helminth",
"infections",
"schistosomiasis",
"eukaryota",
"nucleic",
"acids",
"genetics",
"biology",
"and",
"life",
"sciences",
"non-coding",
"rna",
"organisms"
] | 2018 | Down-regulation of microRNA-203-3p initiates type 2 pathology during schistosome infection via elevation of interleukin-33 |
The plant hormone auxin plays a critical role in regulating various aspects of plant growth and development , and the spatial accumulation of auxin within organs , which is primarily attributable to local auxin biosynthesis and polar transport , is largely responsible for lateral organ morphogenesis and the establishment of plant architecture . Here , we show that three Arabidopsis INDETERMINATE DOMAIN ( IDD ) transcription factors , IDD14 , IDD15 , and IDD16 , cooperatively regulate auxin biosynthesis and transport and thus aerial organ morphogenesis and gravitropic responses . Gain-of-function of each IDD gene in Arabidopsis results in small and transversally down-curled leaves , whereas loss-of-function of these IDD genes causes pleiotropic phenotypes in aerial organs and defects in gravitropic responses , including altered leaf shape , flower development , fertility , and plant architecture . Further analyses indicate that these IDD genes regulate spatial auxin accumulation by directly targeting YUCCA5 ( YUC5 ) , TRYPTOPHAN AMINOTRANSFERASE of ARABIDOPSIS1 ( TAA1 ) , and PIN-FORMED1 ( PIN1 ) to promote auxin biosynthesis and transport . Moreover , mutation or ectopic expression of YUC suppresses the organ morphogenic phenotype and partially restores the gravitropic responses in gain- or loss-of-function idd mutants , respectively . Taken together , our results reveal that a subfamily of IDD transcription factors plays a critical role in the regulation of spatial auxin accumulation , thereby controlling organ morphogenesis and gravitropic responses in plants .
Auxin is a key plant hormone that plays critical roles in the regulation of plant growth and development . A combination of physiological , genetic , biochemical , and molecular studies has greatly enriched our understanding of auxin biosynthesis , transport , and signal transduction [1]–[3] . Increasing evidence indicates that auxin is essential for nearly all developmental processes , including gametogenesis , embryogenesis , lateral organ formation and patterning , branching , and tropic responses [4] , [5] . It is generally believed that most auxin-mediated developmental events are highly dependent on the differential accumulation of auxin within plant organs ( auxin gradients ) , which are mainly attributable to both local auxin biosynthesis and the intercellular polar transport of auxin [1] , [5] . Direct evidence that local auxin biosynthesis is involved in the regulation of plant organogenesis comes from studies of several genes in the auxin biosynthetic pathway of Arabidopsis , including YUCCA ( YUC ) and TRYPTOPHAN AMINOTRANSFERASE of ARABIDOPSIS ( TAA ) . YUCs encode the flavin monooxygenases that catalyze a key step in converting tryptophan into IAA , a main auxin in plants [6] . Overexpression of YUC genes in Arabidopsis substantially elevates the endogenous IAA level and causes distinct phenotypes such as epinastic cotyledons , elongated hypocotyls , and narrow and curly leaves [6]–[8] . Although a yuc single mutant in Arabidopsis does not show an obvious phenotype , the mutation of multiple YUC genes leads to a diversity of auxin-related phenotypes , including reduced apical dominance , crinkled leaves , simple venation , and abnormal flower development , demonstrating that YUC-modulated local auxin biosynthesis is critical for plant morphogenesis and architecture formation [9] , [10] . Consistently , such developmental defects in yuc mutants can be rescued by local expression of iaaM , a bacterial auxin biosynthetic gene , but not by the application of exogenous auxin [9] . The expressions of YUC genes are overlapping and spatiotemporally regulated in various organs [9] , [10] , suggesting that YUCs function redundantly and cooperatively in different organs . Further studies demonstrate that TAA1 and its homologs function in auxin biosynthesis in response to environmental and developmental signals in Arabidopsis . The taa1 plant has a decreased level of endogenous IAA and displays defects in shade avoidance and root-specific ethylene sensitivity , and the simultaneous mutation of TAA1 and its close homologs ( TAR1 and TAR2 ) results in phenotypes that are obviously auxin-related , such as reduced gravitropic response of roots , shortened hypocotyls , and monopteros-like seedlings with a single cotyledon [11]–[13] . Recently , YUCs and TAAs have been shown to function in a key two-step auxin biosynthetic pathway [14]–[16] . Therefore , the regulation of YUCs and TAAs may represent an important mechanism to alter auxin gradients and thus modify plant morphogenesis and responses to environmental cues . Indeed , several transcription factors have been shown to participate in the regulation of YUC expression . STYLISH1 ( STY1 ) , one of the SHORT INTERNODES ( SHI ) transcription factors and NGATHA3 ( NGA3 ) , a member of the B3 transcription factor family , are both proposed to act cooperatively to direct style development , partially through the activation of YUC2- and YUC4-mediated auxin biosynthesis in the apex of the gynoecium [17] , [18] . LEAFY COTYLEDON2 ( LEC2 ) , a B3 domain transcription factor acting as a central regulator of embryogenesis , has been found to be involved in modulating de novo auxin biosynthesis via the activation of YUC2 and YUC4 [19] . Recent studies have also revealed that PIF4-activated YUC8 expression is required for hypocotyl growth under high temperature conditions [20] , and a putative transcription factor SPOROCYTELESS/NOZZLE ( SPL/NZZ ) has been found to be involved in the regulation of lateral organ morphogenesis by repressing transcription of YUC2 and YUC6 [21] . In addition to local auxin biosynthesis , polar auxin transport also contributes greatly to auxin redistribution . The polar auxin transport is mediated by AUX1/LIKE AUX1 ( AUX1/LAX ) , PIN-FORMED ( PIN ) , and ATP-binding cassette class B ( ABCB ) transporters [2] , [22]–[24] . Mutation in AUX1/LAX , which encode auxin influx carriers , influences lateral root formation , root hair development , phyllotaxy , and tropic responses [22] , [24]–[28] . Similarly , single or multiple mutations in auxin efflux carrier genes , PINs and ABCBs , lead to severe defects in multiple developmental processes including embryogenesis , lateral organ formation and patterning , vascular development , and tropic responses [29]–[33] , demonstrating that polar auxin transport is also essential for plant morphogenesis and tropic responses . The INDETERMINATE DOMAIN ( IDD ) transcription factors belong to a plant-specific transcription factor family that contains a conserved ID domain with four zinc finger motifs [34] , [35] . The founding member of this family , maize INDETERMINATE1 ( ID1 ) , is a key regulator of flowering transition [36] . The Arabidopsis genome contains 16 IDD members , and the characterization of several IDD members in Arabidopsis has indicated that IDD genes are involved in regulation of multiple developmental processes . For example , IDD8/NUTCRACKER ( NUC ) is involved in the regulation of flowering time through modulation of sugar metabolism [37] , and IDD14 regulates starch metabolism in response to cold stimulus [38] . IDD15/SHOOT GRAVITROPISM 5 ( SGR5 ) is involved in the gravitropic response of inflorescence stems , possibly through alteration of gravity sensing [39]–[41] . IDD8/NUC , IDD3/MAGPIE ( MAG ) , and IDD10/JACKDAW ( JKD ) regulate root development and patterning [42] , [43] , while IDD1/ENHYDROUS ( ENY ) participates in seed maturation and germination [44] . Recent studies have demonstrated that a rice ID1 homolog is required for flowering induction [45]–[47] , and that the rice Loose Plant Architecture1 ( LPA1 ) , an ortholog of Arabidopsis IDD15 , influences shoot gravitropism and architecture [48] . Here , we identified a gain-of-function mutant of Arabidopsis IDD14 , idd14-1D , which exhibited small and transversally down-curled leaves . Further characterization of loss- and gain-of-function mutants of IDD14 and its close homologs , IDD15 and IDD16 , revealed that these three IDD genes redundantly but differentially regulate multiple aspects of organ development and gravitropic responses . Moreover , we provide evidence that these IDD transcription factors can directly target auxin biosynthetic and transport genes and alter auxin accumulation in multiple organs . These results reveal that IDD-mediated auxin biosynthesis and transport are critical for lateral organ morphogenesis and gravitropic responses in plants .
To gain insight into how lateral organ morphogenesis is controlled in plants , we screened for mutants with altered aerial organ morphology in a transgenic Arabidopsis population harboring a T-DNA activation-tagging plasmid ( pSKI015 ) [49] . A semi-dominant mutant was identified by its smaller and dramatically down-curling leaf phenotype , and thus designated as curlyfolia1-D ( cuf1-D ) and subsequently as idd14-1D ( see below ) ( Figure 1A ) . To examine the genetic nature of cuf1-D , we backcrossed cuf1-D with wild type ( WT ) plants . All F1 plants exhibited an intermediate phenotype as cuf1-D/+ , and F2 plants displayed a phenotypic segregation of WT∶cuf1-D/+∶cuf1-D as 1∶2∶1 ( 82∶173∶87 , P = 0 . 89 , X2-test ) , in which all WT plants were BASTA sensitive . This observation suggests that cuf1-D results from a semi-dominant mutation of a single gene that is likely to co-segregate with a T-DNA insertion event . The most striking phenotype in cuf1-D was the leaf size and shape . Detailed quantification showed that the average areas of fully-expanded leaf blades in heterozygous and homozygous cuf1-D were only about 70% and 47% of that in WT , respectively ( Figure 1B ) . The leaves of cuf1-D were also dramatically curled downward in a transverse direction . The leaf transverse curvature ( TC ) index in WT was about 0 . 08 , whereas the leaf TC index in heterozygous and homozygous cuf1-D reached about 0 . 29 and 0 . 36 , respectively ( Figure 1C ) . Moreover , the lamina of a mature WT rosette leaf displayed an elliptical shape with a leaf length/width index at about 1 . 5 , while the heterozygous and homozygous cuf1-D leaves were comparatively narrow and their leaf indices were about 1 . 9 ( Figure 1D ) . In addition to the morphological changes in leaves , cuf1-D was also late flowering and dwarfed ( Figure 1A , Table S1 ) . These observations demonstrate that the dominant mutation in cuf1-D has pleiotropic effects on lateral organ development and plant architecture . Since cuf1-D is a single gene mutation that is likely to co-segregate with a T-DNA insertion event , we amplified the genomic DNA adjacent to the left border of the T-DNA by thermal asymmetric interlaced PCR ( TAIL-PCR ) . The sequencing analysis indicated that a T-DNA was inserted in the intergenic region between At1g68120 and At1g68130 ( Figure 2A ) . Genotyping analysis showed that the T-DNA insertion co-segregated with the leaf phenotype in heterozygous and homozygous cuf1-D mutants ( Figure 2B ) , suggesting that cuf1-D was caused by the T-DNA insertion event . Because cuf1-D was a dominant mutant , we monitored the transcripts of genes flanking the T-DNA insertion using semi-quantitative reverse transcription PCR ( RT-PCR ) analysis . Compared to those in the WT , the transcripts of At1g68130 ( IDD14 ) were dramatically elevated , while At1g68140 mRNA levels were slightly decreased in cuf1-D ( Figure 2C ) . To determine whether the elevated levels of IDD14 transcripts are responsible for the cuf1-D phenotype , we introduced a p35S::anti-IDD14 construct into cuf1-D and a p35S::IDD14 construct into WT plants , respectively . Transgenic plants overexpressing IDD14 fully recapitulated the phenotype of cuf1-D , while the expression level of At1g68140 in these plants was comparable with that in WT plants ( Figure 2D , S1B ) . Moreover , introduction of p35S::anti-IDD14 into cuf1-D restored the cuf1-D to WT morphology ( Figure 2D ) . These results demonstrate that cuf1-D results from the ectopic expression of IDD14 , and accordingly , the cuf1-D was thus re-designated as idd14-1D . In Arabidopsis , the IDD14 transcription factor is phylogenetically sub-grouped with two close homologs: IDD15 ( SGR5 ) and IDD16 , to form a small subfamily that is distinct from the other IDD family members ( Figure S1A ) [34] , [39] . IDD14 , IDD15 and IDD16 share 52%–62% amino acid identity , and their ID domains are highly conserved with 89%–95% amino acid identity ( Figure 2E ) . To investigate the possible redundancy of IDD15 and IDD16 with IDD14 , we generated transgenic Arabidopsis plants overexpressing IDD15 or IDD16 , respectively . The ectopic expression of either IDD15 or IDD16 resulted in a similar leaf phenotype as observed in idd14-1D ( Figure 2F , S1B ) , suggesting that the three IDD members may have redundant function during plant development . To explore the functions of this IDD subfamily , we examined the tissue-specific expression patterns of these genes in multiple organs of transgenic plants harboring a pIDD::GUS ( β-Glucuronidase ) construct . As shown in Figure 3A–3C , IDD14 was mainly expressed in cotyledons and the vasculature of rosette leaves , and a weak level of expression was observed in hypocotyls and floral organs . However , the GUS signal was undetectable in roots and inflorescence stems . IDD15 was highly expressed in petioles , hypocotyls , roots , floral organs , and especially in inflorescence stems . In inflorescence stems , GUS staining was mainly present in the cortex , endodermis and vasculature tissues ( Figure 3A–3C ) . IDD16 was highly expressed in leaves , hypocotyls , roots , vasculature of cotyledons , floral organs , and in the endodermis and vasculature of inflorescence stems ( Figure 3A–3C ) . RNA in situ hybridization assayed in the inflorescence stems validated the expressions of the three IDD genes detected by the GUS reporter ( Figure 3D ) . Additionally , consistent with the previous finding that IDD members act as transcription factors [38] , [39] , an IDD14-GFP fusion protein in transgenic plants carrying p35S::IDD14-GFP , which recapitulated the phenotype of idd14-1D , was found to be localized in nuclei ( Figure S2A , S2B ) . To gain further insight into the functions of this IDD subfamily , we obtained the T-DNA insertion mutant idd14-1 ( CS367164 ) and idd15-5 ( Salk_087765 ) from the Arabidopsis Biological Resource Center ( ABRC ) , in which the T-DNA is inserted in an exon of IDD14 or IDD15 . Semi-quantitative RT-PCR analyses indicated that the transcripts of IDD14 or IDD15 were undetectable in idd14-1 or idd15-5 , respectively ( Figure 4A , 4B ) . As no idd16 mutant is publically available , we generated IDD16-RNAi transgenic plants with an IDD16-specific cDNA fragment ( Figure 4A ) . Semi-quantitative RT-PCR analysis indicated that IDD16 transcripts were dramatically reduced in transgenic IDD16-RNAi plants ( Figure 4B ) . We then generated double and triple mutants of the three IDD genes , which allowed us to closely examine their differential and redundant functions during plant growth and development . We first examined the aerial organ morphogenesis in these idd mutants . As shown in Figure 4C and 4D , none of the idd single mutants exhibited any obvious organ phenotype . However , the leaves of idd14-1 IDD16-RNAi and idd triple mutants were not only downward-curled in the longitudinal direction , but were also more rotund when compared to WT leaves . Furthermore , idd14-1 IDD16-RNAi and the idd triple mutant had enlarged floral organs and infertile siliques . Careful examination showed that the infertile siliques resulted from the asynchronous elongation of stamen filaments and styles , and thus had poorly pollinated stigmas ( Figure 4C , 4E ) . Manual pollination of styles in these mutants resulted in the development of normal siliques . In contrast to these dramatic phenotypes , the idd15-5 IDD16-RNAi plants had only slightly curled leaves ( Figure 4C , 4D ) . These observations imply that IDD14 and IDD16 have redundant roles in directing leaf and floral organ morphogenesis . Consistent with the previous finding that idd15 displays increased angles between inflorescence stems and siliques [40] , we noticed that the orientation angles of both branches and siliques were obviously increased in idd15-5 ( Figure 4C ) . This phenotype was further enhanced in the idd15-5 IDD16-RNAi and idd triple mutants , but not in idd14-1 idd15-5 plants ( Figure 4C , 5A ) , indicating that IDD15 and IDD16 act cooperatively to control silique and branch orientation . As idd15 has a reduced gravitropic response in inflorescence stems [39]–[41] , it is likely that the altered orientation of branches and siliques is related to the gravitropism defect in the idd mutants . To test this , we investigated the gravitropic responses in gain- and loss-of-function of idd plants . As expected , idd15-5 inflorescence stems exhibited an obviously reduced gravitropic response , and this phenotype was greatly enhanced in idd15-5 IDD16-RNAi and the idd triple mutant ( Figure 5B , S3A ) , demonstrating that IDD15 and IDD16 function coordinately in gravitropic responses . Interestingly , although idd14-1 did not show any defect in gravitropic response , the inflorescence stems of idd14-1D were hypersensitive to gravistimulation ( Figure 5B , S3A ) , indicating that ectopic expression of IDD14 also influences gravitropic responses . Surprisingly , although the IDD genes were found to be expressed in hypocotyls and roots , we did not observe any obvious phenotype in hypocotyls or roots of either idd14-1D or the idd triple mutant , including gravitropic response phenotypes ( Figure S3B , S3C ) . The only aberration we observed was a slightly waved primary root phenotype in the idd triple mutant when grown vertically ( Figure S3D ) . These findings suggest that the three IDDs may primarily function in lateral aerial organs . Previous studies have suggested that IDD14 and IDD15 are involved in the regulation of starch metabolism [38] , [40] . However , the narrow , epinastic leaves in gain-of-function IDD mutant or transgenic plants seem to resemble , to some extent , those observed in auxin overproduction mutants or transgenic plants , such as yucD or p35S::YUCs transgenic plants [6] , [7] , [9] . By contrast , the rotund and curly rosette leaves , abnormal floral phenotype , and gravitropism defect in the loss-of-function idd mutants are also documented in the mutants defective in auxin biosynthesis or transport [2] , [9]–[11] . This led us speculate that auxin accumulation or signaling may also be involved in IDD-mediated organ development and/or gravitropic response . To test this , we used the idd14-1D and idd triple mutant plants as representatives of gain- and loss-of-function idd mutants for our further analysis . We first examined auxin accumulation in their leaf , inflorescence stem , and root , by monitoring the expression of DR5∶GUS or DR5∶GFP , a widely used auxin gradient reporter [50] . On day 3 after leaf initiation , an obvious GUS signal was observed in WT leaf distal tips . A stronger and spatially-expanded GUS signal was observed in idd14-1D leaf tips . No obvious GUS signal , and thus no localized auxin maxima , was observed in the leaf tips of idd triple mutant plants ( Figure 6A ) . Similar differential GUS patterns were subsequently observed in 5-day-old expanding and 15-day-old expanded leaves ( Figure 6A ) . In inflorescence stems , the GUS signal was mainly observed in vascular tissues in WT , whereas strong GUS staining was present in the cortex and endodermis tissues in idd14-1D and weak GUS expression without a tissue-specific pattern was observed in the idd triple mutant ( Figure 6B ) . The increased or decreased auxin accumulation was also found in the meristem regions of primary roots in idd14-1D and the idd triple mutant , respectively ( Figure 6C ) . Further quantification of the GUS activity in these organs confirmed the variations of DR5∶GUS signals observed in idd14-1D and the idd triple mutant ( Figure 6D ) . This observation strongly suggests that alteration of IDD expression affects auxin accumulation in multiple organs . Auxin homeostasis and transport are key factors that determine the accumulation of auxin in plant organs [1] , [2] , [4] . To assess the contributions of auxin homeostasis and transport to IDD-mediated auxin accumulation , we first quantified the endogenous free IAA levels in the idd14-1D and idd triple mutants . Consistent with enhanced or decreased expression of DR5∶GUS reporter observed , the endogenous IAA level was increased by about 13% in idd14-1D but decreased by approximately 19% in idd triple mutant plants as compared to that in WT ( Figure 6E ) . This result implied that the IDD genes may affect auxin biosynthesis . We then measured the auxin transport capability of inflorescence stems in idd14-1D and idd triple mutant . As shown in Figure 6F , the basipetal IAA transport efficiency was increased by over 50% in idd14-1D stems but decreased about 18% in idd triple mutant stems when compared with that in WT stem , demonstrating that IDD also modulates the auxin transport process . To further examine whether auxin signaling is affected in gain- and loss-of-function idd mutants , we monitored the expression of the DR5∶GUS reporter , IAA5 , and IAA29 in response to exogenous IAA treatment in idd14-1D and the idd triple mutant , and observed that the expression of the DR5∶GUS reporter , IAA5 , and IAA29 was normally induced following application of IAA as that in WT plants ( Figure S4A , S4B ) , suggesting that IDD has no effect on auxin perception or signaling . We also observed that the expression of IDD14 , IDD15 , and IDD16 transcripts was not modulated by auxin treatment ( Figure S4C ) . Taken together , our results strongly suggest that the three IDD genes are involved in the establishment of auxin gradients through the regulation of auxin biosynthesis and transport . To identify the genes downstream of IDD , we first carried out a real-time quantitative RT-PCR ( qRT-PCR ) analysis to examine the transcript abundances of genes known to function in auxin biosynthesis and transport in the idd14-D and idd triple mutants . Among the YUC and TAA family genes , the transcription of YUC1 , YUC2 , YUC3 , YUC4 , YUC5 , YUC8 , and TAA1 was found to be elevated in idd14-1D , and the transcription levels of YUC2 and YUC5 were decreased in idd triple mutant , when compared to those in WT plants ( Figure 7A ) . Among 14 genes related to auxin transport , the transcription of AUX1 , PIN1 , ABCB1 , ABCB4 and WAG1 were found to be elevated in idd14-1D but reduced in the idd triple mutant compared to those in WT plants . PIN4 and PINOID ( PID ) transcripts were only elevated in idd14-1D or deceased in the idd triple mutant ( Figure 7A ) . We further investigated the expression of 11genes , which had the apparently altered expression in idd14-1D or the idd triple mutant ( with a significance at P<0 . 01 ) , in the transgenic plants overexpressing IDD15 or IDD16 and the idd15-5 IDD16-RNAi plants . As expected , their differential expressions in gain- and loss-of-function IDD15 and IDD16 plants were much similar to those observed in idd14-1D and the idd triple mutant ( Figure S5A ) , demonstrating that these genes are also downstream of IDD15 and IDD16 . As the IDDs are transcription factors , we speculated that some of the genes differentially expressed in gain- and loss-of-function idd mutants might be directly targeted by IDD . We first identified the genes whose expression was rapidly induced by the activation of IDD , using the chemically-inducible IDD14 transgenic plants . After the transgenic seedlings were treated with the inducer , IDD14 was dramatically induced by 0 . 5 h , and the expression of YUC5 , TAA1 , and PIN1 was obviously elevated within 2 h ( Figure S5B ) , suggesting that three genes might be direct targets of IDDs . To examine whether IDD can directly bind to the promoter regions of YUC5 , TAA1 , or PIN1 , we performed chromatin immunoprecipitation ( ChIP ) assays with both p35S::IDD16-GFP and p35S::IDD14-GFP transgenic plants . It has been reported that the maize ID1 and ID domain proteins could bind to a specific 11 bp DNA consensus motif , T-T-T-G-T-C-G/C-T/C-T/a-T/a-T [35] . As such , we targeted similar possible IDD-binding motifs in the promoter and/or upstream coding regions of YUC5 , TAA1 , PIN1 , YUC2 , and YUC3 , and carried out the ChIP analysis ( Figure 7B ) . As expected , we found that the three fragments containing a putative IDD-binding motif in YUC5 , TAA1 , or PIN1 were greatly enriched by IDD16 after GFP immunoprecipitation , whereas no binding activity was detected in the promoter regions of YUC2 or YUC3 ( Figure 7B ) . Such enrichment was detected in a control DNA fragment of the QUA-QUINE STARCH ( QQS ) promoter , which was previously reported to be a target of IDD14 [38] , but no enrichment was detectable in a control DNA fragment in the ACTIN2 ( ACT2 ) promoter which lacked the putative IDD-binding motif ( Figure 7B ) . Likewise , similar enrichments of these DNA fragments were further confirmed by ChIP assayed with IDD14 protein ( Figure S5C ) . In addition , we indeed visualized the enhanced or attenuated PIN1 accumulation in the roots of idd14-1D and idd triple mutant carrying a pPIN1::PIN1-GFP construct , respectively ( Figure S6 ) . These results illustrate that IDD can directly target YUC5 , TAA1 , and PIN1 , to activate their expression . Since idd14-1D contains a high auxin level while the idd triple mutant has low endogenous auxin content , we attempted to genetically modify auxin biosynthesis to examine whether this could suppress or restore the phenotype observed in gain- and loss-of-function idd mutants . We first generated an idd14-1D yuc2 yuc6 triple mutant through genetic crosses , and observed that loss-of-function yuc2 yuc6 completely suppressed the leaf phenotypes of idd14-1D ( Figure 8A , S7 ) , and partially attenuated the hypersensitivity of idd14-1D inflorescence stems to gravistimulation ( Figure 8B ) . Further , when we overexpressed YUC2 in the idd triple mutant , the ectopic expression of YUC2 fully rescued the infertile silique defect ( Figure 8C ) , but only partially restored the silique orientation in the idd triple mutant ( Figure 8D ) . These results provide further evidence that auxin biosynthesis is genetically downstream of this IDD subfamily . As the amyloplast movement in endodermal cells of idd15 stem has been reported to be defective under gravistimulation [39]–[41] , we also investigated whether ectopic expression of YUC2 has an effect on amyloplast movement in endodermal cells of the idd triple mutant stems . As shown in Figure S8 , the retarded amyloplast movement in endodermal cells of the idd triple mutant stems under gravistimulation was not rescued by ectopically expressed YUC2 ( Figure S8 ) , suggesting that altered auxin accumulation does not affect IDD-mediated amyloplast responsiveness to gravistimulation .
The IDD family has been defined as a plant-specific transcription factor family [35] , [51] , and previous characterizations of a few IDD members have indicated that IDDs are involved in the regulation of transition to flowering and starch metabolism [36]–[38] . IDD14 , IDD15 , and IDD16 belong to a subfamily that is distinct from the other IDD members in Arabidopsis , rice , and maize [34] . The mutation of IDD15 in Arabidopsis and rice reduces gravitropic response in inflorescence stems [39]–[41] , [48] . However , because the loss-of-function idd14 mutant does not have obvious organ phenotype , and IDD16 has not yet been characterized , the functions of this subfamily are still largely unknown . In this study , we characterized both gain- and loss-of-function mutants of this IDD subfamily , and discovered that these three IDD genes function redundantly and cooperatively in regulating organ morphogenesis and gravitropic responses . Gain-of-function of each IDD led to a small , narrow , and down-curled leaf phenotype . Although idd single mutants did not have obvious organ morphological phenotypes ( except the reduced gravitropic response observed in idd15 ) , our further characterization of the idd double and triple mutants clearly demonstrates that IDD14 and IDD16 act redundantly to regulate the morphology of aerial organs and affect fertility , while IDD15 and IDD16 cooperatively control the gravitropic responses and plant architecture . Such redundant and cooperative roles of the three IDD genes in organ morphogenesis and gravitropism are consistent with their differential and overlapping expression patterns in particular organs . IDD14 and IDD16 , but not IDD15 , were expressed in juvenile leaves , whereas IDD15 and IDD16 , but not IDD14 , were highly expressed in inflorescence stems . The three IDD genes were all expressed in floral organs , and the abnormal flower phenotype in idd14-1 IDD16-RNAi was enhanced in the idd triple mutant . Interestingly , although the IDD genes were also expressed in hypocotyls and roots , we could not observe obvious changes in their morphology or gravitropic response in either gain- or loss-of-function idd mutants , except the slightly waved roots in idd triple mutant seedlings . This may be attributable to decreased expression of WAG1 , a gene that is an indirect target of IDD and has been identified as a suppressor of root waving ( Figure 7A ) [52] . Previous studies have shown that IDD15 and IDD14 are involved in the regulation of starch metabolism [38] , [40] . Here , with a combination of phenotypic , genetic , and molecular approaches , we demonstrated that IDD14 , IDD15 , and IDD16 modulate auxin accumulation by affecting auxin biosynthesis and transport , thereby modifying organ morphogenesis and architecture formation . First , the organ phenotype in both gain- and loss-of-function idd mutants appears to be related to altered auxin homeostasis and distribution . For example , the narrow , epinastic leaves in the plants overexpressing IDD are similar to those in auxin overproduction mutants or transgenic plants , such as yucca ( yuc1D ) , yucca6-1D , and p35S::YUCs transgenic plants [6] , [7] , [9] , whereas the rotund and curly rosette leaves and abnormal floral phenotype in the loss-of-function idd mutants are similar to those observed in the mutants defective in auxin biosynthesis or transport [9] , [53] . Disruption of IDD genes also influenced the silique and branch angles , root waving , and gravitropism of inflorescence stems , which have been well documented to be related to polar auxin transport [2] , [54]–[57] . Second , expression analyses using DR5∶GUS and DR5∶GFP reporter clearly indicated that gain- or loss-of-function of IDDs enhanced or reduced auxin gradients , which was further confirmed by the increased or decreased endogenous auxin content and transport ability . Third , the expression of several genes involved in auxin biosynthesis and transport were altered in both gain- and loss-of-function idd mutants , and the IDD proteins could directly bind to the promoter regions of YUC5 , TAA1 , and PIN1 to activate their expression . In addition , genetic manipulation of auxin biosynthesis could fully or partially restore the pleiotropic phenotypes in idd14-1D or the idd triple mutant . These results demonstrate that IDD indeed modulates auxin gradients by promoting auxin biosynthesis and transport . The gravitropic response in plants requires a coordination of three sequential processes: gravity perception , signal transduction , and asymmetric growth response [58] . It is widely believed that the starch-filled amyloplasts ( statoliths ) within specific gravi-sensing cells ( statocytes ) perceive gravity stimulation [59] , [60] . Some other molecules , such as InsP3 and Ca2+ have been found to be involved in gravity signaling [60] . A large body of evidence indicates that auxin plays a key role in gravitropic signaling and asymmetric organ growth , and that it may possibly be involved in gravi-sensing [2] , [60] . For example , many mutants related to auxin biosynthesis and especially transport such as taa1 , aux1 , pin1 , and pin2 , exhibit a defect in gravitropic responses [2] , [11] , [61] . Other mutants with altered silique or branch architecture , such as plethora ( plt ) in Arabidopsis and lazy1 ( la1 ) in rice , also show altered auxin accumulation or transport within their organs [54] , [56] , [62] . Among the three IDD members we characterized , IDD15 and its rice ortholog , LPA1 , have been previously reported to affect the gravitropic response by altering amyloplast sedimentation in the endodermis [39]–[41] , [48] . Recently , IDD14 was also found to mediate starch degradation by directly activating the expression of QQS [38] , suggesting that IDD14 also participates in the regulation of starch metabolism . Our detailed characterization of gain- and loss-of-function of three IDDs provides substantial evidence that IDD-mediated auxin biosynthesis and transport contribute to the organ morphogenesis and also , to some extent , gravitropic responses , because the genetic manipulation of auxin biosynthesis does not alter the responsiveness of amyloplast to gravistimulation but partially restores the gravity sensitivity or defect in gain- or loss-of-function idd mutants . Therefore , it is likely that IDD-mediated auxin accumulation and starch metabolism coordinately control the gravitropic responses . The IDD-regulated starch metabolism might be primarily involved in gravi-sensing while the IDD-regulated auxin gradient may be primarily involved in signaling and responses .
The Arabidopsis thaliana accession Col-0 was used in this study . idd14-1D was isolated from a population generated by T-DNA activation-tagging mutagenesis . idd14-1 ( CS367164 ) and idd15-5 ( SALK_087765 ) were obtained from ABRC . All seeds were sterilized and geminated on 1/2 MS medium after vernalization for 2 days at 4°C , and the plants were grown in a culture room or growth chamber at 22±1°C with an illumination intensity of 80–90 µmol m−2 s−1 and a 16-h light/8-h dark photoperiod , as described previously [63] . Leaf curvature was determined as described previously [64] , [65] . Briefly , the transverse curvature ( TC ) index was defined as TC = 1- cw/pw , where cw and pw are the curved width and the pressed width of leaves , respectively . The longitudinal curvature ( LC ) index was defined as LC = 1- cl/pl , where cl and pl are the curved length and the pressed length of leaves , respectively . To quantify the pressed width or length of leaves , the sixth leaf of 25-day-old plants was dissected transversely or longitudinally on a desk , and then the extreme distances between the margins of the leaf were measured before and after pressing . To quantify the gravitropic responses of roots and hypocotyls , seedlings were grown vertically for 4 days and then turned horizontally , and curvature angles were measured [58] . For quantification of the gravitropic responses of inflorescence stems , 32-day-old plants with inflorescence stems of approximately 4–8 cm were gravistimulated by rotating them 90° in darkness , and stem curvatures were measured as the angles between the growing direction of apex and horizontal base line [39] , [58] . The T-DNA flanking sequence of idd14-1D was amplified by TAIL-PCR [66] . Three primers , P1 , P2 , and pSK-LB2 were used for co-segregation analysis . P1 and P2 were located in the Arabidopsis genome flanking the T-DNA insertion site , and pSK-LB2 was a primer corresponding to the left border of the T-DNA sequence . Similarly , idd14-1 and idd15-5 were genotyped with corresponding primers ( Table S2 ) , and yuc2 and yuc6 were genotyped according to the methods from a previous study [9] . To generate p35S::IDDs and p35S::anti-IDD14 transgenic plants , the IDD coding sequences were amplified by RT-PCR and ligated into the pGEM-T-Easy vector ( Promega , USA ) , and then verified by sequencing . The resulting plasmids were digested with EcoRI and cloned into pVIP96 [67] . The IDD14 cDNA was also cloned into pER8 to generate a chemically inducible IDD14 construct . To investigate the tissue-specific expression of IDD14 , IDD15 , and IDD16 , approximately 2-kb promoter fragments of IDD genes were amplified from genomic DNA and then fused with the β-glucuronidase ( GUS ) gene into pBI101 . To generate the p35S::IDD16-GFP and p35S::IDD14-GFP constructs , an IDD16 or IDD14 coding sequence lacking a stop codon was amplified and then cloned in frame into pMDC83 ( Invitrogen , USA ) . To generate the p35S::IDD16-RNAi construct , a specific IDD16 cDNA fragment was amplified and ligated inversely into pBluescript SK-GUSF [68] , and then an XbaI-BglII digested fragment was sub-cloned into pVIP96 . A YUC2 cDNA fragment was amplified by RT-PCR and cloned into pVIP96myc to generate the p35S::YUC2 construct . All the primers used are listed in Table S2 . All constructs were introduced into Arabidopsis by Agrobacterium tumefaciens–mediated transformation via the floral dip as described previously [69] . At least 20 independent lines harboring a single T-DNA insertion from each construct were generated , and 4–5 independent lines of T3 homozygous plants were used for detailed characterization . Total RNA was isolated using a guanidine thiocyanate extraction buffer [70] . For semi-quantitative RT-PCR or qRT-PCR analysis , cDNA was synthesized from 1 µg of total RNA using SuperScript III Reverse Transcriptase ( Invitrogen , USA ) . qRT-PCR was performed with a Rotor-Gene 3000 thermocycler ( Corbett Research , Australian ) with a SYBR Premix Ex Taq II kit ( Takara , Japan ) . The relative expression level for each gene was normalized to the ACTIN2 and the data were collected from three biological replicates , as described previously [49] . The histochemical GUS assay was carried out according to previously described protocol [71] . The GUS activities were quantified by monitoring cleavage of the β-glucuronidase substrate 4-methylumbelliferyl β-D-glucuronide ( MUG ) , as described [71] . For RNA in situ hybridization , a specific cDNA region of IDD14 , IDD15 , or IDD16 was transcribed in vitro to generate sense and antisense probes using the Digoxigenin RNA labeling kit ( Roche , Switzerland ) . The WT inflorescence stems were fixed and embedded in paraffin ( Sigma-Aldrich , USA ) , and then sectioned to a 10 µm thickness . RNA in situ hybridization was performed according to a previously described method [54] . Aerial organs from 15-day-old plants of WT , idd14-1D , and idd triple mutant plants were used for measurement of free IAA content . The extraction , purification , and analysis of free IAA by gas chromatography-mass spectrometry was performed according to the methods described by Edlund et al . [72] , except that an Agilent GC and a LECO Pegasus TOF mass spectrometer was used , with separation using a DB-5ht column ( Agilent , USA ) [63] . Auxin transport in inflorescence stems was measured according to the methods of a previously published protocol [73] . 25-mm inflorescence segments were cut from 5-week-old plants , and the segments were submerged inversely into an auxin transport buffer ( 100 nM 3H-IAA , 0 . 05% MES , pH 5 . 5–5 . 7 ) in a 0 . 5-ml micro-centrifuge tube . Control experiments were performed by submerging the base of inflorescence stems to measure acropetal IAA movement . After 12 h , 5-mm stem segments were dissected from the non-submerged ends and used to quantify the radiolabeled auxin using a scintillation counter . ChIP assays were performed as described previously [74] . Briefly , 2 g of p35S::IDD16-GFP or p35S::IDD14-GFP transgenic plants grown on 1/2 MS plates for 16 days was harvested , and then submerged in 1% formaldehyde to crosslink the DNA with DNA-binding proteins . The chromatin pellets were extracted and sheared by sonication . 5 µl anti-GFP antibodies ( Abcam , UK ) were used to immunoprecipitate the DNA-IDD16 or DNA-IDD14 complexes . DNA was released with proteinase K and then purified . The enrichment of DNA fragments was determined by quantitative PCR with the primers listed in Table S2 . Plants were grown in soil until the primary inflorescence stems bolted to a height of 4–9 cm , and then gravistimulated by rotating them upside down [40] . 1-cm-long inflorescence stems below the apex were fixed , embedded in paraffin , and sectioned to a 10-µm thickness . A periodic acid-Schiff kit ( Sigma-Aldrich , USA ) was used for amyloplast staining , according to the manufacturer's instructions . | Auxin is a key plant hormone and the spatial accumulation of auxin is essential for lateral organ morphogenesis and gravitropic responses in higher plants . However , the various mechanisms through which spatial auxin accumulation is regulated remain to be fully elucidated . Here , we identify a gain-of-function mutant of Arabidopsis IDD14 that exhibits small and transversally down-curled leaves . Further characterization of both gain- and loss-of-function mutants in IDD14 and its close homologs , IDD15 and IDD16 , reveals that these three IDD transcription factors function redundantly and cooperatively in the regulation of multiple aspects of lateral organ morphogenesis and gravitropic responses . We further demonstrate that these IDD transcription factors influence the spatial accumulation of auxin by directly targeting auxin biosynthetic and transport genes to activate their expression . These findings identify a subfamily of IDD transcription factors that coordinates spatial auxin gradients and thus directs lateral organ morphogenesis and gravitropic responses in plants . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [] | 2013 | The Arabidopsis IDD14, IDD15, and IDD16 Cooperatively Regulate Lateral Organ Morphogenesis and Gravitropism by Promoting Auxin Biosynthesis and Transport |
Trachoma , caused by Chlamydia trachomatis ( Ct ) , is the leading infectious cause of blindness . Sequence-based analysis of the multiple strains typically present in endemic communities may be informative for epidemiology , transmission , response to treatment , and understanding the host response . Conjunctival and nasal samples from a Gambian community were evaluated before and 2 months after mass azithromycin treatment . Samples were tested for Ct by Amplicor , with infection load determined by quantitative PCR ( qPCR ) . ompA sequences were determined and their diversity analysed using frequency-based tests of neutrality . Ninety-five of 1 , 319 ( 7 . 2% ) individuals from 14 villages were infected with Ct at baseline . Two genovars ( A and B ) and 10 distinct ompA genotypes were detected . Two genovar A variants ( A1 and A2 ) accounted for most infections . There was an excess of rare ompA mutations , not sustained in the population . Post-treatment , 76 ( 5 . 7% ) individuals had Ct infection with only three ompA genotypes present . In 12 of 14 villages , infection had cleared , while in two it increased , probably due to mass migration . Infection qPCR loads associated with infection were significantly greater for A1 than for A2 . Seven individuals had concurrent ocular and nasal infection , with divergent genotypes in five . The number of strains was substantially reduced after mass treatment . One common strain was associated with higher infection loads . Discordant genotypes in concurrent infection may indicate distinct infections at ocular and nasal sites . Population genetic analysis suggests the fleeting appearance of rare multiple ompA variants represents purifying selection rather than escape variants from immune pressure . Genotyping systems accessing extra-ompA variation may be more informative .
Trachoma is the leading infectious cause of blindness worldwide [1] . Repeated infection by Chlamydia trachomatis provokes chronic follicular conjunctivitis ( clinically active trachoma ) , which leads to conjunctival scarring , entropion , trichiasis and ultimately blinding corneal opacification . Trachoma is a major public health problem affecting some of the world's poorest regions . Current estimates indicate 84 million have active trachoma , with 7 . 6 million visually impaired from trachomatous corneal opacification [2] . The World Health Organization is leading a global effort to control blinding trachoma through the implementation of the SAFE Strategy: Surgery for trichiasis , Antibiotics to reduce the burden of chlamydial infection , and face washing and environmental improvements to limit transmission [3] . Endemic trachoma is caused by 4 of the 19 recognised serovars of C . trachomatis: A , B , Ba and C . Serovars are distinguished from each other on the basis of surface variations in the Major Outer Membrane Protein ( MOMP ) . As the main antigenic target for strain specific humoral immunity to C . trachomatis , MOMP has been considered a vaccine candidate [4] . MOMP is encoded by the ompA gene , which contains four variable segments ( VS ) interspersed between five conserved segments ( CS ) . Comparative genome sequence analysis has indicated considerable variation in ompA , possibly driven by host immune pressure , and the study of ompA variants may therefore be informative in disease settings [5] , [6] Originally serovars were distinguished according to their recognition by panels of patient sera , however the ompA sequence motifs for each serovar have now been well characterised . Organisms assigned to a serovar group on the basis of their ompA sequence are referred to here as genovars . OmpA genotyping has been used previously to investigate C . trachomatis infections in trachoma endemic populations [7]–[14] , usually with the goal of better understanding C . trachomatis transmission . However the analysis of ompA sequence variation is also relevant to the utility of MOMP as a target for chlamydial vaccine development . In genital infections caused by C . trachomatis D-K genovars , evidence that genovar and strain variants associate with clinically important differences in the biology of infection is marginal [15] , and has not been described in human ocular infection . Here we analyse ompA genotypic diversity before and two months after mass antibiotic treatment of trachoma in Gambian villages [16] , [17] .
The Gambian Government/Medical Research Council Joint Ethics Committee ( SCC 856 ) and the London School of Hygiene and Tropical Medicine Ethics Committee approved the study . All subjects , or their guardians , gave written informed consent , or witnessed consent by thumbprint where appropriate . This study was conducted in 14 trachoma endemic Gambian villages , located within a defined geographical area [16] , [17] , [18] . The villages were surveyed and a population census was conducted . Individuals normally resident in the study area for at least 6 months of the year were enrolled . At baseline the entire available population was examined for signs of trachoma and classified using the WHO Trachoma Grading System [19] . A swab sample was collected from the upper tarsal conjunctiva of each subject for DNA isolation and kept cool until frozen at −20°C later the same day . Swabs of fresh nasal discharge were collected . Following baseline clinical assessment , all participants were offered antibiotic treatment . Adults and children over 6 months old were given a single oral dose of azithromycin ( 20mg/kg up to a maximum of 1g ) . Infants under 6 months were given tetracycline eye ointment ( twice daily , 6 weeks ) . All villages were examined and treated within a 9 day period [17] . Two months after baseline assessment and antibiotic treatment , participants were re-examined , and conjunctival and nasal discharge samples again collected . Between these two time points , the census was updated weekly , together with records of destination and duration of travel and of the presence of any external visitors . DNA was extracted from the swabs and tested using the Amplicor CT/NG kit ( Roche ) [16] . Amplicor extracts from specimens with detectable C . trachomatis DNA were further purified and concentrated using the QIAamp DNA Mini Kit ( Qiagen ) [16] . Infection load was estimated by quantitative real-time PCR for the chlamydial ompA gene using a previously described method [20] . Sequencing of ompA used primers spanning VS1-4 and sequences were comfirmed by a second sequencing pass . A 1076bp fragment was amplified using primers 87: 5′ - TGA ACC AAG CCT TAT GAT CGA CGG - 3′ and 1163: 5′ - CGG AAT TGT GCA TTT ACG TGA G - 3′ . If no amplified product was visible on an agarose gel , nested PCR was performed , with primers 87 ( above ) and 1059: 5′ - GCA AGA TTT TCT AGA TTT CAT C - 3′ used to amplify a 972bp target sequence . PCR products were purified using the QIAquick PCR purification kit ( Qiagen ) and sequenced using BigDye Terminator Cycle Sequencing Ready Reaction kit V3 . 1 ( Applied Biosystems ) with outer primers 97: 5′ - CTT ATG ATC GAC GGA ATT TTC TAT GGG - 3′ and 1047: 5′ - GAT TTT CAT GAT TTC ATC TTG TTC AAC TG - 3′ . Sequencing with inner primers 608: 5′ - CTC TCT GGG AAT GTG GGT GT - 3′ and 627: 5′ - ACA CCC ACA TTC CCA GAG AG - 3′ was performed to close sequencing gaps . Sequences were edited and aligned using DNA*DNASTAR 5 . 07 ( DNASTAR ) , with HAR 13 ( NC_007429 ) as genovar A reference and M33636 for genovar B . Here , a genotype denotes an ompA sequence variant differing from the ompA reference sequence or from another variant by one or more single nucleotide substitutions , and is identified using the letter of its genovar and an arbitrary number . Data were analysed in Stata 9 . 0 , with differences in loads per genotype examined using a two tailed t-test on logtransformed loads . Sequence alignments were imported into DNAsp4 . 00 and Tajima's D value calculated [21] , [22] . P-values for each D test were calculated using 10 , 000 coalescent simulations without the presence of recombination to calculate the proportion of D values generated which were greater than the observed D value . D* and F* indices were calculated as further tests of the neutrality of mutations [23] .
1319 ( 83% ) of 1595 people enumerated at baseline were examined , sampled and treated . At two-months 1344 ( 85% ) were examined and sampled . The overall prevalence of active trachoma in children <10 years was 16% before and 12% two months after treatment , with marked variations in prevalence between villages [16] . The prevalence of C . trachomatis infection was 7 . 2% ( 95/1319 ) before treatment and 5 . 7% ( 76/1344 ) two months after treatment . Of individuals infected at baseline , 30% were still infected two months after treatment and of those infections detected at two months 36/66 ( 59% ) occurred in subjects uninfected at baseline ( Table 1 ) . Most infections ( 74/76; 97% ) detected two months after treatment were in two villages . Almost all residents of these two villages travelled en masse to a religious festival one month after the treatment . This travelling event was very strongly associated with infection at two months [16] . In contrast , in the other 12 study villages all cases of C . trachomatis infection found at baseline had resolved by two months and there were only 2 new cases of infection in previously uninfected individuals . 77/95 ( 81% ) baseline and 64/76 ( 84% ) two-month ocular C . trachomatis samples yielded sequence data . On both occasions sequence data were obtained from all 5 Amplicor-positive nasal specimens . 73 ( 95% ) of the baseline ocular sequences were genovar A and 4 ( 5% ) were genovar B . Overall , ten separate genotypes were identified; 8 genovar A and 2 genovar B . Sequence variation compared to reference strains is shown in Table 2 . For most genotypes single nucleotide polymorphisms ( SNPs ) resulted in amino acid changes in the variable sequence domains of MOMP . Within genovar A baseline sequences , there were eight polymorphic sites , of which five contained singletons ( SNPs found only in a single isolate ) . Tajima's D value for baseline genovar A sequences was −1 . 06 , revealing trend towards an excess of rare mutations , ( p = 0 . 16 ) . This was supported by significantly negative D* and F* indices , indicating an excess of singleton mutations amongst genovar A sequences ( −2 . 59; P = 0 . 02 and − 2 . 45; P = 0 . 02 respectively ) . Only four genovar B sequences were found , therefore frequency based analyses could not be performed . However , addition of these four sequences to the genovar A sequences for calculation of an overall Tajima's D value revealed a significant excess of rare mutations within the baseline dataset as a whole ( D = −1 . 76 ; p = 0 . 018 ) . Genotype frequencies are presented in Table 3 . The dominant strain , A2 , accounted for 74% of baseline ocular isolates . All other strains , except A1 , were detected in only a few individuals . The 14 villages contained 79 family compounds ( fenced areas inhabited usually by the members of one extended family ) . 16 ( 20% ) contained subjects infected at baseline . Seven compounds contained multiple strains; three of which had 3 strains and one 5 different strains . Obvious environmental risk factors which might explain this concentration of diversity were not identified: however the latter compound had an unusually high proportion of its children attending the local primary school ( 7/25; 28% ) compared to ( 30/773: 4% ) in the study area generally . At two months post-treatment only three strains A1 , A2 and A5 were found . The A2 proportion increased to 90% . Rare strains had mostly disappeared . In 23 individuals ocular samples yielded sequence data at both time points . 18 ( 78% ) of these had the same strain at both timepoints: 3 A1 , 14 A2 and 1 A5 ( the only example of A5 at either timepoint ) . 5 ( 22% ) showed a change in genotype: from A1 to A2 in three cases , from A3/A4 to A2 in one case each . 34/35 ( 97% ) newly infected individuals at two-months had the A2 genotype . Infection load data from this population has been previously described [16] , [17] . Geometric mean infection loads for strains A1 and A2 were compared by unpaired , two-sided t-tests on logarithmically transformed data . Chlamydial load was significantly higher in A1 infections before mass treatment: geometric mean for A1 5809 copies ( 95% CI 374–90189 ) ( n = 6 ) and for A2 92 copies ( 95% CI 59–144 ) ( n = 14 ) ( p<0 . 0001 ) . Similarly , after mass treatment geometric mean for A1 was 343 copies ( 95% CI 42–277663 ) ( n = 3 ) compared to 115 copies ( 95% CI 66–202 ) ( n = 19 ) ( p = 0 . 0021 ) . At both baseline and two-months , subjects infected with A1 were more likely than those infected with A2 to have clinically active disease: baseline: 7/10 vs 6/57 ( RR = 6 . 65 , χ 2 = 15 . 63 , p<0 . 0001 ) ; two-months 3/5 vs 7/58 ( RR = 4 . 97 p = 0 . 025 2-tailed Fisher's Exact Test ) . We have previously found that infected individuals with clinical signs of trachoma have higher chlamydial loads than those without signs [16] , [17] . These analyses are not adjusted for potential clustering by village: however A1 only occurred in one village ( village 3 ) . C trachomatis was detected in nasal samples from 5/58 subjects at baseline , and from 5/54 at two months . In seven subjects ompA sequence was determined in both ocular and nasal samples at the same time point: 5/7 ( 71% ) had different genotypes at the two sites: A1 ( ocular ) /A2 ( nasal ) in three cases , with A1 ( ocular ) /A3 ( nasal ) and A2 ( ocular ) /A7 ( nasal ) in one each . Differing genotypes were found in all four individuals in whom baseline ocular and two-month nasal ompA sequence were both determined , and in the two individuals in whom baseline nasal and two-month ocular ompA sequences were both determined .
In this study , 972 bp sequences comprising almost the entire C . trachomatis ompA gene were determined in samples from infected individuals in a trachoma endemic area . Previous trachoma studies have sequenced primarily VS regions: variation in the interspersing ‘conserved’ segments is recognised but not usually examined at the pathogen population level . All variants were confirmed with double pass sequencing methods: dubious calls on the chromatogram were all clarified by resequencing . We discuss the utility of ompA genotyping for determining the existence and nature of selection pressure on the locus , for examining whether variants affect the features of infection or disease , and for distinguishing causes of reemergent infection after treatment . Ten C . trachomatis genotypes were identified at baseline . Excepting B2 , these differed from strains previously sequenced from The Gambia and elsewhere [7]–[11] , [13] . Before treatment most ( 87% ) infections were one of two strains ( A1 and A2 ) . Six of the minority genovar A strains had SNPs resulting in amino acid changes within variable segment domains . A similar pattern of a few dominant strains with several other strains present at low frequency has been described previously [7] , [10] , [14] . The variety of strains in this limited geographical area might suggest that new strains are regularly introduced through mixing with other populations or alternatively that the emergence of new variants is promoted by pressure from the human immune response . To test this frequency based analyses of polymorphism were carried out . Population genetic analysis of baseline genovar A ompA sequences showed negative Tajima's D , Fu and Li's D* and F* statistics , suggesting that in this environment novel genovar A ompA mutations are being eliminated from the population . Despite this , the location of some of the polymorphic sites is intriguing . In genotype A5 the neutralizing antibody epitope which defines serovar A ( 70DVAGLEK76 ) is significantly altered ( 70DEAGLQK76 ) : previously we noted significant alteration in close proximity to this epitope ( 69 ( S→R ) DVAGLEK76 ) in strains which subsequently failed to establish themselves in the community [10] . One would expect that novel mutations which allow immune evasion offer the pathogen a selective advantage ( at least while these strains remain uncommon ) , and ought to spread through the pathogen population until they reach intermediary frequencies . The excess of rare mutations observed at baseline therefore does not support the hypothesis that ompA polymorphisms are maintained within this population by immune selection pressure . Instead it implicates either ongoing negative selection ( where most mutations are deleterious and removed from the population by purifying selection ) or a recent selective sweep ( whereby a single haplotype has reached fixation within the population , driving out diversity at the locus ) . Few studies have applied population genetic methods to analyse selection of C . trachomatis genes , but they have similarly generated little evidence that ompA is under immune selection pressure: both cross sectional studies of genovar A ompA sequences from Tanzania and sequence analysis of genital Ct genovars have found similar evidence of purifying selection in ompA 11 , 24 These data and the existence of individuals within trachoma endemic communities who are often or repetitively infected with the same ompA genovar lead us to question whether the ompA locus is a target of selective pressure in trachoma populations , and consequently whether targeting MOMP will lead to an effective vaccine . Strain-specific differences affecting infection or disease manifestations are described in genital chlamydial infection , but not previously in trachoma . On both occasions strain A1 was associated with clinical signs of active trachoma and with higher mean infection loads to a greater extent than A2 , but it was less common in the community and so not necessarily a more successful pathogen . The sampling method used here has been shown elsewhere to give adequate yields of host RNA [18] , but the infection loads were not standardised , for example against host DNA yield in the sample . In natural infections the number of cells sampled , the proportion of host cells which are infected and the state of the chlamydial developmental cycle within them will all affect the measured load , and the best way to standardise the measurements is not clear . A1 and A2 might amplify differently by PCR , although there was no support for this suggestion in the amplification of standards , and no variation affecting primer binding sites . Differences in sampling , in PCR amplification or in the infection/disease course within the sampled individuals might explain this observation , or alternatively it could result directly or indirectly from variation in ompA . Three differences exist in the ompA sequence of A1 and A2 , of which two cause non-synonymous amino acid substitutions . These might alter the conformation of MOMP or have direct effects on ‘fitness’ , transmission or the host response . The G→A mutation at position 304 introduces a cleavage motif for cathepsin-L , which generates of peptide fragments for antigen presentation [25] , [26] . Whether peptide fragments of A1 and A2 MOMP are therefore presented differently during the generation of adaptive cellular immunity is unknown . Alternatively , strain differences might be unrelated to ompA itself but reflect linkage between ompA genotype and polymorphism ( s ) elsewhere on the chlamydial chromosome leading to differences in fitness or metabolic advantage . Trachoma strains may differ in their laboratory properties , and a recent study found differences in in vitro growth rate , interferon-γ sensitivity and virulence in non-human primates [27] , attributable to variation affecting 22 open reading frames ( ORFs ) in addition to ompA . Both clinical differences between strains , and the purifying selection at the ompA locus could result from variation or selection pressure at linked chlamydial ORFs . Following mass antibiotic treatment there was a modest reduction in the prevalence of infection [17] . Only 3 of the original 10 genotypes were still present . Most ( 90 . 5% ) of these infections were with A2 , and almost all in two villages ( 1 and 3 in Table 4 ) , in which the prevalence of infection actually increased [17] , with strains A1 and A2 continuing to dominate . New infections , 97% with strain A2 , were strongly associated with travel to a festival in Senegal , at which over a million people from the region congregated in basic conditions , where the opportunity to acquire ocular C . trachomatis infection was probably considerable . These data suggest that a remarkable re-infecting exposure to strain A2 occurred in the treated subjects during this event . The persistence of the common A1 or A2 strains in 17 individuals in these villages could be due to treatment failure or to reinfection facilitated by the same unusually effective environment for C . trachomatis transmission . Genotyping provides some evidence that antibiotic treatment was not 100% effective , as strain A5 was found twice , but in the same individual both before and after treatment , strongly suggesting primary treatment failure . Nevertheless antibiotic treatment cleared all baseline infections in the other 12 villages [17] . The surprising demonstration of discordant genotypes in concurrent ocular and nasal samples may imply that these two mucosal surfaces function as distinct sites of infection , despite direct communication via the nasolacrimal duct . This could result from differences in the time course of infection or in the route of inoculation . Autoreinfection of the conjunctiva from extraocular sites such as the nasal mucosa has been suggested , however , a study from Tanzania did not support this hypothesis [28] . Here , the limited nasal genotyping data does not provide support significant transmission between eye and nose . Our study illustrates the use and limitations of ompA sequence data in the molecular epidemiology of C . trachomatis infection . The pattern of ompA sequence diversity remains intriguing and inconsistent with immune selection pressure . Typing systems including other polymorphic loci may lead to better elucidation of key events in ocular C . trachomatis infection . An ongoing extended longitudinal study of C . trachomatis genotypes might better define the population dynamics , and determine implications for the long-term success of mass treatment [14] . | Trachoma is an important cause of blindness resulting from transmission of the bacterium Chlamydia trachomatis . One way to understand better how this infection is transmitted and how the human immune system controls it is to study the strains of bacteria associated with infection . Comparing strains before and after treatment might help us learn if someone has a new infection or the same one as before . Identifying differences between disease-causing strains should help us understand how infection leads to disease and how the human host defences work . We chose to study variation in the chlamydial gene ompA because it determines the protein MOMP , one of the leading candidates for inclusion in a vaccine to prevent trachoma . If immunity to MOMP is important in natural trachoma infections , we would expect to find evidence of this in the way the strains varied . We did not find this , but instead found that two common strains seemed to cause different types of disease . Although their MOMPs were very slightly different , this did not really explain the differences . We conclude that methods of typing strains going beyond the ompA gene will be needed to help us understand the interaction between Chlamydia and its human host . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [
"genetics",
"and",
"genomics/microbial",
"evolution",
"and",
"genomics",
"ophthalmology/eye",
"infections",
"public",
"health",
"and",
"epidemiology/infectious",
"diseases",
"infectious",
"diseases/bacterial",
"infections",
"genetics",
"and",
"genomics/population",
"genetics"
] | 2008 | Chlamydia trachomatis ompA Variants in Trachoma: What Do They Tell Us? |
Understanding of gene regulatory networks requires discovery of expression modules within gene co-expression networks and identification of promoter motifs and corresponding transcription factors that regulate their expression . A commonly used method for this purpose is a top-down approach based on clustering the network into a range of densely connected segments , treating these segments as expression modules , and extracting promoter motifs from these modules . Here , we describe a novel bottom-up approach to identify gene expression modules driven by known cis-regulatory motifs in the gene promoters . For a specific motif , genes in the co-expression network are ranked according to their probability of belonging to an expression module regulated by that motif . The ranking is conducted via motif enrichment or motif position bias analysis . Our results indicate that motif position bias analysis is an effective tool for genome-wide motif analysis . Sub-networks containing the top ranked genes are extracted and analyzed for inherent gene expression modules . This approach identified novel expression modules for the G-box , W-box , site II , and MYB motifs from an Arabidopsis thaliana gene co-expression network based on the graphical Gaussian model . The novel expression modules include those involved in house-keeping functions , primary and secondary metabolism , and abiotic and biotic stress responses . In addition to confirmation of previously described modules , we identified modules that include new signaling pathways . To associate transcription factors that regulate genes in these co-expression modules , we developed a novel reporter system . Using this approach , we evaluated MYB transcription factor-promoter interactions within MYB motif modules .
The advancement in technologies in recent years has resulted in many large data sets cataloging the biological systems at various levels . Biological networks inferred from these data have become an important tool to describe and analyze biological signaling systems [1]–[3] . Depending on the sources of the data , different biological networks include information on protein-protein and protein-DNA interactions , or network structures for gene co-expression , metabolism , phosphorylation , and yet other structured sets that integrate diverse data sources . Identifying novel signaling or gene expression modules from these networks has become a major goal of systems biology . Plant biological networks are mainly gene co-expression networks based on large-scale transcriptome data . Relatively few studies on protein-protein interaction [1] , [4] , [5] , protein-DNA interaction [6] , [7] or phosphorylation [8] have been reported . The gene co-expression networks consist of nodes representing genes and edges representing connections between nodes . An edge between two genes indicates that they have similar expression patterns under various biological conditions . The pair-wise gene expression similarities are mostly measured using the Pearson correlation coefficient [9]–[12] . In addition , association measurements have also been derived using Mutual Rank [13] , the Spearman correlation coefficient [14] , and the partial correlation coefficient [15]–[17] methods . Plant functional networks integrating multiple data types , including co-expression , have also been reported [18]–[21] . Once generated , these co-expression networks are used to identify expression modules to extract biological meaning . An expression module includes a subset of genes from within the network that are highly interconnected with each other but show only limited connection to genes outside the subset . Expression modules usually represent groups of co-expressed genes with condition-specific similar or same expression patterns , suggesting that they likely belong to gene expression units regulated by the same transcription factor ( s ) ( TF ) . Various network clustering methods have been used to identify such modules from plant gene co-expression networks . These include Markov chain clustering ( MCL ) [9] , [10] , [22] , [23] , IPCA [12] , NeMo algorithm [24] , and HQcut [25] . In these methods the clustering algorithms while searching for modules only consider the topology and connectivity of the networks but fail to take into account the properties of the nodes or the genes such as promoter sequences . Motifs in the promoters are only searched after the modules are extracted . This represents a top-down strategy . Here , we describe a bottom-up approach to identify expression modules from a previously published Arabidopsis thaliana gene co-expression network based on the graphical Gaussian model [15] , [26] . Our major interest is to understand how known promoter motifs are distributed across the gene network and to identify gene expression modules that these motifs might regulate . For any given motif , every gene in the network was first analyzed to calculate its probability of belonging to an expression module regulated by that motif . Then , all the top ranked genes were used to extract a sub-network from the original gene co-expression network . From this sub-network , the modular structures will self-manifest , thus enabling discovery of novel signaling pathways . We used this approach to successfully identify novel expression modules for four well studied motifs - G-box , MYB , W-box , and site II element . We validated our predicted promoter-motif interactions using a novel in vivo reporter assay system . The bioinformatics program described here can be used to extract expression modules for any motif of interest .
Gene co-expression networks describe the pattern of co-expression between genes . The connected gene pairs within such networks share similar expression patterns . A subset of genes within such a network might be combined by the presence of a specific motif in their promoters . Depending on their expression patterns , some of these same-motif-containing genes cluster together and form expression modules , while others disperse across the network ( Figure 1A ) . The genes in the former category cluster together at a frequency higher than random distribution . It is assumed that these clustered same-motif-containing genes belong to expression module ( s ) that will be regulated by the corresponding motif in a condition-specific manner . It is also important to note that the promoter motifs tend to show position bias in their distribution relative to the transcription start site ( TSS ) . Consider two groups of genes containing the same motif in their promoters with similar frequency . We can distinguish them by one where a motif is distributed evenly along the promoters and the other where the motif is skewed towards being present closer to the TSS ( Figure 1B ) . The probability for the latter group of genes to be regulated by that motif is higher than the former group . Thus , by studying how a specific motif distributes across the network , it is possible to identify the expression modules it regulates . The key is to distinguish the same-motif-containing genes belonging to expression modules with motif enrichment/motif position bias from those that do not belong . For this , we employed two independent methods . One is based on the hypergeometric distribution to assess motif enrichment and the other is based on the uniform distribution to measure motif position bias towards TSS . Specifically for each motif , a pValue of motif enrichment and a z-score for motif position bias were calculated for every gene within the network in the following manner . For any gene , the gene and its immediately connected neighbor genes within the network are considered as a group . The frequency of the motif present within the promoters of this group of genes is compared to those of the whole genome and a pValue based on the hypergeometric distribution is calculated . The locations of the motif within these promoters are also used to compute a z-score as an indicator of whether the motif has a position-bias distribution towards TSS as described before [27] ( see Material and Methods ) . A large z-score indicates the motif has a biased distribution towards TSS , while a motif with even distribution along the promoters will result in a z-score close to zero . Genes are then ranked according to their pValues . The smaller the pValue the higher the chance that the gene belongs to an expression module regulated by the motif that is under consideration . All genes with pValues smaller than a selected cut-off are used as seeds to generate a sub-network from the original co-expression network . The sub-network is inspected for the existence of densely connected modules that provide information about the propensity of the motif to drive the modular expression of its targets . As an independent method , genes are also ranked according to their z-score . The genes with z-scores larger than a selected cut-off are extracted and used to generate sub-networks . The sub-networks are then inspected for module structures . For our analysis , we used an Arabidopsis gene co-expression network that had been established based on the graphical Gaussian model ( GGM ) [15] , [26] . With a partial correlation co-efficient cut off at 0 . 05 [26] , it contains 16 , 459 genes ( nodes ) and 120 , 276 co-expressed gene pairs ( edges ) ( Table S1 ) . Here , we focused our analysis on the 10 , 385 nuclear-encoded genes connected to 5 or more co-expressed genes , i . e . nodes with > = 5 edges . The bZIP transcription factor family includes 75 members in Arabidopsis that regulate diverse signaling processes in plants [28] . bZIP TFs predominantly bind to the G-box ( CACGTG ) motif in promoters . We analyzed how the G-box motif is distributed across the gene co-expression network . Out of the 10 , 385 genes analyzed , 497 exhibited a pValue for the G-box lower than 0 . 001 ( Table S2 ) , while only 5 genes on average were recovered in permutation experiments with randomized promoter sequences . The estimated false discovery rate ( FDR ) is 1% . A sub-network for these 497 genes is extracted from the original gene co-expression network ( Figure 2 ) . Out of the 497 genes in the sub-network only 291 harbor the G-box motif . The remaining 206 genes are represented in the sub-network because their neighbors possess the G-box motif . Within the G-box sub-network , several densely connected sub-groups of genes or expression modules were identified . Functions of genes in the sub-network are illustrated by their enriched GO term ( Figure 2 , Table 1 ) . Our analyses identified 10 gene modules that are regulated by various developmental or environmental cues such as abiotic and biotic stress , pathogen elicitors , hormones , and different light regimes ( Figure 2 , Figure S1 , Table 1 , and Table S3 ) . Module V , VI , and VIII included genes that are known to be regulated by bZIPs in ABA response pathways [29] , [30] , embryogenesis [31]–[33] , and the ER stress response [34]–[36] . Interestingly , Module V includes genes that are induced by the bacterial pathogen Pseudomonas syringae pv tomato ( Pst ) DC3000 but repressed by the DC3000 hrcc− strain that lacks the type III secretion system used to deliver effector proteins into plant cells . This indicates that the Pst DC3000 pathogen appears to deliver effectors that stimulate ABA signaling pathways through bZIP transcription factors as reported before [37] . In contrast , Module VI includes ER stress genes that are also induced by various pathogens and elicitor treatments . Several genes in Module X are previously categorized as common stress responsive genes [38] but TFs that regulate these genes via the G-box motif have yet to be identified . Thus , Module X identified here is a novel module requiring further studies . Interestingly , some bHLH transcription factor family members also bind to the G-box motif [39] . PIF3 and PIF4 bHLH transcription factors bind to G-box containing photosynthesis genes and the circadian rhythm genes LYH and CCA1 , indicating Module I and II's regulation by bHLH factors [40] , [41] . Module I genes were induced by long exposure to light and Module II genes were induced by short exposure to light . Another bHLH protein AtMYC2 also binds to the G-box element and regulates genes in the jasmonate signaling pathway [42]–[44] which were enriched in module VII . The genes in module IX were enriched for functions in glucosinolate biosynthesis including SUR1 that might be negatively regulated by AtMYC2 [43] , [45] . As an independent measure , we carried out module discovery for the G-box motif via motif position bias analysis . 519 out of 10 , 385 genes analyzed show a z-score for G-box larger than or equal to 3 . A sub-network for these genes was extracted ( Figure 3 ) . On an average , only 1 . 3 genes were identified with a z-score> = 3 in permutation experiments with an FDR of 0 . 3% . Interestingly , this method recovered 9 out of the 10 modules that were also identified via the pValue method ( Figure 2 ) . Modules derived by either method shared a large number of genes , demonstrating the reliability of the analysis . Four additional modules emerged , among them two potentially novel modules regulated by the G-box motif: Module XII is enriched for heat shock proteins and Module XIV contains genes specifically expressed in roots . The majority of the modules identified in our analyses for the G-box motif are consistent with previous studies that focused on individual pathways . In addition , we discovered three novel modules . Importantly , many genes were identified here as part of the known modules for the first time ( Table S3 ) . In addition , our analysis successfully places these genes in a signaling framework that will facilitate further studies on biological functions . Another notable observation is detection of an overlap of the modules for ABA signaling and jasmonate signaling ( Figure 2 ) , suggesting that the regulatory circuits to which these genes respond might be under the control of these two hormones . The interaction and binding of bZIP or bHLH transcription factors with the G-box motif in the promoters of these genes might lead to competition . In fact , antagonistic interaction between the two hormones has been reported before [46] , [47] . In Arabidopsis , the MYB transcription factor family includes >190 members that regulate diverse functions [48] , [49] . We analyzed distribution of two MYB binding motifs , CCwACC and ACCwACC ( with “w” standing for “A” or “C” ) [50] , [51] , across the co-expression network . In the network , 243 genes show pValues for CCwACC or ACCwACC lower than 0 . 01 ( Table S2 ) . A sub-network for these genes is shown in Figure 4 . An inspection of the sub-network revealed 10 expression modules ( Figure 4 , Table 2 ) . A number of these modules are known to function in biosynthesis of various secondary metabolites such as flavonoid ( module III ) , glucosinolate ( II ) , indole derivative ( I ) , anthocyanin ( IV ) , and phenylpropanoids ( VIII ) . Their expression pattern ( Figure S2 ) clearly highlights the activation of diverse metabolic modules in Arabidopsis to cope with distinct environmental stresses . For example , Module V genes were highly induced in response to pathogen elicitors and the bacterial pathogen Pst DC3000 , possibly representing their function in the basal innate immune response . Genes in this module are implicated in different steps of lignin biosynthesis pathway . Module I genes were up-regulated by broader stimuli including methyl jasmonate , the oomycete pathogen Phytophthora , and the fungal pathogen Botrytis . In contrast , the glucosinolate genes in Module II were universally repressed by pathogens . Module VII appears to operate in nitrogen metabolism based on the presence of UPM1 and AT3G58610 genes in this module . Together , the functions collected in these modules are consistent with previous reports about MYB-mediated regulation of diverse metabolic pathways [48] , [52]–[60] . Three of the modules ( VI , IX , & X ) in the sub-network are involved in tissue development . Module IX contains genes specifically expressed in roots and seeds ( Figures S3 ) , indicating a novel module that might control root and seed development . We also noted that module II of the MYB sub-network shared genes with module IX of the G-box sub-network . For example , SUR1 and CYP83A1 genes ( Figure 2 and Figure 4 ) . Gene DFR in module III also appears in module IV of the G-box sub-network . These results indicate some of the genes in these modules are regulated by both the G-box motif and the MYB motif . The position bias analyses of the MYB motif identified 348 genes in the co-expression network with z-scores for the CCwACC or ACCwACC larger or equal to 2 . 2 . For genes with z-score between 2 . 2 and 3 , it is required that there are at least 5 instances of the motifs within the promoters of that gene and its neighbor genes . A sub-network for these genes is shown in Figure 5 . This sub-network revealed 15 modules . Seven of these modules were also identified via the pValue method ( Figure 4 ) . In the remaining 8 modules 2 function in known MYB-regulated pathways: nitrate transport ( XIV ) and wax biosynthesis ( XVII ) . Three of the modules are novel and include genes responding to ABA ( XI ) , auxin ( XV ) , and hypoxia ( XII ) . To assess the FDR in the MYB motif analysis , permutation experiments were conducted with randomized promoter sequences . The permutation was performed 15 times . In each permutation , motif enrichment analysis was conducted for the MYB motif , and the genes with pValue< = 0 . 01 were used to extract a sub-network from the entire gene co-expression network . A typical sub-network is shown in Figure S4 . On average 2 . 7 gene modules were recovered that each contained at least 6 genes from each permutation . Therefore , the FDR for MYB motif module identification is 2 . 7 out of 10 or 27% in the motif enrichment analysis . Similarly , in the motif position bias analysis , on average 3 . 1 gene modules with > = 5 gene numbers were identified among genes with z-score> = 2 . 2 from each permutation ( Figure S5 ) . Thus , there might be up to 3 . 1 false discovered modules or a FDR of 21% ( 3 . 1/15 ) in the analysis based on position bias . Additionally , in each permutation , only 1 gene on average was recovered with both pValue< = 0 . 01 and z-score> = 2 . 2 , and no gene modules was identified that fulfills both requirement . This indicates no falsely discovered modules exist among the 7 MYB-related modules recovered by both methods . The WRKY transcription factors play important roles in plant defense . They bind to the W-box motif [61] . The core sequence of the W-box motif is TTGACy ( with “y” standing for “A” or “G” ) , but various variant forms of the sequence also show binding affinity to WRKY proteins [62] . Here , we analyzed the W-box motif variant kTTGACy ( with “k” standing for “G” or “T” ) identified in our previous study [27] . There are 388 genes whose pValues for this W-box motif is less than 0 . 001 with a FDR of 1 . 1% . A sub-network for these genes is shown in Figure 6 . From this sub-network , five expression modules can be recognized ( Table 3 ) . The majority of the genes in modules I and II are regulated by pathogen responses . The genes in Module II are primarily induced by Microbe Associated Molecular Patterns ( MAMPs ) or by pathogens , while the genes in module I were also strongly induced by salinity stress in Arabidopsis roots ( Figure S6 ) . Interestingly , genes in Module II were repressed by Pst DC3000 at 6 hour post infection . However , at the same time point , these genes were not repressed by the DC3000 hrcc− mutant ( a mutant that's unable to deliver effectors into plant cells ) [26] . Thus , it appears that the pathogen Pst DC3000 actively delivers effectors into plant cells that interfere with plant signaling pathways and suppress the induction of these genes , presumably for the benefit of the pathogen . The majority of genes in module III can be characterized as common stress responsive genes because they are induced by different types of abiotic or biotic stress [38] . Interestingly , the majority of genes in module IV are specifically expressed in the roots under normal growth condition ( Figure S7 ) but are repressed by salinity or osmotic stress in roots ( Figure S6 ) . In contrast , these genes do not respond to MAMPs or pathogen treatments . These observations raise the possibility that WRKY-mediated signaling might regulate root development . Consistent with these observations , WRKY75 has a function in root hair development [63] . However , any regulatory influence of WRKY75 on genes in module IV has not yet been analyzed . Finally , module V genes are also specifically expressed in roots ( Figure S7 ) and no specific function for WRKY in the regulation these genes are known . Using the motif position bias analysis , 357 genes were identified with a z-score> = 3 and with a FDR of 2 . 4% . The recovered modules included 3 modules ( I , II , V ) identified by the motif enrichment method and 2 additional modules with genes specifically expressed in roots ( VI ) or siliques ( VII ) ( Figure 7 and Figure S7 ) . The Site II element motif TGGGCy , bound by TCP transcription factors is present in the promoters of many cell-cycle genes , ribosomal protein genes , and nuclear-encoded mitochondrial protein genes [64]–[67] . Our motif position bias analysis resulted in 1 , 161 genes with z-scores for TGGGCy larger or equal to 3 with a FDR of 0 . 4% . The sub-network for these genes is shown in Figure 8 . Thirteen modules were identified from the sub-network ( Table 4 ) . Consistent with previous reports [64]–[67] , modules enriched with cell-cycle genes ( V , VI , VII ) , ribosomal proteins genes ( I ) , and mitochondrial proteins genes ( IX ) were identified . Our analysis revealed that some nuclear-encoded chloroplast genes may also be regulated by the site II element motif ( module II , XI , XII ) . Additionally , two novel modules ( III , VIII ) harbor genes functioning in protein folding and one ( IV ) contains genes encoding members of the proteasome complex . Yet another novel module ( X ) includes a number of fatty acid biosynthetic genes . Thus , our analysis indicates that site II element motif might regulate a broader array of biological processes than previously thought . Many of the functions that are highlighted show strong relationships to housekeeping functions of plant cells . Using the motif enrichment analysis method , 161 genes were recovered with pValue< = 0 . 001 for the motif TGGGCy at a FDR of 3 . 6% . Therefore , for the site II element motif , the position bias analysis recovered more genes and performed better than the motif enrichment analysis . The above analysis identified gene expression modules for individual motifs . Here , these modules were incorporated into a single network . Shown in Figure 9A is a sub-network consisting of the top 6 , 000 co-expressed gene pairs from the original GGM network ( the whole GGM network is too big to depict here ) . Among the 3 , 756 genes in this sub-network , 1 , 056 ( 28% ) are regulated by at least one of the four motifs . Gene modules regulated by the W-box motif appear in multiple clusters across the network . The modules regulated by G-box , MYB , or site II elements have similar distribution pattern . A number of modules within the network are regulated by two motifs: MYB & G-box , W-box & G-Box , or G-box & site II elements ( Figure 9B ) . These modules are similar to those identified via single motif analysis . For example , module I from this analysis is regulated by the site II element and shares many genes with the site II element module VI from the single motif analyses ( see Figure 8 ) . Module II is regulated by both the G-box and MYB motifs and shares many genes with the G-box module IX ( see Figure 2 ) and the MYB module II ( see Figure 4 ) from single motif analyses . The structure of the combined motif sub-network is more complex than the one derived from single motif analyses , while the single motif analyses provide the basis to reveal the modular structures within this network . To compare our bottom-up module discovery approach described here with the top-down approach , we used previously published Arabidopsis Gene Co-expression Network ( AGCN ) generated from 1 , 094 Affymetrix ATH1 microarray data sets via the AtGenExpress project [9] . The AGCN network contained 6 , 206 genes and was clustered into 527 modules using the MCL algorithm via a top-down approach [9] . Using the same motif enrichment and motif positions bias analysis employed in our bottom-up approach , we identified AGCN modules that are regulated by the G-box , MYB , WRKY , and the site II element motifs ( Table S4 ) . The results of comparative analyses are shown in Table 5 and Figures S8 , S9 , S10 , S11 , S12 , S13 , S14 , S15 . The two approaches were considered to share a common identified motif-driven module if the respective modules from each approach share common genes between them . 7 out of the 14 modules regulated by G-box motif identified via our bottom-up approach were not recovered by in the AGCN network using the top-down approach ( Figure S8 and S9 ) . These include the modules responding to ABA ( V ) and heat shock ( XII ) , and modules related to flavonoid ( IV ) and glucosinolate ( IX ) metabolism . Similarly , our bottom-up method also identified 10 unique modules for the MYB motif ( Figure S10 & S11 ) . While both methods identified similar number of modules for WRKY motif ( Figure S12 & S13 ) , the top-down method recovered 2 more distinctive modules for the site II elements ( Figure S14 ) . Overall , more unique modules were identified via our bottom-up approach . The MCL clustering program used in the top-down approach on AGCN network generated 3 large clusters ( cluster No . 1 , 2 , and 3 ) with more than 500 genes in each cluster ( Table S4 ) [9] . These three clusters include 2 , 684 genes that represent 43% of all the genes in the AGCN network . These clusters are large and include a mix of real targets of modular regulation with many non-targets . Therefore , prioritizing true target genes from this large cluster size for downstream analyses is not straightforward . For example , the largest cluster ( No . 1 ) of the AGCN network contains 1 , 362 genes . The enrichment of the G-box motif in this cluster suggests that all genes within the cluster are regulated by this motif . In contrast , our bottom-up approach analysis on the GGM network revealed that only 62 genes out of these 1 , 362 genes are regulated by the G-box motifs ( Figure S15 ) . These G-box regulated genes did not spread evenly across the whole sub-network , but occupied certain distinctive sub-domains within it . Thus , our bottom-up approach was able to differentiate the genes potentially regulated by G-box motif from those non-targets , resulting in a more refined and precise gene regulation model than those obtained via the top-down approach . From our analysis it is apparent that a single motif can regulate multiple expression modules . These modules might be regulated by different transcription factors ( TFs ) from the TF family which bind to that motif . For example , the 19 modules identified for the MYB motif ( Figures 4 and 5 ) can be driven by different MYB transcription factors . An important task that remains for our understanding of transcriptional networks will be to distinguish the specificities within a TF family , i . e . which member or members of TF family drive the expression of individual modules . At the same time , the TFs that regulate genes in the module might not be a part of the modules themselves in the co-expression network . This is because TFs themselves may not be regulated at the transcriptional level but may be regulated at the translational or protein turnover levels and thus might have expression patterns different from the genes in the modules . Therefore , analyzing a co-expression network in isolation is not sufficient to identify the TFs responsible for regulating the expression modules . To this end , we developed a rapid screening system to test the transcription factor–promoter interactions . The setup employs the Arabidopsis At4g22920 gene that encodes stay green ( SGR ) protein as a reporter . SGR protein is required for dismantling chlorophyll-protein complexes , leading to chlorophyll degradation [68] , [69] . Transient over-expression of the SGR gene under the control of CaMV 35S promoter induces yellowing of leaves in Nicotiana benthamiana ( Figure 10A ) . In the screening system , the SGR gene was placed behind a promoter of interest and transiently co-expressed with a selected TF in N . benthamiana ( Figure 10B ) ( see Materials and Methods for details ) . If the over-expressed TF can bind to the promoter of interest and drive the expression of SGR gene , the infiltrated N . benthamiana leaves will turn yellow ( Figure 10A ) . In a pilot experiment , the SUR1 gene promoter was linked to the SGR gene and co-expressed with seven different Arabidopsis MYB TFs or an actin gene as a negative control . Only AtMYB28 and AtMYB29 caused leaf yellowing ( Figure 10C; spot #1 and #2 ) . Thus , this straightforward screen established interaction between MYB28 and MYB29 transcription factors and the SUR1 promoter . Next , we were interested in determining which MYB TFs regulate the five expression modules ( Figure 4 ) involved in different secondary metabolic pathways . Using the SGR screening approach , eight promoters from these 5 different expression modules ( Figure 4 ) were selected , and screened against 82 different Arabidopsis MYB TFs . The TSB1 promoter of Module I displayed exceptionally high basal expression levels in the leaves and was excluded from further experiments . The analyses identified 34 interactions between 18 AtMYB TFs and 7 promoters ( Table 6 ) . For each promoter , at least one MYB protein was identified as driving its expression . As a further validation of our SGR reporter assay , a luciferase-based assay was performed to measure the promoter activity [70] , [71] . Four selected promoters were cloned in front of the luciferase gene and co-expressed with different AtMYB TFs in N . benthamiana . Luciferase activities were then measured 48 or 72 hours later as an indication of the promoter activities ( Figure S16 ) . We tested 29 of the 34 interactions identified using the SGR system , and confirmed 23 of them . This demonstrates the usefulness of the rapid SGR-based screening system and its value in the analysis and verification of predictions made by the program . Among the interactions recovered by both reporter systems are the interaction between the SUR1 and APK promoters and several regulators of glucosinolate synthesis , including ATR1 , HIG1 , HAG2 , and PMG2 , which is consistent with previous reports [53] , [54] . The gene CYP98A3 is from module V of the MYB sub-network . Module V is enriched with lignin biosynthesis genes that are induced by pathogen treatment . This is consistent with previous reports that infection by pathogens induced lignification in plants [72]–[74] , although mechanistic details are not known . MYBs are important regulators of lignin biosynthesis [75] but the exact MYB ( s ) that regulate pathogen induced lignification have yet to be identified . Our results showed that several MYBs drive the expression of the CYP98A3 promoter ( Table 6 , Figure S16 ) . Among them , the MYB14 , MYB15 and MYB32 genes themselves were also induced by pathogen treatments ( Figure S17 ) . These MYBs might act as master regulators of the lignification process in the response leading to pathogen resistance .
We describe a bottom-up strategy to identify gene expression modules from gene co-expression networks that are regulated by known promoter motifs . Two independent methods were used to identify genes belonging to modules regulated by specific motifs: based on motif enrichment and motif position bias . For the G-Box , W-Box , and the site II elements , the cut-offs were set at a pValue of 0 . 001 for motif enrichment analysis and a z-score of 3 for position bias analysis . Many known and a number of novel modules were identified with a FDR of ∼1% , indicating very high confidence . To recover additional modules for the MYB motif , the cut-offs were lowered to 0 . 01 for the pValue and 2 . 2 for the z-score . From this , 18 modules were identified with a FDR of 21%–27% representing moderate confidence . However , the overlap of modules between the motif enrichment analysis and the motif position bias analysis for MYBs revealed high confidence . Thus , two different stringency levels may be chosen depending on the nature of the motifs . Even at high stringency levels , our analysis identified more modules than other module analysis based on gene co-expression networks , such as the AGCN network ( Table 5 ) or by Vandepoele et al . [76] . For the G-Box motif , the analysis by Vandepoele et al . [76] recovered modules enriched with GO terms for response to cold , photosynthesis , starch metabolism , and response to ABA . Our analysis identified 14 modules and includes many additional GO-terms ( Table 1 ) . The site II element motif analysis by Vandepoele et al . recovered modules enriched for the GO term ribosome biogenesis and assembly . Our analyses identified 13 modules consisting of 6 known and 7 novel modules . Promoter motifs have long been shown to have position bias towards TSS [27] , [77]–[79] . This feature has been widely used as supporting evidence for the validity of a bona-fide motif in motif discovery algorithms . For example , the AMADEUS platform calculates localization bias based on a binned enrichment score [80] , while the FIRE program uses mutual information to detect motif position bias [79] . Here , a z-score based on uniform distribution was used to measure motif position bias [27] . Our analysis provides evidence that motif position bias could be used as an effective tool to identify gene expression modules . In the four motifs studied here , our analysis based on motif position bias performed as well as ( for G-box and W-box motif ) or even better ( for MYB and the site II element motifs ) than analyses based on motif enrichment . For some of the identified motif–module combinations , the motif was localized with position bias within the modules without enrichment . Therefore , application of motif position analysis to other known plant promoter motifs has the potential to lead to the discovery of additional novel signaling modules that so far have escaped recognition . Our approach to identify motif based gene expression modules presents a novel step to understand the regulatory mechanisms underlying gene co-expression networks . An important task for gene network analysis is to identify hub genes which serve as the key regulators that determine the expression of other genes within the network . Genes with the most number of connections are usually treated as hubs . Here , we argue that for co-expression modules driven by a specific motif , hub genes should be the TFs that bind to the motif and regulate gene expression . These TFs might not be part of the gene co-expression network and can form regulatory networks themselves ( Figure 11A ) . For the modules identified from our analysis , the potential regulatory motif and TF family that govern the structure of a co-expression module can be identified . In turn , the rapid TF-promoter interaction screening system based on the SGR gene provides a fast method to identify the exact transcription factor ( s ) that drives the expression of a specific module , thus revealing the specificities for the TFs within the same family . For example , our results indicated that the MYB-motif containing SUR1 promoter is only activated by a subset of MYB TFs , while CYP98A3 promoters are activated by another subset of MYB TFs . On the other hand , some MYB TFs do not activate any of the selected promoters whose targets might reside in the MYB modules we have not tested . It is intriguing how such specificities between different MYB TFs and different MYB motif containing promoters are achieved . The specificities might be determined by different MYB motif variants , or the nucleotides flanking the core MYB motifs , or the combinatorial effects from other motifs in the same promoters . As another advantage , our analysis also benefits from an existing library collection generated in our laboratory for the expression of plant proteins in N . benthamiana for protein microarray productions including 1 , 100 Arabidopsis transcription factors [5] , [8] ( Ma et al . , unpublished data ) . Finally , coupling the gene co-expression network , module analysis , and gene expression visualization provides a powerful way to study gene signaling systems . First , applying gene expression visualization on co-expression modules can easily determine if the response of the genes are mirrored by the same stimulus , i . e . W-box module I , II , III ( Figure S6 ) , or whether genes share similar or identical expression pattern in particular tissues , i . e . W-box module IV , V ( Figure S7 ) . Second , by comparing different expression modules , general frameworks of signaling pathways can be outlined . For example , Figure S18 shows the expression of three modules induced by pathogens , namely via the G-box , MYB , and W-Box motif , respectively . Both MYB and W-Box modules are induced by MAMPs and pathogens , and were repressed by Pst DC3000 . However , only the W-Box module was repressed by ABA treatment . Therefore , these two modules represent two different branches of the basal immunity pathways regulated by MYB and WRKY transcription factors respectively . The MYB module mainly contains lignin biosynthesis genes and our rapid SGR screening system identified MYB 14 , 15 , or 32 could be their regulators . A model for such regulation is depicted in Figure 11B which can be further tested using different MYB mutant lines . As discussed before , the bZIP module might be induced by Pst DC3000 effector proteins delivered into plant cells via ABA pathway . It will be interesting to test the potential repression of the W-Box modules by pathogen effectors in dependence on ABA . In conclusion , we provide a robust approach useful for the identification of gene co-expression modules regulated by known promoter motifs that can be extracted from gene co-expression networks . These predicted TF-promoter interactions could be verified easily using a novel rapid screening system based on SGR reporter gene expression . The algorithm will be available freely for downloading to aid in the identification of expression modules based on motifs selected by the user .
We used an Arabidopsis gene co-expression network based on the Graphical Gaussian model described before [15] , [26] . The software package GeneNet was used when constructing the network [16] , [81] . From this network , 120 , 276 gene pairs with absolute values of partial correlation co-efficient > = 0 . 05 ( pValue< = 7 . 03E-49 ) were chosen for the analysis , which contained 16 , 456 genes ( Additional data file 1 ) . The Arabidopsis promoter dataset was downloaded from TAIR ( ftp://ftp . arabidopsis . org/Sequences/blast_datasets/TAIR10_blastsets/upstream_sequences/TAIR10_upstream_1000_20101104 ) . The promoters are defined as the first 1 , 000 bp upstream of the 5′ UTR or upstream of translation start codon if no 5′ UTR data were available of the 33 , 602 TAIR 10 gene loci . Our algorithm works with any promoter motifs described as IUPAC consensus word sequences , consisting of the nuclides A , C , G , T , and wobble nucleotides r ( A or G ) , y ( C or T ) , s ( G or C ) , w ( A or T ) , m ( A or C ) , k ( G or T ) , or n ( any base ) . Many plant promoter motifs are registered as such consensus word sequences in the AGRIS and PLACE databases [82] , [83] . We chose four well-known motifs for the current study . Motif enrichment was assessed based on hypergeometric distribution . For a given motif , a pValue of motif enrichment was calculated for every gene in the network . Suppose a gene and all the genes immediately connected with it form a group of genes with M promoters in total , and a motif presents in m promoters among them . Within the K promoters in the whole Arabidopsis genome , the motif presents in k promoters . A pValue for that motif and gene combination is calculated as: Motif position bias towards TSS was assessed based on the uniform distribution [27] . For a given motif , a z-score of motif position bias was calculated for every gene in the network . Suppose a motif appears n times in the promoters of a gene and all the immediately connected genes . The locations of these n motif instances relative to TSS is p1 , p2 , … , pn , and their mean value is p . A z-score for that motif and gene combination is calculated as:where L is the length of the promoters , and l is the length of the motif . The motif position is the midpoint of the motif relative to TSS . For orientation , we describe p = 0 as the position at TSS , and p = −1000 at position of 1000 bp upstream of TSS . For a given motif , genes with pValue of motif enrichment smaller or equal to cut-off were selected . A sub-network was extracted from the gene co-expression network for these genes . A sub-network can also be extracted for all the genes with z-score value larger or equal to a selected cut-off value . Network visualization was carried out using the neato program with the “stress Majorization” algorithm which is included in the software package Graphviz 2 . 21 [84] , [85] . The lay-out of the sub-network is then visually inspected for modules . GO enrichment analysis was then conducted by genes within these modules . Permutation experiment on randomized promoters was carried out to measure false discovery rate . Two steps were employed to randomize promoter sequences . First , each of the 33 , 602 promoter sequences in the TAIR Arabidopsis promoter dataset was randomized within itself . The order of nucleotides was completely shuffled but the total numbers of each type of nucleotide were kept the same . Then the resulting promoter sequences were randomly assigned to each of the 33 , 602 genes without replacement . Gene expression module discovery was then carried out on these randomized promoters and false discovery rate calculated . We used an in-house developed software package called MotifNetwork to conduct the above mentioned motif enrichment analysis , motif position bias analysis , sub-network extraction , and permutation analysis . The algorithm is provided through our website ( http://dinesh-kumarlab . genomecenter . ucdavis . edu/downloads . html ) and upon request for academic use . Transcription profiling of Arabidopsis gene expression in different tissues or gene expression regulation upon treatments with different abiotic stresses , hormones , pathogen elicitors , pathogens , and different light regimens were obtained from the AtGenExpression project [86] , [87] . The data were downloaded from WeigelWorld ( http://www . weigelworld . org/resources/microarray/AtGenExpress ) and TAIR ( http://www . arabidopsis . org/portals/expression/microarray/ATGenExpress . jsp ) . Data were processed as previously described [38] . Table S5 lists the treatments used in the gene regulation profiling experiment in Figure S1 , S2 , S6 , S17 , and S18 . Table S6 lists the tissues used in the tissues expression profiling experiments in Figure S3 and S7 . A TF-promoter interaction screen system was developed based on the stay green gene ( SGR ) . A gateway vector , SPDK2388 , was generated with a gateway cassette placed in front of SGR . The promoter::SGR construct was generated via gateway cloning of the selected promoters ( 1000 bp ) . Previously , we built an expression library for expressing Arabidopsis proteins in plants [5] , [8] ( Ma et al . , unpublished data ) , which include the 82 TF genes used in this analyses . For SGR-based screening , selected promoters were cloned into SPDK2388 , and transferred into Agrobacterium tumefaciens GV2260 . Over-night cultures of Agrobacterium with selected promoter vectors were centrifuged and re-suspended to O . D600 = 0 . 1–0 . 3 with infiltration medium ( 10 mM MgCl2 , 10 mM MES , 200 mM acetosyringone ) , and mixed with TFs Agrobacterium O . D600 = 1 . 0 . The mixed Agrobacterium cultures were then spot-infiltrated into 5 week-old N . benthamiana leaves . The infiltrated spots were inspected at 48 to 96 hours after infiltration for signs of yellowing . TF-promoter interactions were also analyzed with the dual luciferase system according to the protocol described in [71] . Briefly , selected promoters were cloned into the pBGWL7 [88] vector to make Promoter::LUC cassette , and transferred into A . tumefaciens GV2260 . The transferred Agrobacteria were then co-infiltrated into 5 week old N . benthamiana leaves with Agrobacteria containing a vector to constitutively express hRenilla genes and Agrobacteria containing different TFs . Leaf disc of 1 cm in diameter from the infiltrated spot were collected and used for luciferase and Renilla fluorescence measurement using the Dual-Luciferase Reporter Assay System ( Promega , Fitchburg , WI ) as described in [71] . | Gene co-expression networks unite genes with similar expression patterns . From these networks , gene co-expression modules can be identified . A specific family of transcription factor ( s ) may regulate the genes within a co-expression module . Thus , module identification is important to decipher the gene regulatory network . Previously , module identification relied on clustering the gene network into gene clusters that were then treated as modules . This represents a top-down approach . Here , we introduce a reverse approach aiming at identifying gene co-expression modules regulated by known promoter motifs . For a given promoter motif , we calculated the probability of each gene within the network to belong to a module regulated by that motif via motif enrichment analysis or motif position bias analysis . A sub-network containing the genes with a high probability of belonging to a motif driven module was then extracted from the gene co-expression network . From this sub-network , the modular structure can be identified via visual inspection . Our bottom-up approach recovered many known and novel modules for the G-box , MYB , W-box and site II elements motif , whose expression may be regulated by the transcription factors that bind to these motifs . Additionally , we developed a rapid transcription factor-promoter interaction screening system to validate predicted interactions . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [] | 2013 | Incorporating Motif Analysis into Gene Co-expression Networks Reveals Novel Modular Expression Pattern and New Signaling Pathways |
Differential DNA methylation is an essential epigenetic signal for gene regulation , development , and disease processes . We mapped DNA methylation patterns of 190 gene promoter regions on chromosome 21 using bisulfite conversion and subclone sequencing in five human cell types . A total of 28 , 626 subclones were sequenced at high accuracy using ( long-read ) Sanger sequencing resulting in the measurement of the DNA methylation state of 580427 CpG sites . Our results show that average DNA methylation levels are distributed bimodally with enrichment of highly methylated and unmethylated sequences , both for amplicons and individual subclones , which represent single alleles from individual cells . Within CpG-rich sequences , DNA methylation was found to be anti-correlated with CpG dinucleotide density and GC content , and methylated CpGs are more likely to be flanked by AT-rich sequences . We observed over-representation of CpG sites in distances of 9 , 18 , and 27 bps in highly methylated amplicons . However , DNA sequence alone is not sufficient to predict an amplicon's DNA methylation status , since 43% of all amplicons are differentially methylated between the cell types studied here . DNA methylation in promoter regions is strongly correlated with the absence of gene expression and low levels of activating epigenetic marks like H3K4 methylation and H3K9 and K14 acetylation . Utilizing the single base pair and single allele resolution of our data , we found that i ) amplicons from different parts of a CpG island frequently differ in their DNA methylation level , ii ) methylation levels of individual cells in one tissue are very similar , and iii ) methylation patterns follow a relaxed site-specific distribution . Furthermore , iv ) we identified three cases of allele-specific DNA methylation on chromosome 21 . Our data shed new light on the nature of methylation patterns in human cells , the sequence dependence of DNA methylation , and its function as epigenetic signal in gene regulation . Further , we illustrate genotype–epigenotype interactions by showing novel examples of allele-specific methylation .
After deciphering the sequence of the human genome , the study of epigenetic processes which initiate and maintain heritable patterns of gene expression and gene function without changing the DNA sequence , has moved into the center of research [1] . The epigenome comprises different modifications of histone proteins including acetylation , ubiquitination , phosphorylation and methylation working in concert with methylation of the DNA [2] , [3] . In mammals , DNA methylation predominantly occurs at CpG dinucleotides , the majority of which are methylated under normal cell conditions [4] . CpG sites are underrepresented in the human genome but cluster in CpG-islands which overlap with the annotated transcriptional start sites ( TSS ) of about 70% of all human genes [5] and mostly are unmethylated in normal differentiated cells [6] . DNA methylation has been shown to play important roles in the regulation of gene expression , development , genomic imprinting , X chromosome inactivation , and genome stability [7]–[9] . Erroneous DNA methylation contributes to the development of human cancer and multifactorial diseases [10]–[12] . Various high-throughput technologies for the analysis of DNA methylation in human genomes have been developed recently [13] , [14] . In principle , these technologies are based on three approaches to discriminate the methylated and unmethylated cytosines in CpG sites . 1 ) Digestion of genomic DNA with methylation sensitive restriction enzymes to discriminate and/or enrich methylated and unmethylated DNA and employ two-dimensional electrophoresis [15] , PCR [16] , microarray [17] or paired-end sequencing [18] for analysis . These methods only provide methylation data related to the restriction enzyme recognition sites . 2 ) Enrichment of methylated or unmethylated fractions of genomic DNA with antibodies against methylated cytosine , methyl-CpG binding domains or other protein domains and readout by microarray or DNA sequencing [19]–[23] . The resolution of this approach is limited by the fragment size . 3 ) Bisulfite conversion of DNA leading to the selective deamination of cytosine but not 5-methyl cytosine [24] , [25] and the sequencing of subsequently generated PCR products either directly [26] or after subcloning as done here . Next generation ultra-deep sequencing methods were recently used for the analysis of the bisulfite converted genomic DNA from Arabidopsis [27] , [28] , as well as for analysis of bisulfite converted DNA enriched for CpG island sequences in mouse [29] . The suitability of these methods for establishing reference maps of DNA methylation has been evaluated recently in silico [30] . The sequencing of subcloned single DNA molecules , as carried out in this study , provides the most reliable and detailed information of the methylation pattern for every single CpG site in a relatively long region of about 300 to 500 base pairs ( bps ) , when analyzed by conventional Sanger sequencing . Furthermore , it provides qualitative and quantitative information of allele-specificity of DNA methylation . Drawbacks of this method are the relatively high costs for conventional Sanger sequencing and the time-consuming need to establish suitable primers for each amplicon of interest . Therefore , we focused our work on chromosome 21 , which is the smallest human autosome . It is of special biomedical relevance due to its association with genetic diseases including trisomic 21 causing Down syndrome , which is the most common genetic cause of reduced cognitive abilities .
We aimed to establish a comprehensive map of DNA methylation at promoter regions on chromosome 21 . All protein-coding genes on chromosome 21 annotated in Ensembl , UCSC and RefSeq gene were investigated in a window from 2000 bps upstream to 500 bps downstream of the annotated transcriptional start site for CpG density and GC content . For this study , we selected genes which show an enriched CpG density in their promoter region . This includes genes which contain a CpG island in their promoter as defined by the widely used Takai/Jones criteria [31] and also genes with weaker CpG islands [20] , [32] , [33] . To increase coverage , we investigated more than one amplicon for a subset of genes , in particular for those with well-annotated alternative transcriptional start sites ( TSS ) . In total , we analyzed the DNA methylation pattern of 297 amplicons from 190 gene promoters by using bisulfite conversion , subcloning and sequencing as the major experimental method . The study was performed in five cell types , viz . human peripheral blood ( mainly leukocytes ) , fibroblast , the human embryo kidney cell line HEK293 , the human hepatocellular liver carcinoma cell line HepG2 and fibroblast cells derived from a patient with Down syndrome ( trisomic 21 ) . A statistical summary of the analysis is provided in Table 1 . All methylation data obtained here are presented in an integrated web platform ( http://biochem . jacobs-university . de/name21/ ) for visualization and download which also includes additional technical information . Furthermore , the results are provided as custom annotation tracks to be displayed in the UCSC Genome Browser [34] . The DNA methylation levels of all amplicons in all studied cell types are shown in Figure 1A . Methylation of some of our amplicons has been studied previously . In such cases , our data in general fit well with previous results obtained for the overlapping DNA region in the same tissues ( Text S1 ) . 57% ( 168/297 ) of the amplicons show similar methylation level ( methylation difference <30% ) in all five cell types . The remaining amplicons are differentially methylated between two or more cell types . We clustered all five cell types according to their average DNA methylation levels for all amplicons ( Figure 1C ) . The results show that DNA methylation levels are more similar between related cell types , e . g . between transformed cell line HEK293 and cancer cell line HepG2 and between the two types of primary cells used here , blood and fibroblasts . As seen previously , the average CpG island methylation in cultured cells or cancer cell lines is higher than in primary tissues [29] , [35] , [36] . We tested the effect of 5-azacytidine treatment on methylated regions in the HEK293 cells . As shown in Figure 1B and Text S2 , we observed a heterogeneous response in which about 40% of the amplicons were massively demethylated but about 10% were almost completely refractory to demethyation . As shown in Figure 1D , the methylation levels of all amplicons studied showed a bimodal distribution with an enrichment of highly methylated and unmethylated sequences , a result that as has been observed previously as well [20] , [26] , [29] . Most of our amplicons are situated in promoter regions and show low methylation ( 62% of them having methylation levels <30% ) . However , 25% of the amplicons are highly methylated with methylation levels >70% . The bimodal methylation level distribution was also observed at the level of the individual CpG sites and clones analyzed ( Figure 1D ) . Trisomic cells are expected to exhibit an 1 . 5-fold increase in gene expression levels when compared to normal cells . However , it is known that epigenetic modifications can alter such effects . For example , DNA methylation is involved in dosage compensation by X-chromosome inactivation in females [37] . To test if similar compensatory effects are mediated by methylation changes of the promoters on chromosome 21 in trisomic patients , we investigated DNA methylation patterns in trisomic fibroblasts and compared it to normal fibroblasts . In total , DNA methylation of 252 amplicons ( corresponding to 169 genes ) can be compared between trisomic and normal fibroblast cells . The results indicate that only a small number of amplicons ( 7 out of 252 ) are differentially methylated with a methylation difference >30% ( in fact , trisomic fibroblast and normal fibroblast were most similar among all pairs of cell types studied here ) . Therefore , DNA methylation does not appear to be a general mechanism of global gene dosage correction in cells trisomic for chromosome 21 . The data from trisomic fibroblast cells were not included in the following analyses in order to prevent statistical overweighting of fibroblast methylation data . Using our high-quality set of DNA methylation profiles with single base pair and single allele resolution , we addressed the question of the stability of natural DNA methylation patterns . To this end , a subset of data was extracted only containing the PCR products with unimodal distribution of methylation levels among the clones and average methylation levels between 20 and 80% . Then , we compared the methylation levels of all clones from each PCR product in the set with the average methylation level of the respective PCR product ( Figure 2A ) . Our data show that the methylation levels of clones from each particular PCR product were very similar to the average of the respective PCR product indicating that the methylation levels of individual cells from one tissue are similar . We then examined the type of the DNA methylation pattern . There were two extreme possibilities , with the methylation either distributed in a site specific pattern ( meaning that each site is either fully methylated or unmethylated ) or in a stochastic pattern which would predict that the average methylation of each site equals the average methylation of the PCR product ( Figure 2B ) . A stochastic pattern would only preserve the average methylation of clones but not the sites of methylation . Visual inspection of the data ruled out a strict site specific pattern . To determine if the methylation is stochastically distributed or if there are particular preferences to methylate some sites , we extracted the average methylation levels of all CpG sites of PCR products from the unimodal set and compared to the average methylation level of the respective PCR products . As shown in Figure 2C , the results significantly differ from what would be expected by a stochastic distribution , because the average methylation levels of sites cluster at levels higher and lower than the average methylation level of the corresponding PCR product . To further examine the statistical significance of this finding , the p-value of the methylation patters of each site was calculated by exact binominal test assuming a stochastic methylation pattern . As shown in Figure 2D , the fraction of sites with small p-values was much larger than statistically expected . We conclude that the methylation level of CpG sites is not strictly site specific , but there are significant differences in the methylation of individual sites , which cannot be explained by statistical fluctuation – we call it a relaxed site specific pattern . We often observed that amplicons next to each other in the same CpG island had different methylation states ( see Figure 3 for an example ) . In our data set there are 164 examples , where more than one amplicon located on one CpG island was studied in one of the non-trisomic cell types . In 35 of these cases ( 21% ) , the amplicons were differentially methylated ( with a methylation difference >30% ) . Differential methylation within one CpG island happened more often in HEK293 ( 13 out of 41 , 32% ) and HepG2 cells ( 11 out of 41 , 27% ) , than in fibroblast cells ( 6 out of 40 , 15% ) and leukocytes ( 5 out of 42 , 12% ) . It was reported that there is an unmethylated core region surrounding TSS within 1 kb of genes [26] . We checked if the amplicons on the same CpG island also show such tendency . Among the 35 cases with methylation difference on the same CpG island , 11 were not informative , because there are either two TSSs for one gene or different TSSs for two genes annotated . Hence it was not possible to determine the distance between TSS and amplicon . In 20 out of the remaining 24 cases ( 83% ) , we observed the methylation of amplicons gradually decreased when approaching the TSS of the respective gene both from upstream and downstream . An example for this is shown in Figure 3 . The GC content and CpG density are two critical properties for the biological effects of CpG islands . We calculated the GC content and CpG density of all studied amplicons and compared with DNA methylation . As shown in Figure 4A , the amplicons with high GC content and CpG density tend to be low methylated while those with low GC content and CpG density tend to be highly methylated , which is consistent with previous results [20] , [26] , [29] . The correlation between periodic distribution of CpG and DNA methylation has been reported recently [38] . We studied the distribution of CpG pairwise distances in all amplicons in our dataset . We observed three significantly overrepresented CpG distances of 9 , 18 , and 27 bps in highly methylated ( >70% ) amplicons , when comparing with low methylated ( <30% ) amplicons ( Figure 4B ) . Simulations using data sets in which the observed DNA methylation levels were randomly connected to the amplicon sequences , indicated that these overrepresentations are highly significant with p-values <1 . 1×10−4 ( Figure 4B ) . The same result was also obtained after separating amplicons for methylation <50% and >50% ( data not shown ) . The overrepresentation of pairs of CpG sites in distances of 9 bps and multiples of this in the sequence of highly methylated DNA could be correlated to the preference of the DNMT3a DNA methyltransferase for methylation of CpG pairs in that particular distance [38] , [39] . With our large dataset of methylation states of CpG sites , we studied the effect of flanking sequence on DNA methylation . The flanking sequences ( 20 bases ) of each methylated CpG site were collected using different methylation thresholds ( Figure 5 and Text S3 ) and the occurrences of all four bases at each position were compared with the flanking sequences of all CpG sites used as reference . The results revealed that several bases at different positions were significantly over- or underrepresented in the flanking sequences of methylated CpG sites with p-values <1 . 25×10−4 ( Text S3 ) . These differences were reproducible when using different thresholds of methylation level ( Text S3 ) and using different combinations of data from only two or three cell types ( data not shown ) . No statistically significant fluctuations were detected when using randomized methylation data for the same analysis . From the overall point of view , A/T flanks correlate with DNA methylation while G/C flanks are correlated with the absence of DNA methylation . This result reflects the previous observation that bona fide CpG islands with high GC content and CpG density tend to be unmethylated while non-CpG islands tend to be methylated ( see above ) . However , it is noteworthy that this tendency was not equally observed at all flanking positions . DNA methylation is known to lead to gene silencing [8] . However , so far this observation was mainly based on experiments with individual genes . We compared our gene promoter methylation data with gene expression data extracted from a serial analysis of gene expression ( SAGE ) database [40] . Matching data sets were available for HEK293 and leukocytes . In both cell types , high methylation of amplicons is correlated with low gene expression , and high gene expression is correlated with low methylation ( Figure 6A ) , demonstrating that DNA methylation and gene expression are inversely correlated . Some methylated amplicons from expressed genes have been excluded from this analysis , because there was more than one amplicon analyzed for that particular gene and at least one was found to be unmethylated or because there were alternative start sites annotated for the gene ( Text S4 ) . The distribution of RNA polymerase II pre-initiation complex ( PIC ) is another indicator of gene expression . Kim et al ( 2005 ) mapped the PIC binding sites across the genome by immunoprecipitation of TFIID-bound DNA from primary fibroblast cells [41] . We extracted all 92 PIC binding positions that are within 2 . 5 kb of the annotated TSS of genes on chromosome 21 . These PIC binding sites are related with 69 gene promoters , out of which 67 genes ( with 93 amplicons ) are with DNA methylation data . 95% of these gene promoters ( 88 out of 93 ) exhibit low levels of DNA methylation ( <30% ) in fibroblast . Comparing with the proportion of low methylated genes in the whole dataset of fibroblast ( 67% ) , the absence of DNA methylation is highly significant in the genes occupied with PIC ( with p-values <10−10 according to exact binominal test ) ( Figure 6B ) . The absence of methylated genes in the PIC occupancy data set is consistent with the inhibitory role of DNA methylation on gene expression . Beside DNA methylation , histone modification is another important mechanism of epigenetic regulation of gene expression . Therefore , we assessed the correlation of DNA methylation and the following histone modifications: ( i ) histone H3 lysine 4 trimethylation ( H3K4me3 ) and ( ii ) histone H3 lysine 9 and 14 acetylation ( H3K9ac/H3K14ac ) , all of which are known as activating marks [3] . Their distributions were mapped across the nonrepetitive portions of chromosome 21 and 22 in HepG2 cells [42] , which allowed us to correlate them with the HepG2 DNA methylation data . The distances between the regions with DNA methylation data and histone modifications were calculated up to 10 kb . The results indicate a strong correlation between the absence of DNA methylation and the presence of H3K4me3 and H3K9ac/K14ac , up to distances of 1 kb ( Figure 7A ) . Polycomb-mediated histone H3K27 methylation was found to pre-mark genes for de novo methylation in cancer [43] , [44] . Suz12 is one core subunit of PRC2 and is essential for its activity . A genome-wide mapping of Suz12 binding sites in human embryonic stem ( ES ) cells revealed that Suz12 is enriched at a special set of developmental genes which are generally repressed to maintain pluripotency of ES cells and preferentially activated during ES cell differentiation [45] . It was also demonstrated that Suz12 co-occurred with histone H3K27 trimethylation at most genes [45] . We extracted the published Suz12 binding data on chromosome 21 and correlated them with our DNA methylation data . 17 amplicons from 13 genes were exactly overlapping with Suz12 binding sites ( Text S5 ) . We observed that almost half of all amplicons with polycomb binding show increased methylation in the cancer cell line HepG2 and/or the transformed cell line HEK293 which corresponds to a high enrichment when compared to the whole dataset where only about 15% show increased methylation ( Figure 7B ) . To check the biological functions associated with genes having methylated or unmethylated promoters , we analyzed the gene ontology ( GO ) categories of genes with methylated promoter ( >50% ) in at least one of the cell types , genes that were low methylated ( <30% ) in all studied cell types and genes with differentially methylated promoter among cell types ( methylation difference >30% in at least three pair wise comparisons ) by using GOTM ( Gene Ontology tree machine: http://bioinfo . vanderbilt . edu/gotm/ ) [46] . All genes analyzed with GO annotations were used as reference list . The analysis revealed that genes with methylated or unmethylated promoters are significantly overrepresented in distinct GO categories ( Figure 8 ) . The genes with methylated promoters are overrepresented in “sensory perception” and “physiological response to stimulus” categories or they encode for structural genes like collagens while unmethylated genes are significantly overrepresented in “transferase activity” and “ATP binding” categories . Genes with differentially methylated promoters are overrepresented in “immune response” and “cell-cell signalling” categories . Different methylation levels of the two alleles of one gene within one cell ( allele-specific DNA methylation ) has been observed in imprinting regions , where methylation of one allele occurs on a parent of origin basis [47] and in X chromosome inactivation in females [48] . There are also reports about sequence dependent allele-specific methylation in non-imprinted loci the in human genome [49] . Using our single allele resolution DNA methylation dataset , we checked for the presence of allele-specific DNA methylation on chromosome 21 . To this end , the data were filtered for a biphasic distribution of DNA methylation levels . Hits were then manually inspected for the occurrence of SNPs in the sequenced region which can be used to differentiate the alleles . Our results indicated the presence of allele-specific methylation in three regions in leukocytes derived from a healthy individual ( Figure 9 and Text S6 ) . Two amplicons ( 176_1 and 176_2 ) on a CpG island overlapping with the first exon of the CBR1 gene ( carbonyl reductase 1 ) showed strong allele-specific methylation in leukocytes ( Figure 9A ) . For amplicon 176_2 the C allele was highly methylated ( average: 66% ) , while G allele was low methylated ( average: 7% ) . For amplicon 176_1 about half of the clones showed a deletion and were partially methylated , while all clones without the deletion were unmethylated . This result suggests that the allele-specific methylation spans the whole CpG island . In other tissues , we did not observe a SNP in these amplicons . In HepG2 , biphasic methylation was observed , while in HEK293 and fibroblast both amplicons were completely unmethylated ( Text S6 ) . Amplicon 23_2 located on the first exon of gene C21orf81 showed massive allele-specific methylation in leukocytes with the A and C alleles methylated to 28% and 94% on average ( Figure 9B ) . Amplicon 23_1 immediately adjacent to 23_2 also showed biphasic methylation in leukocytes , but did not contain an SNP . In other tissues , no allele-specific methylation was observed ( Text S6 ) . Amplicon 197_intern located in an internal region of gene DSCR3 showed ASM in leukocytes with the A allele either being methylated or unmethylated and the C allele always being unmethylated ( Text S6 ) . This result agrees with data obtained by Yamada et al ( 2004 ) who identified this region of allele-specific methylation in blood previously [16] . For this amplicon , fibroblast and trisomic fibroblast contained the SNP but were both unmethylated , HepG2 showed only the A allele and was unmethlylated and HEK293 showed biphasic methylation but only contained the C allele .
For many years , it was believed that CpG islands are mainly unmethylated and there exists little difference in the DNA methylation of different cell types . Here , we show for five different cell types that CpG islands frequently are methylated and that there is a substantial difference in the methylation pattern of different cell types . Based on the most commonly used criteria for defining CpG islands ( GC content ≥55% and CpG observed vs . expected ≥0 . 65 [31] ) , 14% of the amplicons on CpG islands show dense methylation in different cell types . We also observe that a significant number of CpG islands exhibit substantial differences in their average DNA methylation levels in different parts . This result underscores the importance of the position where DNA methylation is studied , in order to draw valid biological conclusions . Our data confirm that high levels of DNA methylation in CpG rich promoters are strongly associated with down-regulation of gene expression . However , the inverse relation does not hold , because there are many examples of unmethylated genes that are not expressed , possibly due to lack of expression caused by other mechanisms than DNA methylation like absence of the relevant activating transcription factors . With bisulfite subcloning and sequencing technology , we provide methylation data at single-allele resolution , which makes it possible to investigate cell-specific mosaicism and allele-specific methylation patterns . Methylation levels of different cells in the same tissue are very similar in general . We observed allele-specific methylation at two regions on chromosome 21 that were not identified before and also confirmed allele specific methylation at one region previously shown to be specifically methylated in the maternal allele [16] . Interestingly , our data show that allele-specific methylation in all three cases is not observed in all tissues analyzed . Further studies have to determine if the methylation differences were due to parent of origin dependent methylation that has been lost in some tissues or if some tissues underwent allele-specific methylation changes that were triggered by the genetic polymorphisms between the two alleles , which would provide an example of genotype-epigenotype interactions . Allele-specific methylation might contribute to allele-specific expression that is a widespread phenomenon in the human genome [50] , [51] . There are interesting connections between DNA sequence and DNA methylation . We and others observed a strong anti-correlation of DNA methylation with CpG density and GC content , indicating that DNA methylation declines the more the sequence resembles a CpG island [20] , [26] , [29] . This trend is also reflected in our flanking sequence preferences of DNA methylation showing that highly methylated CpG sites are flanked by A/T rich sequences while unmethylated ones tend to be embedded in G/C rich sequences . The underlying mechanism of this phenomenon is not clear at present , however , the flanking sequence preferences suggest that not all neighboring base pairs are of equal importance for the DNA methylation state . One potential connection between DNA methylation and DNA sequence could be the flanking sequence preferences of the DNA methyltransferases . We observed an overrepresentation of A at −1 and T at +1 flanking position which agrees with the experimental flanking preferences of Dnmt3a and 3b [52] . This observation suggests a potential link to the sequence preference DNMT3a and 3b and confirms a similar conclusion based on a smaller data set [52] although preferences observed here at larger distances to the CpG site are slightly different , which is most likely due to the smaller amount of data used in the previous analysis . Dnmt3a and 3L have been shown to form heterotetramers with two active sites in a distance of 8–10 bps and both Dnmt3a/3L and Dnmt3a tend to co-methylate CpG sites in that distance [38] , [39] . The similar periodicity in the occurrence of CpG sites was also observed in the differentially methylated regions from 12 maternally imprinted mouse genes [38] . In the Arabidopsis genome , a periodicity of 10 nucleotides was found for CHH methylation [27] . Here , we observe for the first time a genome wide enrichment of pairs of CpG sites in distances of 9 , 18 and 27 bps in highly methylated DNA when compared to unmethylated DNA . This effect could be due to the preferential methylation of sites in such distances . The occurrence of multiples of 9 could be related to the multimerisation and protein nucleofilament formation observed with Dnmt3a and Dnmt3a/3L [38] , [39] . However , our results clearly illustrate that DNA methylation is a dynamic mark , the pattern of which cannot be explained at the level of the DNA sequence alone , because 43% of the amplicons ( 129 out of 297 ) showed significant differences in DNA methylation between at least two out of the five cell types analyzed . In extrapolation to the total number of cell types in the human body , it is likely that most of the CpG islands will show differential methylation in some cell types . DNA methylation is tightly connected to other forms of epigenetic signaling . We observed a strong anti-correlation of DNA methylation and histone H3K4me3 and H3K9ac/K14ac indicating that these activating marks and the repressive DNA methylation mark are mutually exclusive at the genomic scale . Similarly as reported previously [21] , [53] , we observe that methylation of polycomb marked genes tends to be increased in cancer or transformed cells suggesting that polycomb signaling is connected to DNA methylation , perhaps by recruitment of DNA methyltransferases by polycomb proteins . Our data also shed some light on the biological function of epigenetic gene regulation . Unmethylated genes are significantly overrepresented in functional categories , which suggest important metabolic functions of these unmethylated genes similarly as originally proposed for CpG island associated genes [54] . The functional categories of genes with methylated promoters suggest that DNA methylation has an important effect of on regulation of cell type specific genes determining both cellular physiology and morphology .
Blood DNA was extracted using QiaAmp Blood DNA Mini kits ( Qiagen ) . Human embryo kidney ( HEK293 ) cells were grown in Dulbecco's Modified Eagle's Medium ( DMEM ) with 10% ( v/v ) fetal bovine serum ( FBS ) at 37°C in 5% ( v/v ) CO2 . Human hepatocellular carcinoma cells HepG2 ( ATCC: HB 8065 ) were cultured in 90% RPMI 1640 supplemented with 10% FBS at 37°C in 5% ( v/v ) CO2 . Fibroblast cells ( CCD-1059SK ) were cultured in Eagle's Minimum Essential Medium ( EMEM ) supplemented with 10% FBS at 37°C in 5% ( v/v ) CO2 . Trisomic 21 fibroblast cells ( Coriell Cell Repositories: AG08941 ) were grown in EMEM with Earle's salts and non-essential amino acids supplemented with 15% FBS at 37°C in 5% ( v/v ) CO2 . The genomic DNA from cells was extracted using QIAamp DNA Mini kit ( Qiagen ) . For the demethylation treatment of HEK293 cells , 2 µM 5-azacytidine was added to the media . During three days of treatment , the media was exchanged daily and cells were harvested five days after starting the experiment . Bisulfite methylation analysis was performed basically as described [55] . Briefly , the CpG island searcher program ( http://www . cpgislands . com ) [56] or the CpGPlot program ( http://www . ebi . ac . uk/Tools/emboss/cpgplot/index . html ) were used to check the presence of a CpG island or CpG rich region in gene promoter . RepeatMasker software ( http://www . repeatmasker . org ) was used to identify the presence of repeat sequences . Bisearch ( http://bisearch . enzim . hu ) [57] and Methprimer ( http://www . urogene . org/methprimer/index1 . html ) [58] programs were used to design primers . For bisulfite conversion , 200–300 ng genomic DNA were digested with BamHI or SphI ( 40 U ) at 37°C overnight , converted with bisulfite as described [55] and used for PCR . The PCR products were purified by ChargeSwitch PCR Clean-Up Kit ( Invitrogen ) and subcloned using the StrataClone kit ( Stratagene ) . Around 40 clones for each amplicon were picked and sequenced . Plasmid DNA of clones was isolated by an automated alkaline lyses procedure , which includes template purification by PEG-precipitation and subsequently adjustment to similar molarity . DNA sequences were determined using ABI BigDye Terminator chemistry ( BigDye Terminator v3 . 1 K ) and 3730xl ABI 96-capillary sequencer systems equipped with capillaries of 50 cm separation length . The BiQ analyzer software was used to perform quality control and to derive DNA methylation patterns from the sequencing results [59] . BDPC was used to present the methylation pattern , prepare the figures and WEB presentation and compile methylation data [60] . The flanking sequences ( 20 bases ) of each methylated or unmethylated CpG site in all amplicons ( from non-trisomic cell types ) were extracted using different thresholds of methylation level ( ≥90% , ≥80% , ≥70% and ≥60% ) . The over- or under-representation of bases ( observed/expected ) in the flanks of methylated CpG sites was analyzed with Microsoft Excel BINOMDIST function . The flanking sequences of all CpG sites were used as reference . Figure 5 and Text S3 summarize the bases with significant p-value ( p-value <1 . 25×10−4 corresponding to p-value <0 . 01 when considering Bonferroni multiple testing correction ) by using four different thresholds . For an additional statistical validation , we randomized the methylation data of CpG sites 10 times and extracted the respective ( randomized ) flanking sequences . No significant p-value was observed for the flanking sequences from the randomized datasets . We also performed the analysis by using different combination of data from only three cell types and still observed the same results . The distance distribution of pairwise distances between CpG site and the frequency of observed/expected of all pairwise distances was determined in all amplicon sequences within distances of 2–200 bps using the DISTRIDIST program ( http://biochem . jacobs-university . de/cgi/distridist . cgi ) . Amplicons with low GC ( ≤0 . 55 ) and observed/expected CpG dinucleotide content ( ≤0 . 65 ) were excluded from this analysis , because the poor CpG density of them could have caused a bias in the data analysis . For each distance value , the ratio of averaged observed/expected CpG pairwise distances was calculated comparing highly methylated amplicons ( HM , 70%–100% methylation ) with low methylated amplicons ( LM , 0–30% methylation ) . This ratio ( HM/LM ) reflects the over- or underrepresentation of CpG pairwise distances in the highly methylated amplicons as compared to low methylated amplicons . Since there was no difference in results between the different cell types , data were finally averaged over all cell types . To determine the statistical significance of the data , we randomized the methylation level of all amplicons 12 times and performed the same calculation . The randomized distributions were used to derive standard deviation and mean and to calculate p-values assuming normal distribution . The TFIID binding positions on chromosome 21 in IMR90 fibroblast cells were extracted from published results [41] and converted from NCBI Build 34 to NCBI Build 36 . The p-value for the enrichment of unmethylated genes in the amplicons with PIC binding , comparing to the whole dataset , was calculated assuming a binomial distribution . The analysis was performed using GOTM ( http://bioinfo . vanderbilt . edu/gotm/ ) , which is based on hypergeometric test to show the overrepresented gene ontology categories ( p-value <0 . 01 ) [46] . 148 out of 190 genes studied on chromosome 21 were used as reference gene list for the statistical analysis . The other 42 genes are without functional annotation in GOTM , so they could not be included in the analysis . The p-value was also calculated according to BINOMDIST function on the basis of the overrepresentation of gene ontology categories in methylated , unmethylated or differentiately methylated genes when comparing to all genes , as a confirmation of the significance of results . The published histone H3K4me3 and H3K9ac/K14ac modification data in HepG2 cells with a p-value <10−4 were extracted [42] and the positions of them were converted from NCBI Build 33 to NCBI Build 36 . The minimal distance was calculated between regions with DNA methylation and histone modification data . All amplicons correlated with histone H3K4me3 or H3K9ac/K14ac were used as reference to show the normal distribution of amplicons according to methylation level . The published SUZ12 binding data on chromosome 21 in ES cells were extracted from supplementary table S7 of Lee et al . , 2006 [45] . Only amplicons which are exactly overlapping with SUZ12 binding sites were included in this analysis . The different proportions of higher methylated ( methylation increase >30% ) amplicons in HepG2 or HEK 293 cells , comparing to that in normal leukocytes and fibroblast cells , in whole dataset and in amplicons with SUZ12 binding were calculated respectively . The p-value was calculated based on BINOMDIST function . The SAGE database ( http://cgap . nci . nih . gov/SAGE/AnatomicViewer ) presents ranked expression data for different cell types [40] . For each gene the relative tag occurrence as tag per 200 , 000 is provided . Standardization of the relative expression is the key task in this experiment , because one has to discriminate low expression caused by an intrinsically weak promoter from low expression caused by silencing of a strong promoter . Therefore , we quantified the expression potential of each promoter using the six highest expression levels of each gene . This maximal expression was then used to normalize the expression level in HEK293 and leukocytes , which resulted in a percentage of gene expression . Normalized expression data were then compared with the methylation data . | Epigenetics is defined as the inheritance of changes in gene function without changing the DNA sequence . Epigenetic signals comprise methylation of cytosine bases of the DNA and chemical modifications of the histone proteins . DNA methylation plays important roles in development and disease processes . To investigate the biological role of DNA methylation , we analyzed DNA methylation patterns of 190 gene promoter regions on chromosome 21 in five human cell types . Our results show that average DNA methylation levels are distributed bimodally with enrichment of highly methylated and unmethylated sequences , indicating that DNA methylation acts in a switch-like manner . Consistent with the well-established role of DNA methylation in gene silencing , we found DNA methylation in promoter regions strongly correlated with absence of gene expression and low levels of additional activating epigenetic marks . Although methylation levels of individual cells in one tissue are very similar , we observed differences in DNA methylation when comparing different cell types in 43% of all regions analyzed . This finding is in agreement with a role of DNA methylation in cellular development . We identified three cases of genes that are differentially methylated in both alleles that illustrate the tight interplay of genetic and epigenetic processes . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"molecular",
"biology/histone",
"modification",
"genetics",
"and",
"genomics/gene",
"expression",
"genetics",
"and",
"genomics/chromosome",
"biology",
"genetics",
"and",
"genomics/genome",
"projects",
"molecular",
"biology/dna",
"methylation",
"genetics",
"and",
"genomics/epigenetics"
] | 2009 | DNA Methylation Analysis of Chromosome 21 Gene Promoters at Single Base Pair and Single Allele Resolution |
Malaria , dengue fever , and filariasis are three of the most common mosquito-borne diseases worldwide . Malaria and lymphatic filariasis can occur as concomitant human infections while also sharing common mosquito vectors . The overall prevalence and health significance of malaria and filariasis have made them top priorities for global elimination and control programmes . Pyrethroid resistance in anopheline mosquito vectors represents a highly significant problem to malaria control worldwide . Several methods have been proposed to mitigate insecticide resistance , including rotational use of insecticides with different modes of action . Anopheles sinensis , an important malaria and filariasis vector in Southeast Asia , represents an interesting mosquito species for examining the consequences of long-term insecticide rotation use on resistance . We examined insecticide resistance in two An . Sinensis populations from central and southern China against pyrethroids , organochlorines , organophosphates , and carbamates , which are the major classes of insecticides recommended for indoor residual spray . We found that the mosquito populations were highly resistant to the four classes of insecticides . High frequency of kdr mutation was revealed in the central population , whereas no kdr mutation was detected in the southern population . The frequency of G119S mutation in the ace-1 gene was moderate in both populations . The classification and regression trees ( CART ) statistical analysis found that metabolic detoxification was the most important resistance mechanism , whereas target site insensitivity of L1014 kdr mutation played a less important role . Our results indicate that metabolic detoxification was the dominant mechanism of resistance compared to target site insensitivity , and suggests that long-term rotational use of various insecticides has led An . sinensis to evolve a high insecticide resistance . This study highlights the complex network of mechanisms conferring multiple resistances to chemical insecticides in mosquito vectors and it has important implication for designing and implementing vector resistance management strategies .
Malaria and filariasis are two of the most important vector-borne parasitic diseases in Southeast Asia . Although China and several other countries in the region have reported a marked downward trend in malaria cases , high malaria incidence has been observed in the major neighboring endemic countries such as Myanmar , Bangladesh , and India [1] . Cross-border migration provides increased opportunities for malaria infections . Further , the high frequencies of natural disasters augment the risk of imminent outbreaks of malaria . Therefore , malaria surveillance and vector control become very important tools to prevent malaria outbreaks in low-transmission areas [1] . Currently , insecticide-treated bed nets ( ITNs ) and indoor residual spray ( IRS ) are the primary vector control tools in the Global Strategy for Malaria Control and the Roll Back Malaria program [2] and in the Global Fund to Fight AIDS , Tuberculosis and Malaria [3] . Pyrethroids are currently the only class of insecticide approved for use on ITNs [4] due to their high toxicity to insects , rapid rate of knockdown , strong mosquito excito-repellency , and low mammalian toxicity [5] . Reducing vector-human contact by the use of ITNs has been shown effective in reducing malaria transmission [6] , [7] . However , the emergence and spread of insecticide resistance has significantly hampered the efficacy of ITN programs [8] . Insecticides remain the most important vector control method; however , insecticide resistance poses a major threat to vector-borne disease control due to lack of other viable alternatives . Several methods have been proposed to mitigate insecticide resistance in vector mosquito populations , including insecticide rotation strategies [1] and combinational use of insecticides with different modes of action . Resistance to any particular insecticide is mitigated , since the selection pressure is removed before resistance is developed . Combinational use of insecticides with different modes of action is currently applicable to IRS only , with the assumption that mosquito vectors exhibit low resistance to the new insecticides under consideration . Adding IRS to the ITN program becomes an increasingly popular malaria control strategy worldwide due to increasing resistance to pyrethroids used in ITNs [8] . WHO-recommended products for IRS include four classes of insecticides: pyrethroids , organochlorines , organophosphates , and carbamates [1] . However , the long-term consequences of insecticide rotation and combinational use of insecticides on mosquito insecticide resistance are not clear . The mosquito Anopheles sinensis is the most important malaria vector in China and other Southeast Asian countries [9]–[14] . In southern China , An . sinensis play an important role in the natural transmission of both malaria and filariasis ( Wuchereria bancrofti ) [15]–[17] , as well as Romanomermis jingdeensis [18] and Setaria digitata [19]–[21] . The major breeding sites of An . sinensis in China are rice fields where various classes of insecticides have been used in agricultural pest control regimes in rotation [22] , [23] . Although An . sinensis is not the intended target in this pest control regimes , it has been directly exposed to the insecticides over a period of four decades [24]–[26] . Therefore , An . sinensis in China also represents an interesting model in which to examine the consequence of long-term rotational use of various insecticides on resistance evolution . In this study , we examined the extent and distribution of insecticide resistance in An . sinensis against the four classes of insecticides recommended by WHO for malaria vector control by IRS . It also indirectly examined the long-term consequences of rotational use of insecticides on mosquito resistance evolution . Using An . Sinensis populations from central and southern China , general correlations between insecticide-specific resistance and geography can be examined . The information obtained from this investigation can be used to guide insecticide rotation strategies . Another major objective of this study was to examine the importance of target site insensitivity and various metabolic detoxification enzymes in resistance to the major classes of insecticides used in IRS . Pyrethroids and organochlorines function as neurotoxins that act by prolonging sodium channel activation whereas organophosphates and carbamates kill insects by inhibiting acetylcholinesterase found in the central nervous system [27] . The voltage-sensitive sodium channel proteins are the major target site for pyrethroids and DDT , and a mutation at codon 1014 the para sodium ion channel gene causes knockdown resistance ( kdr ) [27] . On the other hand , a mutation at codon 119 of the acetylcholinesterase ( ace-1 ) gene that leads to a single amino acid substitution of glycine to serine in the binding pocket of acetylcholinesterase may confer resistance to organophosphates and carbamates . In addition to target-site insensitivity , metabolic detoxification enzymes—including cytochrome P450 monooxygenases ( P450s ) , carboxylesterases , and glutathione S-transferases ( GSTs ) —may also augment insecticide resistance [27] . The possible pleiotropic role of metabolic enzymes on resistance and the relative significance of each mechanism on resistance to multiple classes of insecticides in An . sinensis are unknown . This information is particularly valuable to the development of reliable molecular-based resistance surveillance tools .
No specific permits were required for the described field studies . For mosquito collection in rice paddies , oral consent was obtained from field owners in each location . No sites were protected by law and this study did not involve endangered or protected species . The study was conducted in malaria endemic sites in southern China ( Yunnan Province ) and central China ( Anhui Province ) ( Fig . 1 ) . Yunnan Province has the highest malaria incidence in China and is responsible for about 50% of officially reported malaria cases in China [28] . Malaria in Yunnan province is mesoendemic with perennial circulation of both P . vivax and P . falciparum parasites [28] , [29] . Anhui Province is hypo-endemic , with Plasmodium vivax as the predominant malaria species [30] . The Yunnan site was located in Yingjiang and Lianghe Counties , Dehong Prefecture , and the Anhui site was in suburbs of Bengbu City . Rice is the major agricultural crop in these study sites , with 1–2 harvests per year . Due to severe insect pest damage to the rice , insecticide use for pest control has been very intensive , with several rounds of sprays administered during each growing season . From 1960 to 1990 , insecticides were extensively used in agriculture in China because the government routinely subsidized pesticide expenses by as much as 85% . The pesticides used were predominantly organochlorines and organophosphates from the 1970s up to the early 1980s . Since the mid-1980s , pyrethroids have been the dominant insecticides with pyrethroids-treated areas constituting more than one third of the total insecticide-treated area in China [26] , [31] . In addition to their agricultural use , pyrethroids have had various public-health applications—as indoor sprays or incense , impregnated in bed nets , or as tools in public sanitation . Other insecticides , including organophosphates and carbamates , have been used widely but less extensively than pyrethroids . During May–August 2012 , Anopheles sinensis mosquito larvae and pupae were collected from irrigated rice fields and small ponds with aquatic plants , using standard 350-ml dippers . We used adults reared from field-collected larvae for this study to minimize the influence of mosquito age and blood feeding history on resistance measurements [4] , [32] . For each site , we collected mosquito larvae from >100 breeding sites in each of the five villages , separated by 5–10 km from each other , to avoid using genetically-related siblings in the subsequent resistance analysis . We collected a total of 4 , 000 anopheline larvae per site . The collected mosquito larvae were transported to the local rearing facility to be reared into adults . All adult mosquitoes were identified to species using the published morphological keys of Dong [33] . An . sinensis adult mosquitoes were provided with fresh 10% sucrose solution daily . After the mosquitoes were identified to species , An . sinensis female adult mosquitoes at 3–4 days post emergence were tested for susceptibility to five insecticides belonging to four classes ( 0 . 05% deltamethrin , 0 . 75% permethrin , 5% malathion , 0 . 1% bendiocarb , and 4% DDT ) , using the standard WHO resistance tube assay [4] . The discriminating dose used for each insecticide should kill 99 . 9% susceptible mosquitoes [4] . As a susceptible mosquito control , we used a laboratory susceptible strain that has been maintained in the insectary of the Jiangsu Institute of Parasitic Diseases in Wuxi , China , for more than 10 years with no insecticide exposure . For each insecticide , a total of 100–150 female mosquitoes were tested in insecticide susceptibility bioassays , with 20 mosquitoes per tube . Equal number of mosquitoes were exposed to the corresponding control papers impregnated with silicone oil ( deltamethrin/permethrin control ) , olive oil ( malathion/bendiocarb control ) , and resila oil ( DDT control ) . After a 1-hr exposure , mosquitoes were transferred to recovery cups and maintained on 10% sucrose solution for 24 hrs , and the number of surviving mosquitoes was recorded . Here , we defined resistant for the mosquitoes alive 24 hours after the end of the bioassay and susceptible for the mosquitoes knocked down during the 60 min exposure time or within 24 hr recovery period [34] . Mosquitoes were considered knocked down if they were unable to walk from the center to the border of a 7-cm filter paper disc , either alone or when they were mechanically stimulated [4] . After the resistance/susceptible status were recorded , one leg of each mosquito was removed and preserved individually in 95% alcohol for subsequent DNA analysis , and the remainder of the body was immediately tested for metabolic enzyme activities . Therefore , only fresh mosquitoes were tested for metabolic enzyme activities . Our definition of susceptible mosquitoes was based on knockdown phenotype , rather than death phenotype . As such , metabolic enzyme activities of susceptible mosquitoes could be measured because fresh mosquitoes were used . This definition allowed us to determine the association between metabolic enzyme activities and resistance with little bias because the resistant and susceptible mosquitoes were exposed to the insecticide in the same manner and our analysis computed the ratio of metabolic enzymes in the resistant mosquitoes to the susceptible mosquitoes . A total of 1 , 103 female adult mosquitoes were used for bioassay in the study . Three metabolic enzymes were analyzed: cytochrome P450 monooxygenases ( P450s ) , glutathione S-transferases ( GSTs ) , and carboxylesterases ( COEs ) . We followed our previously published protocol to measure monooxygenase and GST activities [23] . Mean absorbance values for each tested mosquito and enzyme were converted into enzyme activity and standardized based on the total protein amount . Total protein was measured for each mosquito using the method of Bradford [35] . All measurements were done in duplicate . COE activity was measured following the method of Hosokawa and Satoh [36] . Briefly , 900 µl of p-nitrophenyl acetate solution ( 1 mM ) was transferred to 1 . 5-ml test tubes and incubated at 30°C for 5 min , then 100 µl of mosquito homogenates was added and vortexed for 5 sec . The reaction mixture was transferred to 1 . 0-ml semimicro cuvettes , and the release of p-nitrophenol was measured using a UV/VIS spectrophotometer at 405 nm for 2 min . Spontaneous hydrolysis was used as the blank . COE activity was calculated as µmol of p-nitrophenol formed per min per mg protein , using the formula ( Δabsorbance/min – Δblank/min ) ×1 . 0/16 . 4×0 . 05× protein ( mg/ml ) . An absorption coefficient of 16 , 400 M−1·cm−1 was used [37] . For each mosquito population and each insecticide , 100 female adult mosquitoes were tested . One leg of each mosquito was used for DNA extraction for subsequent PCR-based mosquito species confirmation and mutation detection in kdr and ace-1 genes . DNA extraction was done with the SYBR Green Extract-N-Amp™ Tissue PCR Kit ( SIGMA ) following the manufacturer's protocol . Extracted DNA was stored at 4°C or used immediately . Molecular identifications of An . sinensis species were done by using species-specific primers and amplification of the ITS2 and 28S rDNA regions ( D1 and D2 ) [38] . A total of 100 mosquitoes per site were randomly selected and tested molecularly , and all of them identified as An . sinensis . Detection of point mutation of the kdr gene at codon 1014 was done by using the allele-specific PCR ( AS-PCR ) methods developed by Zhong et al [23] . A PCR-RFLP method was developed to rapidly determine point mutation of the ace-1 gene at codon 119 following the method used in An . gambiae [39] . Briefly , we designed a pair of primers ( Forward primer 467F: GTGCGACCATGTGGAACC , Reverse primer 660R: ACCACGATCACGTTCTCCTC ) based on the An . gambiae ace-1 gene sequence ( GenBank accession: BN000066 ) to amplify a 193-bp fragment that flanks the target codon position 119 in the ace-1 gene ( ace-1R ) . The PCR products of 20 individuals from each population were sequenced in both forward and reverse directions ( GenBank accession: KF697669- KF697683 , KF709027-KF709034 ) . The PCR product was digested by AluI restriction enzyme , which results in 118-bp and 75-bp fragments when there is a homozygous G119S mutation . A total of 577 and 414 mosquitoes were tested for kdr and ace-1 mutations , respectively . Genotype frequencies were calculated and deviation from Hardy-Weinberg equilibrium ( HWE ) was analyzed using the web-based program ‘GENEPOP’ [40] . Agricultural activity is particularly intense in the two study sites . Insecticides are commonly used for agricultural pest control . An insecticide usage survey was conducted following a standardized questionnaire that included questions on crops harvested and insecticides used , including brand name , time and operation dosage for crop treatment and for human health protection . For each site , questionnaire surveys on 20 households were administered . Soil and water samples were collected from four mosquito sampling sites in Anhui Province to determine residual insecticide concentrations for deltamethrin ( pyrethroid ) and chlorpyrifos ( organophosphate ) in July 2012 . For each sampling site , water samples were collected at four equidistant points in 250-ml aliquots , 5 cm from the water's surface . The aliquots were combined in 1-L amber glass bottles and analyzed in duplicate . Soil samples were extracted as 10-cm plugs at four points in each sampling site . The samples were combined and manually blended until homogeneous . Water and soil samples were chilled ( 4°C for water and −20°C for soil ) until analysis . A negative water and soil samples for controls were collected from an abandoned cornfield that has not been planted or sprayed with insecticides for at least two years . A positive control sample was prepared by adding diluted deltamethrin and chlorpyrifos to the negative water and soil samples at a concentration in ppm . Each water sample was directly extracted in a separatory funnel with methylene chloride . The organic fractions were combined , dried with anhydrous sodium sulfate , and filtered . The solvent was stripped in vacuo before a final dilution in hexane for analysis using a GC-MS equipped with an electron impact ion source . Soil samples were prepared by mixing 5 . 0 g ( dry weight ) sample with anhydrous sodium sulfate . The sample was extracted with a 1∶1 ( v/v ) methylene chloride:acetone solution under ultrasonicating conditions . The organic layers were combined and dried with sodium sulfate , and the solvent was evaporated in vacuo prior to the addition of activated copper to remove sulfur contamination . The solution was filtered and evaporated to dryness . The residue was taken up in 9∶1 hexane:acetone and eluted through a plug of silica conditioned with 9∶1 hexane:acetone . The sample was eluted and dried , and the residue was reconstituted in 9∶1 hexane:acetone for analysis using a GS-MS equipped with an electron-impact ion source in selective-ion monitoring mode . Sample chemical analysis was conducted by the National Center of Agricultural Standardization and Supervision ( Anhui ) under the China National Center for Quality Supervision and Testing of Agricultural-Avocation Processed Food . Mosquito mortality rates after a 24-hr recovery period were calculated for each insecticide . If control mortality was greater than 5% but less than 20% , then the observed mortality was corrected according to the mortality rates of the respective control groups ( control paper ) using Abbott's formula following the WHO test procedures [41] . If the control mortality was below 5% , it was ignored and no correction was necessary . If the control mortality was above 20% , the tests were discarded . We classified mosquito resistance status according to WHO criteria [4]—i . e . , resistant if mortality is <90% , probable resistant if mortality is 90–98% , and susceptible if mortality is >98% . Univariate analysis of variance ( ANOVA ) was conducted using the arcsin transformation of the mosquito mortality rate to determine among-population differences in mosquito mortality rates in the insecticide susceptibility bioassay . One-tailed Mann-Whitney tests were used to compare the enzyme activities in the two field populations and the lab susceptible strains . To determine the role of target site mutation and metabolic detoxification enzymes on phenotypic resistance , we conducted the following three analyses . First , the kdr and ace-1 allele frequency was calculated in each population . The odds ratio ( OR ) of kdr and ace-1 gene mutation on resistance ( survival or death in resistance bioassay ) was calculated , and the statistical significance was determined using the Chi-Square ( χ2 ) test . Second , the mean enzymatic activity was calculated for mosquitoes that survived the bioassay ( resistant ) and those that died in the bioassay ( susceptible ) for each insecticide tested , and the relative enzyme activity ratio of resistant individuals to susceptible individuals was presented . A t-test was used to determine whether this ratio value was significantly different from the null expectation of 1 ( same enzyme activities between resistant and susceptible individuals ) . Third , we used the CART method to determine the relative contributions of target site mutations ( kdr and ace-1 genes ) and metabolic detoxification enzymes ( P450s , GSTs and COEs ) to phenotypic resistance . The CART method is a nonparametric statistical method that recursively partitions the multidimensional space defined by the explanatory factors into subsets as homogeneous as possible [42] . In the CART analysis , the dependent variable was resistant or susceptible status of a mosquito , and factors analyzed were kdr or ace-1 mutation ( binomial variables ) , and P450 , GST and COE enzyme activities ( continuous variables ) . We used the Gini impurity criterion to determine variable splits and identified optimal trees from repeated cross-validations to find the smallest trees whose model errors fell within 1 standard error of the minimum error [43] . The relative importance of each variable is measured by the predictor importance score . A raw variable importance score is constructed by locating every node split by a variable and summing up all the improvement scores generated by the variable at those nodes ( if the variable acted as a surrogate , add up all those improvement scores as well ) . The raw importance score is rescaled so that the best score is always 100 and all other variables are scaled down proportionately [44] . The CART v6 . 0 software ( Salford Systems , Inc ) was used to construct classification and regression trees [44]–[46] . An . sinensis kdr haplotypes . Yunnan: KF697669-KF697673; Anhui: KF697674-KF697683 . An . sinensis ace-1 haplotypes . Yunnan: KF709027-KF709030; Anhui: KF709031-KF709034 .
The standard WHO resistance tube bioassay found that the mortality rate of the mosquitoes against the various control papers was generally <5% , and in 5–20% in 6 ( 12 . 2% ) tests . The corrected mortality rates were all below 90% for the five insecticides tested in both the Anhui and Yunnan sites ( Fig . 2 ) . According to WHO criteria [4] , An . sinensis mosquitoes from the two study sites were resistant to pyrethroids , carbamates , organophosphates , and organochlorines . Further , with the exception of malathion , resistance was extremely prevalent , as more than 50% of mosquitoes survived the diagnostic dose for resistance , and in some cases more than 90% of the tested mosquitoes survived the bioassay ( Fig . 2 ) . In general , the Anhui population was more resistant than the Yunnan population , as the Anhui population exhibited a significantly lower mortality rate than the Yunnan population for three insecticides tested ( permethrin , DDT and malathion ) ( P<0 . 01 ) ( Fig . 2 ) . Two types of non-synonymous kdr mutation at position 1014 ( TTG to TTT and TGT ) were observed in the Anhui population . The mutation ( TTG → TTT ) lead to a change from leucine to phenylalanine ( L1014F ) and the mutation ( TTG → TGT ) leads to a leucine to cysteine substitution ( L1014C ) . The L1014F mutation was predominant ( 70 . 0–88 . 9% ) and the L1014C mutation ( 11 . 1–26 . 7% ) was less common in the Anhui population ( Table 1 ) . Wildtype allele frequency was low ( <9% ) . Significantly higher kdr mutation frequencies ( both L1014F and L1014C alleles ) were found in the deltamethrin-resistant mosquitoes than in the susceptible mosquitoes for the Anhui site ( Table 1 ) . Except for DDT-susceptible individuals , all genotype frequencies at the kdr locus conformed to HWE ( P>0 . 05 ) . Interestingly , no kdr mutation was detected in the Yunnan population despite high levels of phenotypic resistance ( Table 1 ) . No kdr mutation was detected in the lab susceptible strain . To detect the target site mutation in the ace-1 gene ( ace-1R ) , a 193-bp fragment was amplified by PCR . The PCR product was digested by AluI restriction enzyme , resulting in 118-bp and 75-bp fragments when a G119S resistant mutation was present ( Fig 3 ) . Comparison of PCR-RFLP method with direct sequencing on 40 individuals revealed 100% consistency in ace-1 mutation detection . The G119S allele was detected in both the Anhui and Yunnan populations , with 58 . 9% and 38 . 5% frequencies , respectively . No ace-1 mutation was detected in the lab susceptible strain . For the Yunnan population , ace-1 alleles were at HWE , but the Anhui population exhibited a significant heterozygote excess ( P<0 . 05 ) . Significantly higher G119S allele frequency was observed in malathion-resistant mosquitoes than in susceptible mosquitoes in both the Anhui and Yunnan populations ( P<0 . 001 ) , but such a phenomenon was not observed in the bendiocarb-resistant or susceptible populations for both sites ( Table 2 ) . The median P450 activity of the lab strain was 25 . 3 pmol 7-HC/min/mg protein ( ranging from 19 . 6 to 34 . 5 ) , the median GST activity was 0 . 229 µmol cDNB/min/mg protein ( ranging from 0 . 06 to 0 . 44 ) , and the mean COE activity was 0 . 104 µmol p-nitrophenol formed/min/mg protein ( ranging from 0 . 04 to 0 . 21 ) . For field-collected mosquitoes , significantly elevated levels of cytochrome P450s , GSTs and COEs were found in the Anhui and Yunnan populations compared with the susceptible lab strain ( Fig 4 ) . Overall , F-tests found significantly higher variances in field-collected mosquitoes than in the laboratory strain for both the Anhui and Yunnan populations ( F<0 . 01 for all tests; Fig . 4 ) , suggesting much higher within-population variability in the metabolic detoxification enzyme activities . Comparison of metabolic enzyme activities between the mosquitoes that survived the bioassay ( resistant ) and those that died in the bioassay ( susceptible ) would reveal the role of the metabolic enzyme activities in resistance . We found that deltamethrin- permethrin- and malathion-resistant mosquitoes consistently showed significantly higher P450 activities than susceptible mosquitoes ( Fig 5A ) . Further , malathion-resistant mosquitoes showed significantly higher GST activities ( Fig . 5B ) and COE activities ( Fig . 5C ) than susceptible mosquitoes , suggesting all three metabolic detoxification enzymes may play a role in malathion resistance . One unsolved key question in insecticide resistance research is the determination of the relative contributions of target site insensitivity and various metabolic detoxification enzymes to resistance . To answer this question , we conducted a CART analysis that simultaneously took kdr and ace-1 genotypes and three metabolic enzyme activities into consideration . Because resistance to malathion and bendiocarb involves the same target site ( ace-1 ) and resistance to deltamethrin , permethrin and DDT involves a different target site ( kdr ) , we analyzed resistance to malathion and bendiocarb separately from the other three insecticides . The CART analysis revealed that for resistance to deltamethrin , permethrin , and DDT , the most important variable was P450 activity , followed by GST or COE activity , and kdr mutation played a small role ( Fig 6; Table S1 ) . For resistance to malathion , the most important variable was COE activity , followed by GST and P450 activity , and G119S mutation was less important . These were consistent for the Yunnan and Anhui populations ( Fig . 6 ) . The role of COEs in resistance to bendiocarb varied between the Yunnan and Anhui populations , and the important role of GSTs was consistent . Overall , these results suggest metabolic resistance was the most important resistance mechanism for the four classes of insecticides tested . Overall , rice is the major crop in the two study sites . Other crops include wheat , corn , sugarcane , banana , and vegetables . Pyrethroids , organochlorine , organophosphates and carbamates were all used in agricultural pest control ( Table S2 ) . The organophosphate-based insecticides were more commonly used than pyrethroids . The common insecticides used for agricultural pest control included lambda-cyhalothrin , beta-cypermethrin , dichlorvos , chlorpyrifos , malathion and propoxur . The insecticides used for indoor residual spraying and bed net treatment included dimefluthrin , alpha-cypermethrin , meperfluthrin and D-prallethrin . Although the insecticide usage surveys were conducted in limited number of houses ( n = 20 per site ) , the survey results showed diverse types of insecticides being used in the study sites . Residual insecticide analysis in samples from the Anhui study site detected chlorpyrifos in soil samples with concentration ranging from 21–130 ppb , but no chlorpyrifos was detected in the water samples ( Table S3 ) . Deltamethrin was not detected in either the soil or water samples . The analytes of interest ( deltamethrin and chlorpyrifos ) were all detected in the positive control water and soil samples , confirming the soundness of the analytic technique .
The study demonstrated that field populations of An . sinensis from central China ( Anhui ) and southern China ( Yunnan ) developed high resistance to four classes of insecticides tested , including pyrethroid ( deltamethrin and permethrin ) , organochlorine ( DDT ) , organophosphate ( malathion ) and carbamate ( bendiocarb ) . The An . sinensis population from Anhui was more resistant to permethrin , DDT , and malathion than the population from Yunnan , while resistance levels to deltamethrin and bendiocarb were similar . Interestingly , among the 300 samples from Yunnan tested for kdr mutation , no kdr mutation was detected . Therefore , the target site kdr mutation at L1014 is not the resistance mechanism to pyrethroids and DDT in the Yunnan population . On the other hand , kdr mutation reached near fixation for the Anhui population , and our CART analysis suggests that kdr mutation only played a small role in pyrethroids and DDT resistance . Similarly , target site mutation at the ace-1 gene makes a small contribution to resistance to organophosphate and carbamate . This is consistent for both the Anhui and Yunnan populations . Therefore , we conclude that under long-term high insecticide selection pressure , mosquitoes have evolved a strong metabolic resistance to various classes of insecticides . Because the same bioassayed mosquitoes in our study were tested for kdr and ace-1 mutations and three detoxification enzyme activities , using the CART statistical method , we were able to statistically determine the importance of these individual factors in resistance to multiple insecticides . This analysis found that kdr and ace-1 mutations played a small role in resistance to the four classes of insecticides tested , P450 activity was the most important mechanism in mosquito resistance to pyrethroids and organochlorines and COE monooxygenases was most important to organophosphate and carbamate resistance . When a mosquito population lacked kdr mutation ( as in the Yunnan population ) , GSTs played a large role in pyrethroid resistance in comparison to the Anhui population , in which kdr mutation was very prevalent . Previous investigations on metabolic activity in African An . gambiae confirmed that monooxygenases played an important role in the resistance to pyrethroids while COE was the major metabolic mechanism for organophosphates [47]–[49] . This correlates with our findings . Several studies have found correlation between L1014F allele frequency and resistance to pyrethroids in An . gambiae populations from West Africa [48] , [50]–[53] . Thus , it is important to experimentally verify the results from the CART analysis that only statistically teased out the specific contribution of each resistance mechanism . For example , the metabolic enzymes may be inhibited using various synergists prior to insecticide resistance bioassay , such as synergists PBO ( 4% pyperonyl butoxide ) , DEF ( 0 . 25% S . S . S-tributyl phosphotritioate ) , DEM ( 8% diethyl maleate ) and TPP ( 10% Triphenyl phosphate ) which are known inhibitors of multi-function oxidases ( MFOs ) , glutathione S-transferases ( GSTs ) , non-specific esterases ( NSEs ) and carboxylesterase ( COEs ) , respectively [54]–[58] . Through comparison of mosquitoes pre-exposed to synergists with the appropriate controls , the effects of specific metabolic enzymes on resistance can be determined . The lack of kdr mutation in An . sinensis population in Yunnan Province was not unique to our particular study site ( Lianghe and Yingjiang counties ) . We have previously examined An . sinensis from Mengla and Yuanyang counties of Yunnan Province and did not detect any kdr mutation in 186 samples [23] , despite the fact that these mosquito populations have been , and are currently , experiencing strong insecticide selection pressures . Low kdr mutation frequency ( <30% ) was reported in an An . sinensis population from another southern China province , Guangxi [59] . The high kdr mutation frequency observed in the central China Anhui study site was consistent with other studies on An . sinensis throughout central China . For example , we detected >90% kdr mutation frequency in Hunan , Hubei and Jiangsu Provinces , China [23] . Tan et al [60] reported a 95–100% kdr mutation frequency in an An . sinensis population from Jiangsu Province in central China . In addition to the predominant L1014F kdr allele , we found L1014C allele with considerable frequency . Further , we detected a significant positive association between L1014C mutation and resistance to deltamethrin , but not to DDT and permethrin resistance . The role of L1014C mutation on insecticide resistance should be further investigated by increasing the number of sample sites . It is not clear what caused the lack of kdr mutation in the Yunnan population . One interesting possibility is that the lack of gene flow between central China and the mountainous Yunnan Province prevents kdr mutation from being spread to the population in Yunnan . We are currently examining the An . sinensis population genetic structure in China to determine the role of gene flow on the spread of kdr mutation . The high level and multiple insecticide resistance of An . sinensis in China may result from prolonged and extensive use of insecticide for agricultural pest control and public health disease vector control . The detected insecticide residues of organophosphates ( chlorpyrifos ) in soil in Anhui site further suggest that insecticide residues in the larval environment of mosquitoes through agricultural pest control spray may be an important factor that selected for insecticide resistance . It is possible selection pressure from larval environment may be very strong as mosquito larvae are confined to the aquatic habitats with residual insecticides and constantly exposed to insecticides in the aquatic habitats . Adult mosquitoes are mobile and may exhibit behavioral avoidance to the insecticide [61] , [62] . Whether selection pressure from larval exposure to insecticides favors metabolic resistance more than mutational target site resistance is an interesting question for future research . Resistance to multiple classes of insecticides is becoming a common problem in various disease vector species . Reported multiple resistance in mosquito vectors includes An . gambiae [47]-[49] , [63] , An . arabiensis [64] , An . funestus [65] , Culex quinquefasciatus [47] , Aedes aegypti and Ae . albopictus [66] in Africa , and An . culicifacies , An . subpictus , An . nigerrimus , An . peditaeniatus [56] and Cx . quinquefasciatus [67] , [68] in Asia , and Ae . Aegypti in South America [69] . Multiple insecticide resistance impedes the current front-line vector-borne disease control programs , which are primarily based on the use of pyrethroids . The present study suggests that long-term use of various classes of insecticides , in rotation or combination , will eventually select vector populations resistant to the major classes of insecticides . Our results have important implications for Anopheles and other mosquito vector control strategies . Organophosphate and carbamate insecticides may have limited applications in disease vector control , and rotation or combinational use with these insecticides may not be effective as the mosquito populations are already highly resistant . Adding appropriate synergists to the IRS formulation may help improve the effectiveness of the insecticides . Developing and implementing alternative efficient vector control methods that are not reliant on pyrethroids , organophosphate and carbamates—such as home improvement [70] , odor-baited traps [71] , larval resource reduction [72] and biological control [73] , microbial insecticides and new classes of insecticides — presents an urgent challenge . | Malaria and lymphatic filariasis are two of the most important mosquito-borne parasitic diseases worldwide , which can occur as concomitant human infections . The Anopheles sinensis mosquito is the most common malaria and lymphatic filariasis vector in Southeast Asia . Disease-vector control is an important part of the global malaria and filariasis control strategies . Pyrethroid insecticides are the major vector control agents , and rotational or combinational use with other classes of insecticides has been proposed to mitigate the problem of pyrethroid resistance . Therefore , assessing resistance to multiple classes of insecticides is important because vector control programs may not be effective , or even fail , if mosquito vectors are already resistant to the insecticides being used in rotation or combination . The field populations of An . sinensis mosquito from central China ( Anhui ) and southern China ( Yunnan ) developed high resistance to four classes of insecticides tested , including pyrethroids ( deltamethrin and permethrin ) , organochlorine ( DDT ) , organophosphate ( malathion ) , and carbamate ( bendiocarb ) . This study examined two questions: 1 ) the consequences of long-term insecticide rotation use on insecticide resistance , and 2 ) the role of metabolic detoxification and target site insensitivity in resistance to chemical insecticides . The information obtained from this investigation is valuable in informing current and future vector control strategies . | [
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] | [
"arthropoda",
"public",
"and",
"occupational",
"health",
"invertebrates",
"disease",
"vectors",
"mosquitoes",
"medicine",
"and",
"health",
"sciences",
"infectious",
"diseases",
"helminth",
"infections",
"filariasis",
"epidemiology",
"global",
"health",
"biology",
"and",
"life",
"sciences",
"lymphatic",
"filariasis",
"malaria",
"parasitic",
"diseases",
"animals",
"insects",
"organisms"
] | 2014 | Multiple Resistances and Complex Mechanisms of Anopheles sinensis Mosquito: A Major Obstacle to Mosquito-Borne Diseases Control and Elimination in China |
During animal development , cellular morphogenesis plays a fundamental role in determining the shape and function of tissues and organs . Identifying the components that regulate and drive morphogenesis is thus a major goal of developmental biology . The four-celled tip of the Caenorhabditis elegans male tail is a simple but powerful model for studying the mechanism of morphogenesis and its spatiotemporal regulation . Here , through a genome-wide post-embryonic RNAi-feeding screen , we identified 212 components that regulate or participate in male tail tip morphogenesis . We constructed a working hypothesis for a gene regulatory network of tail tip morphogenesis . We found regulatory roles for the posterior Hox genes nob-1 and php-3 , the TGF-β pathway , nuclear hormone receptors ( e . g . nhr-25 ) , the heterochronic gene blmp-1 , and the GATA transcription factors egl-18 and elt-6 . The majority of the pathways converge at dmd-3 and mab-3 . In addition , nhr-25 and dmd-3/mab-3 regulate each others' expression , thus placing these three genes at the center of a complex regulatory network . We also show that dmd-3 and mab-3 negatively regulate other signaling pathways and affect downstream cellular processes such as vesicular trafficking ( e . g . arl-1 , rme-8 ) and rearrangement of the cytoskeleton ( e . g . cdc-42 , nmy-1 , and nmy-2 ) . Based on these data , we suggest that male tail tip morphogenesis is governed by a gene regulatory network with a bow-tie architecture .
Morphogenesis involves the coordinated change in the shape of cells and tissues during development , eventually giving rise to functional structures in the adult animal . Such coordinated change must occur at the correct time and in the proper position . In the case of structures that differ between the sexes , this process must also be regulated sex-specifically . While many genes and pathways are known that regulate development , the identity of genes that link regulation to the execution of morphogenesis have been more difficult to ascertain [1] . The many different cues and signals that must be integrated to control morphogenesis , combined with the complexity of the molecular machinery associated with this process , suggest that a large number of genes and gene products are involved . To elucidate the molecular mechanisms underlying morphogenesis , the first step is thus to determine what components are involved in its regulation and execution and to determine how they interact in a network . In the pursuit of such an aim , it is advantageous to use a simple model structure that still demonstrates all the properties of cellular morphogenesis . The model we use is the male tail tip of Caenorhabditis elegans . This structure is made up of four epithelial ( "hypodermal" ) cells , hyp8–hyp11 , which are born during embryogenesis . Embryonic morphogenesis of hyp8–hyp11 leads to the formation of a pointed , whip-like tail tip . The tail tip retains this shape throughout the lifespan of the hermaphrodite . However , during the last larval stage ( L4 ) of males , these conical cells are dramatically remodeled to form the rounded tip of the adult [2]–[4] . Male tail tip morphogenesis begins when hyp8–11 fuse to form a syncytium; fusion is followed by detachment of the cells from the overlying cuticle . Towards the middle of the L4 stage , the syncytium changes its shape from conical to round and moves anteriorly; these morphogenetic events cease at the end of the L4 stage ( Figure 1A ) . In hermaphrodites , the tail tip cells do not fuse and do not change shape . The tail tip model thus allows the study of a sexual dimorphism at the cellular level . Another advantage of this model is that male-specific mutations in C . elegans can be propagated through the self-fertile hermaphrodites , even if the mutations affect male fertility , mating ability , or viability . A few mutations in genes involved in the regulation of tail tip morphogenesis have been described previously [5]–[7] . Some of these mutations impede or completely block tail tip morphogenesis , resulting in the retention of the pointed larval tail tip in the adult male , a phenotype that we call "Lep" ( Figure 1C ) . This phenotypic designation is derived from the term "leptoderan , " which—in the taxonomic literature—describes the unretracted , pointed male tail tip in some nematodes related to C . elegans [8] , [9] . Other mutations cause precocious onset of ( and thus an extended total period for ) tail tip retraction , which results in over-retracted ( "Ore" ) and thus abnormally shortened adult male tails [6] ( Figure 1C ) . Studies of these mutants have revealed a few of the important regulatory components for tail tip morphogenesis . DMD-3 was suggested to be a central regulator of tail tip morphogenesis , as it is required for tail tip retraction in males and is sufficient for inducing ectopic morphogenesis in hermaphrodite tail tips [7] . The gene encoding this DM-domain transcription factor , dmd-3 , is a homolog of dmrt1 in vertebrates and doublesex in Drosophila [10] , [11] . DMD-3 functions cooperatively and partially redundantly with a closely related factor , MAB-3 [7] , also involved in somatic sex determination [12] . TRA-1 , the most downstream global regulator in the sex-determination pathway , inhibits the expression of these genes in hermaphrodites [7] , [13] , [14] . The initiation of dmd-3 expression at the proper developmental stage is controlled by the "heterochronic" pathway [7] . Finally , maintenance of dmd-3 expression levels is regulated by Wnt signaling and a feedback loop involving both MAB-3 and DMD-3 [7] . Only one effector of a cellular process is known to be downstream of DMD-3 and MAB-3 , namely the fusogen-encoding gene eff-1 , which is important for the many fusions that occur between epithelial cells in C . elegans [15] , [16] . DMD-3 and MAB-3 induce expression of eff-1 through an unknown post-transcriptional mechanism [7] . To find additional genes with roles in tail tip morphogenesis , we carried out a genome-wide , post-embryonic RNAi-feeding screen . This screen identified 212 candidates . Starting with these candidates , we used network model-building [17] , transgenic reporter lines and expression epistasis analysis to construct a first draft of the gene regulatory network for tail tip morphogenesis . We found that dmd-3 expression is regulated by a nuclear hormone receptor ( NHR-25 ) , a new heterochronic gene , Hox anteroposterior patterning factors , and GATA transcription factors . NHR-25 is in turn negatively regulated by dmd-3 and mab-3 . In addition , dmd-3 and mab-3 negatively regulate other signaling modules , including the TGF-β pathway . We also found that DMD-3 and MAB-3 regulate the localization or expression of genes involved in vesicular trafficking/endocytosis , cell polarity and cytoskeletal organization . Our data thus strongly support the hypothesis that DMD-3 , MAB-3 and NHR-25 are central regulators of tail tip morphogenesis . The genetic architecture for tail tip morphogenesis which emerges from this analysis closely resembles the "bow-tie" architecture , a possibly universal characteristic of robust , evolvable systems [18] . A bow-tie network has many inputs and outputs that are connected through a conserved core . Versatile weak linkages form the interface between the core and the input and output . In such a network , there is not a simple flow of information from input to output through the core; instead , there is extensive global and local feedback regulation found at every level [19] . A bow-tie architecture has been found to underlie a variety of biological networks: metabolic networks [20] , [21] , the Toll-like receptor signaling network [22] , the epidermal growth factor receptor signaling network [23] , the osmolarity glycerol signal-transduction pathway in yeast [24] , stress response networks [25] and the immune system [26] , [27] . In fact , it has been proposed that the bow-tie architecture of regulatory networks is ubiquitous because this structure ensures not only robustness but also evolvability [18] , [19] . We found evidence for the existence of each aspect of a bow-tie architecture in the gene network governing tail tip morphogenesis . To our knowledge , this is the first time that this kind of network architecture has been explicitly identified in the context of development and morphogenesis . However , we believe that other developing systems are also governed by bow-tie genetic networks , supporting the proposition that this network architecture is universal .
To identify tail tip morphogenesis genes , we cultured animals on dsRNA-expressing bacteria from L1 to adult and looked for evidence of defective morphogenesis . Using the Ahringer RNAi-feeding library [28] , we screened 16 , 131 genes , approximately 83% of the genome . Genes that gave a positive Lep or Ore phenotype were screened again; only repeatable positives were kept as candidates . We identified 212 genes , of which 190 produced Lep phenotypes , 14 produced Ore phenotypes , and 8 produced both Lep and Ore phenotypes in a single experiment ( Figure 1D ) . Positives in each category were analyzed for GO-attribute enrichment using FuncAssociate [29] . Relative to the genome , Lep positives showed enrichment of GO-attributes associated with components or processes involved in morphogenesis , such as anchoring junctions and cell migration . However , Ore and Ore/Lep positives were enriched for genes involved in cell division/cytokinesis , and nuclear/chromosome organization ( Figure 1D ) . The 212 candidates are listed and categorized by developmental pathway or annotated cellular process ( if known ) in Table S1 . Raw RNAi data are publicly available along with representative images via the "Male Tail Tip Database" ( MTTdb , at http://wormtails . bio . nyu . edu ) . We identified components of conserved and widely studied developmental regulatory pathways ( e . g . Hox anteroposterior patterning factors , TGF-β signaling module , heterochronic pathway , and GATA transcription factors ) . Importantly , the screen also identified components of fundamental cellular structures and processes that are likely to be important for the execution of morphogenesis . These include vesicular trafficking/endocytosis , cellular polarity , cytoskeleton , cell junctions , and nuclear export/import . Genes with the most severe and/or most penetrant RNAi phenotypes were selected for further study ( Table 1 ) . Hox genes play central roles in shaping animal body plans by distinguishing different fields of cells along the anteroposterior axis [30] . Hox proteins regulate not only the expression of other transcription factors , but also of genes involved in specific processes such as morphogenesis [31] . Our screen has shown that posterior Hox patterning is crucial for tail tip morphogenesis . RNAi knockdowns of the Abd-B homologs php-3 and nob-1 postembryonically cause Lep phenotypes ( Figure S1A , Tables S1 and S2 ) . Likewise , a php-3 null allele , ok919 , results in a 100%-penetrant Lep phenotype of moderate severity ( Figure S1B , Table S2 ) . The severity of the php-3 ( ok919 ) phenotype is increased with nob-1 RNAi treatment ( Table S2 ) . A nob-1::gfp translational fusion construct ( containing a genomic fragment with the nob-1 gene and 9 Kb of the 5′-regulatory region ) is expressed in the tail tip cells hyp8–11 throughout larval development in both sexes ( Figure 2A and data not shown ) . A php-3::gfp translational fusion driven only by the short 500-bp intergenic region between nob-1 and php-3 shows variable expression in the nuclei of hyp8–11 during tail tip morphogenesis ( data not shown ) . This latter transgene only partially rescues the tail tip phenotypes of php-3 mutants ( Table S2 ) , suggesting that regulatory elements upstream of both php-3 and nob-1 are required for appropriate expression of php-3 . Using a tail tip-specific promoter ( from the gene lin-44 [32] ) , we observed that expression of the PHP-3::GFP fusion protein product in hyp8–11 is sufficient to rescue the php-3 ( ok919 ) Lep phenotype ( Figure S2A and Table S2 ) . Mosaic animals that only express PHP-3::GFP in a subset of tail tip cells do not show rescue ( Figure S2B ) . These data suggest that Hox-mediated patterning by PHP-3 and NOB-1 is carried out cell-autonomously and is required in all tail tip cells ( hyp8–11 ) for proper morphogenesis . GATA transcription factors play important regulatory roles during the differentiation of multiple cell types in normal development [33] , [34] and during tumorigenesis [35] . RNAi knockdown of the gene for the GATA factor egl-18 resulted in Lep phenotypes ( Figure S1C , Tables S1 and S2 ) . An egl-18 null allele , ok290 , causes Lep phenotypes of varying severity with 45% penetrance ( Figure S1D , Table S2 ) . Another GATA transcription factor that was missed in our RNAi screen , elt-6 , shares an operon with egl-18 [36] . We repeated the RNAi treatment against elt-6 and observed low-penetrance , low-severity tail tip defects ( Table S2 ) . RNAi treatment for elt-6 in the egl-18 ( ok290 ) mutant strain , however , dramatically increased the penetrance and severity of the tail tip defects , such that tail tip morphogenesis failed altogether in some individuals ( Table S2 ) . Interestingly , an egl-18::gfp translational fusion is expressed in the nuclei of the main body epidermal syncytium hyp7 , but appears to be excluded from the tail tip cells hyp8–11 ( Figure 2B ) . This observation suggests that EGL-18 and ELT-6 function in cells adjacent to the tail tip to regulate tail tip morphogenesis cell-nonautonomously . Furthermore , transforming wild-type animals with the egl-18/elt-6 operon regulated by the lin-44 promoter disrupted morphogenesis ( data not shown ) , suggesting that morphogenesis requires exclusion of these GATA factors from hyp8–11 . In other epidermal cells , it has been observed that EGL-18 and ELT-6 repress cell fusion [37] . Thus , their exclusion from the tail tip cells may be required to allow tail tip cell fusion and subsequent morphogenesis . TGF-β signaling is a major conserved cell-signaling module which regulates multiple processes during the development of all animals and also during the progression of cancer [38] . The screen identified two components of this pathway . RNAi treatment against sma-3 , which encodes a Smad protein , and daf-4 , encoding the TGF-β receptor , resulted in Lep phenotypes ( Figure S1E , Tables S1 and S2 ) . A sma-3 null allele , e491 , resulted in a 57%-penetrant low-severity Lep phenotype ( Figure S1F , Table S2 ) . Of the other known components of the TGF-β pathway , only sma-2 showed any RNAi phenotype: RNAi against sma-2 significantly enhanced the penetrance of the Lep phenotype of sma-3 ( e491 ) ( Table S2 ) . A transgenic strain expressing a sma-3::mCherry translational reporter shows expression at low levels in the cytoplasm of the tail tip of both sexes at the beginning of the L4 stage . In males , the SMA-3::mCherry fusion protein enters the nuclei of hyp8–11 prior to tail tip morphogenesis and remains in both the cytoplasm and nuclei during morphogenesis ( Figure 2C ) . In hermaphrodites , SMA-3::mCherry remains cytoplasmic throughout L4 and never enters the nuclei ( data not shown ) . The dynamic localization pattern of SMA-3::mCherry suggests that TGF-β-mediated gene expression occurs concurrently with tail tip morphogenesis . Consistent with this hypothesis , DAF-4::YFP fusion protein localizes to the plasma membranes of hyp8–11 during tail tip morphogenesis ( Figure 2D ) . Taken together , the RNAi results , loss-of-function mutant phenotypes and expression patterns of sma-3 and daf-4 suggest that TGF-β signaling is required during tail tip morphogenesis . Previous work has shown that Wnt signaling plays a crucial role in the regulation of tail tip morphogenesis [5] . Consistent with those findings , the RNAi screen identified additional genes that are in or interact with the Wnt pathway . RNAi treatments against sys-1 ( β-catenin ) and lit-1 ( Nemo-like kinase ) each resulted in Lep phenotypes ( Table S1 , Figure S3C , S3D ) . A transgenic strain expressing a SYS-1::GFP fusion protein shows faint cytoplasmic expression in hyp8 and hyp11 but not in hyp9 or hyp10 ( Figure S3C ) . A strain expressing a LIT-1::GFP fusion protein shows nuclear expression in hyp9 and hyp10 but not in hyp8 or hyp11 prior to morphogenesis ( Figure S3D ) . We identified multiple nuclear hormone receptor genes in our screen: nhr-9 , nhr-23 , nhr-25 and nhr-165 . NHR-25 is the C . elegans homolog of FTZ-F1 [39] which is a highly conserved protein with diverse functions regulating embryonic patterning [40] , [41] , and ecdysone-mediated molting in Drosophila [42] . Both nhr-25 RNAi and a hypomorphic allele , ku217 , showed Lep phenotypes ( Figure S1G and S1H , Tables S1 and S2 ) . Furthermore , nhr-25 RNAi treatments on the ku217 strain showed a complete lack of male tail morphogenesis in 33% of the animals ( N = 24 , Table S2 ) , a phenotype reminiscent of the mab-3 ( e1240 ) ;dmd-3 ( ok1327 ) double mutant [7] . We were not able to produce transgenic lines expressing NHR-25 fusion proteins due to lethality , as previously reported [43] . Instead , we made a transgenic strain expressing a transcriptional reporter containing the 5′- and 3′-regulatory regions for nhr-25 . This reporter shows a dynamic expression pattern in which expression in hyp8–11 is intense prior to and at the beginning of morphogenesis , rapidly shuts off in late L4 , and is never on in adult animals ( N = 34 adults ) ( Figure 2E ) . RNAi against another nuclear hormone receptor gene , nhr-165 , showed a low-penetrant , mild , yet reproducible Lep phenotype upon RNAi treatment ( Table S1 , and Figure S3A ) . An nhr-165::gfp translational reporter shows nuclear expression in the lateral hypodermis ( hyp7 , but not hyp8–11 ) near the time of larval molts . This expression is highest and seen in the largest number of cells in the tail prior to the L3–L4 molt ( Figure S3A ) . The heterochronic pathway ensures that tail tip morphogenesis is initiated precisely at the beginning of the L4 stage [6] . We identified genes in this pathway: the zinc-finger transcription factor gene blmp-1 , and known suppressors of the miRNA let-7 [44] . BLIMP-1 is a transcriptional repressor that regulates ecdysone-mediated molts in Drosophila [45] and differentiation of multiple cell types in humans and mice , such as lymphocytes [46] and primordial germ cells [47] , [48] . Intriguingly , BLIMP-1 is a target of let-7-mediated degradation in Reed-Sternberg cells , a Hodgkin-lymphoma cell line , suggesting a possibly conserved interaction with the heterochronic pathway , of which let-7 is a member [49] , [50] . RNAi treatments directed against blmp-1 produced Ore phenotypes ( Table S1 ) . A deletion allele , tm548 , produces an Ore phenotype with 100% penetrance ( Table S2 ) due to precocious initiation of tail tip morphogenesis during the L3 stage ( N = 26 , data not shown ) . A transgenic line expressing a BLMP-1::GFP fusion protein shows both nuclear and cytoplasmic expression in hyp8–11 throughout development , although cytoplasmic expression is most intense during tail tip retraction ( Figure 2F ) . Of the 41 suppressors of let-7 lethality identified by Ding et al . [44] , six were positives in our screen . RNAi knockdown of two ( pri-2 , npp-6 ) resulted in the Ore phenotype , of two others ( spg-7 , smo-1 ) in the Lep phenotype and of two further genes ( cdt-1 , xpo-2 ) in both phenotypes in a single experiment . Four of these genes—pri-2 , npp-6 , cdt-1 , and xpo-2—are predicted by N-Browse [17] to interact in a subnetwork that includes other genes for which RNAi knockdown also produced Ore phenotypes ( Figure 3 ) . It is thus possible that these additional genes also influence the timing of tail tip morphogenesis . It is still unclear why both Ore and Lep phenotypes are observed in a single experiment . One possible explanation is that precision in timing of tail tip morphogenesis is lost when certain gene-products are removed . In this case , morphogenesis might begin too early in one animal and too late in another , leading to Ore tails and Lep tails , respectively . Post-transcriptional regulation appears to play an important role during the coordination of tail tip morphogenesis , as our screen has identified multiple genes that encode RNA splicing factors , kinases , phosphatases and ubiquitinating enzymes . One such gene is ubc-12 , which is a part of the NED-8 conjugating system and has been shown to be important in epidermal differentiation during embryogenesis in C . elegans [51] . RNAi-knockdown of ubc-12 resulted in larval lethality in the RNAi hyper-sensitive rrf-3 ( - ) background ( Table S4 ) . However , we identified ubc-12 in a pilot screen which was carried out in the wild-type background; ubc-12 RNAi treatments produced a highly penetrant and severe tail tip defect ( Figure S3B ) . UBC-12::GFP expresses intensely in the cytoplasm of hyp10 just prior to and during tail tip retraction ( Figure S3B ) . We identified many genes known to play central roles during the execution of morphogenesis . Such genes encode proteins involved in vesicular trafficking , endocytosis , cell-cell communication , cytoskeletal rearrangement , establishment of cellular polarity and cell-cell transport . Adjacent to the tail tip cells lie the dendritic projections of the PHC neurons . Two genes that produced Lep phenotypes upon RNAi treatment are expressed in these neurons but not in the tail tip cells: the calcipressin-encoding gene rcn-1 [54] and ptl-1 , a gene encoding a tau-like microtubule-associated protein [55] , [56] . A null allele of ptl-1 , ok621 , produced Lep phenotypes with 23% penetrance ( Table S2 ) . PTL-1::GFP and RCN-1::GFP fusion proteins are expressed in the PHC neurons prior to and during hypodermal morphogenesis . In adults , PTL-1::GFP is expressed in most neurons of the tail ( data not shown ) . RCN-1::GFP is expressed in most tail neurons and in the support cells of the phasmid neurons ( socket cells , Figure 4H ) . This pattern is consistent with the previously described adult expression pattern of RCN-1 [54] . With a list of genes required for tail tip morphogenesis , we next sought to characterize the interactions between these genes . We constructed a working hypothesis for these interactions using N-Browse , a publicly available database which integrates the information from numerous genome-wide studies to build gene networks [17] . We manually entered our candidate genes into N-Browse , excluding those that did not have predicted functions or known interactors . N-Browse produced a genetic network that included not only our candidate tail tip morphogenesis genes , but additional genes predicted to be nearest-neighbor interactors . To the resulting N-Browse network , we manually added gene interactions ( edges ) based on published work not represented in N-Browse [6] , [7] , [44] , [51] , [57]–[61] ( Figure 5 ) . This analysis predicts the involvement of genes not identified in our screen . We tested two of these predictions by repeating RNAi knockdown with different methods and/or by scoring larger numbers of males . We could thus validate roles in morphogenesis for elt-6 and vav-1 . elt-6 has genetic interactions with egl-18 ( positive in our screen ) in other contexts . It showed a low-penetrance Lep phenotype when the RNAi experiment was repeated and more males were scored . In addition , elt-6 ( RNAi ) enhances the Lep phenotype of egl-18 ( ok290 ) mutants ( Table S2 ) . Also , the network model predicts that vav-1 has interactions with php-3 , cdc-42 and inx-12 ( all positives in our screen , Figure 5 ) . Although vav-1 treatment by RNAi via feeding did not result in detectable phenotypes , administering RNAi against vav-1 by soaking did cause tail tip defects ( data not shown ) . We next asked whether the network model—developed from information in other systems—has biological relevance for tail tip morphogenesis . To this end , we tested a selection of the predicted interactions by genetic and expression epistasis analyses . The results are detailed below .
Here , we used systemic RNAi to identify components that are involved in male-specific morphogenesis of the tail tip of C . elegans . RNAi via feeding in C . elegans provides a simple yet powerful means for identifying the regulatory and structural components and pathways of developmental processes [28] , [62] . The methodology we employed ( Materials and Methods and Figure 1B ) allowed us to quickly score for subtle defects in morphogenesis at high magnification . Many of the genes we identified have roles in embryonic processes and are lethal when knocked down ( e . g . nob-1 , cdc-42 ) . This justifies our approach to treat worms with RNAi postembryonically and it underscores the power of RNAi as a tool for identifying postembryonic roles of genes that have essential embryonic functions . To minimize the number of false negatives , we performed the screen on the RNAi hypersensitive strain , rrf-3 [63] . For a number of reasons , however , we believe that there are still other tail tip genes to be identified . First , our screen did not cover the entire genome ( approx . 83% ) . Second , because of complete larval lethality , about 2% of the genes in our screen were not scored , including the known tail tip regulator lin-41 [6] ( Table S4 ) . Third , of the previously known tail tip genes , only lin-44 , which encodes for the Wnt ligand , was found in the screen . Other known tail tip genes , tlp-1 and dmd-3 , which have representative clones in the library , were missed . Finally , two genes ( elt-6 and vav-1 ) not found in the screen , but predicted by the N-Browse network analysis , turned out to have RNAi-induced tail tip phenotypes when treated in a different genetic background ( i . e . elt-6 RNAi in the egl-18 ( ok290 ) background ) or by soaking instead of feeding ( vav-1 ) . The number of false positives is likely to be very small since only genes which were positive in the primary and secondary screen were considered . The 212 candidate genes identified in this process were significantly enriched with morphogenesis-related GO attributes relative to the genome at large , consistent with what would be predicted if our screen were successful . Some candidates were studied further to elucidate their roles in regulating or effecting tail tip morphogenesis . Developmental decisions , such as the initiation of morphogenesis , require the input of multiple signaling pathways and result in a coordinated response by many different components of the cell . The response must be robust against perturbations from the internal and external environment . Robustness and precision of biological processes are thought to be facilitated by a bow-tie ( or hourglass ) architecture of the gene regulatory network [18] , [19] . Characteristics of bow-tie networks include the following . ( 1 ) Many inputs and outputs are connected to a conserved core . ( 2 ) Versatile weak linkages form the interface between input and core and between core and output . ( 3 ) Systems control is facilitated by positive and negative feedback at every level . ( 4 ) Modularity and partial redundancy or degeneracy are two other properties of the bow-tie network architecture that contribute to the robustness of biological systems [19] , [64] . We used the data about gene interactions described here in combination with published information to delineate a first draft of the gene regulatory network underlying tail tip morphogenesis in C . elegans males . Although the reconstruction of this network has only just begun , we already find many features that are consistent with bow-tie architecture . A network of interactions is called modular if it can be subdivided into relatively autonomous components ( modules ) that are built of highly connected parts but are more loosely connected to other modules [65] . Modularity is a major contributor to the robustness and evolvability of a system , since perturbations and mutations can occur within a module with minimal effects on the whole system [19] . It has been proposed that modularity facilitates evolutionary change by allowing new connections to be made between modules without disrupting the core function of the modules [66] . Modularity has been observed in many networks [67] , [68] . The modules of metabolic networks have bow-tie structure , just like the networks themselves [21]; that is , bow-tie architecture can be nested . In the gene regulatory network of male tail tip morphogenesis , we find evidence for many conserved regulatory and effector modules . Regulatory modules include Hox patterning , the sex-determination and heterochronic pathways and TGF-β signaling . We also identified tail tip roles for additional Wnt pathway components , i . e . the SYS-1 beta-catenin , the MIG-1 Frizzled-like receptor , and the LIT-1 Nemo-like MapK . Effector modules consist of conserved components controlling vesicular trafficking and endocytosis , establishment of cellular polarity , cytoskeletal rearrangement , and cellular fusion . Degeneracy describes the coexistence of structurally or mechanistically different components that can perform similar roles or are interchangeable under certain conditions [64] . Degeneracy confers robustness because , in a system composed of partially redundant elements , one element can compensate for the failure of another . One mechanism that generates degeneracy is gene duplication . In the male tail tip morphogenesis system , we find several examples for degeneracy due to gene paralogy . We think that MAB-3 and DMD-3 function partially redundantly because the phenotype of the mab-3 ( e1240 ) ;dmd-3 ( ok1327 ) double mutant is more severe than that of dmd-3 ( ok1327 ) or mab-3 ( e1240 ) alone [7] . Similarly , two other pairs of paralogs function semi-redundantly: the Hox genes php-3 and nob-1 and the GATA transcription factors egl-18 and elt-6 , which form an operon . In both cases , removal of both genes results in a much more severe disruption of morphogenesis than removal of only one gene . Both modularity and degeneracy contribute in a major way to the robustness of a system [19] . Indeed , the majority of RNAi knockdown phenotypes suggest that male tail tip morphogenesis is very robust against genetic perturbations . In most cases , the effects of RNAi ( as well as some of the mutations tested ) were subtle and the penetrance was low , suggesting that there is extensive buffering of the system against partial depletion of individual transcripts ( or reduced functionality of genes ) . The architecture of the genetic network regulating male tail tip morphogenesis in C . elegans is congruent with the bow-tie model , since we find evidence for all the characteristics of bow-tie networks . We find modularity , degeneracy , a conserved core , weak linkage and positive and negative feedback loops connecting spatial , sexual and temporal inputs to the cellular responses required for morphogenesis . To our knowledge , this is the first time that a genetic network regulating a morphogenetic process has been specifically investigated for bow-tie architecture . However , it is likely that other morphogenetic events—e . g . , the development of the eye in flies and possibly mammals and the formation of the pharynx in C . elegans—are also controlled by bow-tie regulatory networks . In both examples , components have been identified which are likely to be part of the conserved core . Drosophila eye development is in part controlled by the products of eight eye specification genes . Deletion of either one of these genes leads to a drastic reduction or loss of the adult eye , whereas ectopic expression of all but one results in retinal development outside of normal eye tissue [88] , [89] . The fly eye specification genes are conserved with orthologs in mammals . Expression of one of them , dachshund , is regulated by at least 36 upstream factors , including the TGF-β signaling pathway , the transcription factor Zerknüllt and several other patterning genes ( e . g . krüppel , snail and dorsal ) [88] , suggesting the existence of an extensive input fan in this system . The FoxA transcription factor PHA-4 is a central regulator of pharynx development in C . elegans . pha-4 is the only zygotic gene that deletes the entire pharynx when mutated [90] . FoxA transcription factors are conserved from Cnidaria to mammals and are always associated with the digestive tract [90] . PHA-4 has hundreds of targets , many of which are directly regulated at the transcriptional level [91] , [92] . In this system , feedback loops have been identified as well [90] . Thus , there is evidence for a conserved core , an output fan and system control as elements of a bow-tie network architecture for pharynx morphogenesis . Finding bow-tie networks in multiple developmental systems supports the notion that this architecture is a universal feature of evolved gene regulatory networks and is favored by selection due to its robustness . The male tail tip thus provides a simple model for investigating not only morphogenetic mechanisms , but also the properties of a universally important genetic regulatory architecture .
Genetic manipulations and culturing of C . elegans were performed as previously described [98] . We use the following nomenclature for transgenes ( similar to that used by Ziel et al . [99] ) . Transcriptional reporters are designated by the name of the gene associated with the promoter , followed by a “>” and the reporter gene to which it is fused ( e . g . , dmd-3>yfp ) . Translational reporters are designated by the gene , followed by “::” and the reporter to which it is fused ( e . g . , sma-3::mCherry ) . Unless otherwise stated , the endogenous promoter is used to drive expression of translational reporters . If a different promoter is used , we use both designations ( e . g . , lin-44>php-3::gfp represents the php-3 gene fused to the gfp gene , with expression driven by the promoter of the lin-44 gene ) . Unless otherwise stated , the unc-54 3′UTR is used for all constructs . The protein product of a construct is designated with capital letters ( e . g . , PHP-3::GFP ) . No construct employed cDNA; all introns were included . Strains with transgenes generated for this paper are listed in Table S3 . Other strains used for this study are listed below . CB4088 = him-5 ( e1490 ) V . This is the otherwise wild-type , male-producing strain used as the background genotype in this study ( from Caenorhabditis Genetics Center , CGC ) . BW2020 = ctIs57[nob-1::gfp + rol-6] ( a gift from Zhongying Zhao , University of Washington , Washington ) . Hermaphrodites of this strain were crossed with CB4088 males to obtain a him-59 ( e1490 ) V; ctIs57 strain . DF125 = php-3 ( ok919 ) III; him-5 ( e1490 ) V . Made by crossing CB4088 males with RB998 hermaphrodites . RB998 = php-3 ( ok919 ) III ( from CGC ) . DF159 = rme-8 ( b1023 ) I; him-5 ( e1490 ) V . Made by crossing CB4088 males with DH1206 hermaphrodites . DH1206 = rme-8 ( b1023 ) I ( from CGC ) . DF160 = blmp-1 ( tm548 ) I; him-5 ( e1490 ) V; fsIs3[dmd-3>yfp + unc-122>gfp] . DF161 = blmp-1 ( tm548 ) I; him-5 ( e1490 ) V . Made by crossing CB4088 males with hermaphrodites carrying the blmp-1 ( tm548 ) allele ( from Shohei Mitani , National BioResource Project , Tokyo Women's Medical University School of Medicine , Tokyo , Japan ) . DF163 = sma-3 ( e491 ) III; him-5 ( e1490 ) V . Made by crossing CB4088 males with CB491 hermaphrodites . CB491 = sma-3 ( e491 ) III ( from CGC ) . DF164 = egl-18 ( ok290 ) IV; him-5 ( e1490 ) V . Made by crossing CB4088 males with JR2370 hermaphrodites . JR2370 = egl-18 ( ok290 ) IV ( from CGC ) . DF167 = him-5 ( e1490 ) V; nhr-25 ( ku217 ) X . Made by crossing CB4088 males with MH1955 hermaphrodites . MH1955 = nhr-25 ( ku217 ) X ( from CGC ) . DF171 = him-5 ( e1490 ) V; bIs34[rme-8::gfp + rol-6] . Made by crossing CB4088 males with DH1336 hermaphrodites . DH1336 = bls34[rme-8::gfp + rol-6] ( from CGC ) . DF177 = him-5 ( e1490 ) V; nhr-25 ( ku217 ) X; fsIs3[dmd-3>yfp + unc-122>gfp] . DF178 = mab-3 ( e1240 ) II; dmd-3 ( ok1327 ) him-5 ( e1490 ) V; bIs34[rme-8::gfp + rol-6] . DF196 = him-5 ( e1490 ) V; xnIs8[pJN343: nmy-2::mCherry + unc-119 ( + ) ] . This strain was made by crossing CB4088 males to hermaphrodites carrying the transgene xnls8 which were a generous gift from Jeremy Nance ( NYU Skirball Institute , New York , New York ) . DF197 = mab-3 ( e1240 ) II; dmd-3 ( ok1327 ) him-5 ( e1490 ) V; xnIs8[pJN343: nmy-2::mCherry + unc-119 ( + ) ] . DF199 = ptl-1 ( ok621 ) III; him-5 ( e1490 ) V . Made by crossing CB4088 males to RB808 hermaphrodites . RB808 = ptl-1 ( ok621 ) III ( from CGC ) . JJ1473 = unc-119 ( ed3 ) III; zuIs45[nmy-2::gfp + unc-119 ( + ) ] ( from CGC ) . KC447 = rrf-3 ( pk1426 ) II; him-5 ( e1490 ) V . A generous gift from King L . Chow ( Hong Kong University of Science and Technology , Hong Kong , China ) . KC529 = eri-1 ( mg366 ) IV; him-5 ( e1490 ) V . ( from K . L . Chow ) . UR157 = fsIs2[dmd-3>yfp + unc-122>gfp]I ? ; him-5 ( e1490 ) V . A generous gift from Douglas Portman ( Rochester University , New York ) . UR279 = mab-3 ( e1240 ) II; dmd-3 ( ok1327 ) him-5 ( e1490 ) V ( from D . Portman ) . WM79 = rol-6 ( n1270 ) II; neEx[lit-1::GFP + rol-6 ( su1006 ) ] ( from CGC ) . Hermaphrodites of this strain were crossed with CB4088 males to obtain a rol-6 ( n1270 ) II; him-5 ( e1490 ) V; neEx[lit-1::GFP + rol-6 ( su1006 ) ] strain . The genome-wide RNAi-feeding screen was carried out in the RNAi-hypersensitive background rrf-3; him-5 ( strain KC447 ) . RNAi effects on embryogenesis were bypassed by feeding siRNA-expressing bacteria to synchronized L1 larvae . Following a recommendation by K . Chow ( pers . comm . ) , L1 larvae were plated onto a thin film of agar ( 1 . 5 ml per 60 mm plate ) containing 2 mM IPTG and 100 µg/ml ampicillin . Worms were cultured to adulthood ( 3 days ) at 20°C at which time a square of agar with worms was removed and mounted directly onto a glass slide and covered with a coverslip ( Figure 1A ) . All scoring was carried out at 400x with a Zeiss Axioskop equipped with Nomarski differential interference contrast . Images were recorded with a C4742-95 “Orca” Hamamatsu digital camera and Openlab software , ver . 3 . 0 . 9 ( Improvision ) . The secondary screen was carried out in the same way but in a different RNAi hypersensitive background , eri-1 ( strain KC529 ) . RNAi clones consistently conferring a Lep or Ore tail tip phenotype were sequenced to confirm the targeted genes . All scoring data and images are available via our male tail tip database , MTTdb , at http://wormtails . bio . nyu . edu . Translational fusions and transcriptional reporters were constructed by overlap-extension PCR as previously described [100] . The 5′ upstream sequence and coding sequences for php-3 ( -500 bp to the stop codon ) , egl-18 ( -3691 to stop ) , rcn-1 ( -4797 to stop ) , ptl-1 ( -2105 to stop ) , arl-1 ( -550 to stop ) , abcx-1 ( -437 to stop ) , cdc-42 ( -1814 to stop ) , sys-1 ( -3411 to stop ) , inx-12 ( -3975 to stop ) , blmp-1 ( -4916 To stop ) , nhr-165 ( -1559 to stop ) , and ubc-12 ( -363 to stop ) , were PCR-amplified from genomic DNA and fused to gfp and the unc-54 3′-UTR amplified from pPD95 . 75 ( Addgene ) . The 5′-upstream sequence and coding region of sma-3 ( from -1169 bp ) , amplified from genomic DNA , was fused to mCherry [101] and to cfp , which were PCR-amplified from pGC326 ( a gift from E . J . Hubbard ) and pPD136 . 61 ( Addgene ) , respectively . To make the DAF-4 reporters , the upstream sequence and coding region ( from -5091 ) was fused to yfp that was amplified from pPD136 . 64 ( Addgene ) . A transcriptional reporter of daf-4 fused the upstream sequence ( -4865 to -1 ) to the NLS and gfp amplified from pPD122 . 13 ( Addgene ) . The transcriptional reporter for nhr-25 fused the upstream sequence ( -9100 to -1 ) to the NLS and gfp from pPD122 . 13 ( Addgene ) followed by the 3′-UTR for nhr-25 ( stop to +760 bp ) . Transcriptional reporters for inx-12 ( -3975 to -1 ) and inx-13 ( -1557 to -1 ) were fused to yfp ( pPD136 . 64 ( Addgene ) ) and cfp ( pPD136 . 61 ( Addgene ) ) . Transgenes were microinjected at concentrations ranging from 5–20 ng/µl along with 100 ng/µl pRF4 ( rol-6 ( su1006 ) ) as injection marker . Multiple lines were analyzed for each construct using epifluorescence ( Axioskop with mercury lamp , 400 or 1000x ) . Representative images of fluorescence expression patterns are available via the MTTdb database at http://wormtails . bio . nyu . edu . Strain names and primer sequences are provided in Table S3 . Genes identified in our screen were manually entered into N-Browse2 ( http://Aquila . bio . nyu . edu/NBrowse2/nbrowsetest . jnlp ) [17] . Only a subset of these genes showed annotated interactions , and only those with an interaction one or two edges away from another candidate gene were added to our network ( Figure 5 ) . Information from other studies [6] , [7] , [44] , [51] , [57]–[60] which show genetic or direct interactions with known or newly identified tail tip genes were also included ( Figure 5 ) . | Morphogenesis is a process in which cells change their shape and position to give rise to mature structures . Elucidation of the molecular basis of morphogenesis and its regulation would be a major step towards understanding organ formation and functionality . We focus on a powerful model for morphogenesis , the four-celled tail tip of the C . elegans male , which undergoes morphogenesis during the last larval stage . To comprehensively determine the components that regulate and execute male tail tip morphogenesis , we performed a genome-wide RNAi screen . We identified 212 genes that encode proteins with roles in fundamental processes like endocytosis , vesicular trafficking , cell–cell communication , and cytoskeletal organization . We determined the interactions among several of these genes to reconstruct a first draft of the genetic network underlying tail tip morphogenesis . The structure of this network is consistent with the "bow-tie architecture" that has been proposed to be universal and confers evolvability and robustness to biological systems . Bow-tie networks have a conserved core which is linked to numerous input and output components . Many components of the network underlying tail tip morphogenesis in C . elegans are conserved all the way to humans . Thus , understanding tail tip morphogenesis will inform us about morphogenesis in other organisms . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"developmental",
"biology",
"genetics",
"biology",
"morphogenesis",
"gene",
"networks",
"genetics",
"and",
"genomics"
] | 2011 | A Bow-Tie Genetic Architecture for Morphogenesis Suggested by a Genome-Wide RNAi Screen in Caenorhabditis elegans |
Cystic echinococcosis ( CE ) is a chronic , complex and neglected disease caused by the larval stage of Echinococcus granulosus . The effects of this neglect have a stronger impact in remote rural areas whose inhabitants have no chances of being diagnosed and treated properly without leaving their jobs and travelling long distances , sometimes taking days to reach the closest referral center . In 1980 our group set up a control program in endemic regions with CE in rural sections of Rio Negro , Argentina . Since 1997 , we have used abdominopelvic ultrasound ( US ) as a screening method of CE in school children and determined an algorithm of treatment . To describe the training system of general practitioners in early diagnosis and treatment of CE and to evaluate the impact of the implementation of the field program . In 2000 , to overcome the shortage of radiologists in the area , we set up a short training course on Focused Assessment with Sonography for Echinococcosis ( FASE ) for general practitioners with no previous experience with US . After the course , the trainees were able to carry out autonomous ultrasound surveys under the supervision of the course faculty . From 2000 to 2008 , trainees carried out 22 , 793 ultrasound scans in children from 6 to 14 years of age , and diagnosed 87 ( 0 . 4% ) new cases of CE . Forty-nine ( 56 . 4% ) were treated with albendazole , 29 ( 33 . 3% ) were monitored expectantly and 9 ( 10 . 3% ) were treated with surgery . The introduction of a FASE course for general practitioners allowed for the screening of CE in a large population of individuals in remote endemic areas with persistent levels of transmission , thus overcoming the barrier of the great distance from tertiary care facilities . The ability of local practitioners to screen for CE using US saved the local residents costly travel time and missed work and proved to be an efficacious and least expensive intervention tool for both the community and health care system .
Hydatidosis or cystic echinococcosis ( CE ) is a disease caused by the larval stage of the cestode Echinococcus granulosus . It is endemic in sheep raising areas and is among the most neglected diseases in the world today [1] , [2] . Despite a global burden calculated at 1 009 662 disability adjusted life years ( DALYs ) , CE continues to be excluded from funding associated with conditions related to low socioeconomic status [3] . The effects of this neglect have a stronger impact in remote rural areas whose inhabitants have no chances of being diagnosed and treated properly without leaving their jobs and travelling long distances , sometimes taking days to reach the closest referral center . Rio Negro , situated on the northern border of Patagonia in the south of Argentina , is one such place . The main endemic area for CE measures 143 , 048 km2 with a population at risk ( e . g . shepherds , people in contact with stray dogs , etc ) of 85 , 509 people [4] , [5] , [6] . The health structure in the endemic area is comprised of three tertiary care hospitals ( Roca , Viedma , Bariloche ) , 10 rural ( secondary care ) hospitals and a network of 57 rural or suburban health outposts ( primary care ) . The entire health system is staffed by nurses or other paramedics/clinical officers . Distances between a health outpost and its referral rural hospital can be as long as 120 Km by rural roads , often made impassable by the harsh weather conditions of Patagonia ( ice , snow and rain in winter ) . Likewise , distances between rural hospitals and the tertiary care hospitals can be as long as 300 Km . making diagnosis , treatment and control of CE extremely difficult ( Figure 1 ) . To address these problems , in 1997 our group set up a program to conduct annual ultrasound examinations of groups at risk including shepherds and children living in endemic areas . We treated active cysts with albendazole ( ABZ ) and monitored the response to treatment using a portable US scanner that was transported by bus to the various rural sites . The goal was to provide a cost-effective way to deliver care for CE in patients living in remote rural areas . The sensitivity and specificity of US of abdominal CE in rural settings compared to plain radiography or computed tomography ( CT ) has been assessed in a previous work . In that study , 1054 schoolchildren living in rural areas were evaluated by US . Twenty-seven cases ( 100% ) diagnosed with abdominal CE were re-examined by specialists in a tertiary care center with US , radiography and CT scan . They confirmed the diagnosis in 24 patients while 3 cases turned out to be negative . The sensitivity of US was 100% and the specificity was 95 . 6% [7] . In Argentina , US surveys for CE in rural areas are generally carried out by radiologists . Since they are in short supply and mostly employed in tertiary centers , the number of US exams that could be conducted in primary care settings was limited . To address this discrepancy , we set up a two-day FASE course in 2000 for general medicine physicians without previous experience in US . At the end of the course , the physicians learned to diagnose abdominal CE , treat positive cases with ABZ according to a specific algorithm and monitor the response to therapy with US under specialist supervision . The aim of this study is to describe the training system of general practitioners in the diagnosis and treatment of CE and to evaluate the impact of the implementation of the field program .
Childhood screening is done as a part of the school health program through the Ministry of Health for Rio Negro . Each child's parents gave written consent prior to be scanned . All participants in the ultrasound training course gave written consent . The Medical Committee of Control Program Against Hydatidosis in Rio Negro ( established by Decree 6412/06 of the Ministry of Health of the Province of Rio Negro ) gave consent for the study .
Between 2000 and 2008 , 180 general practitioners and general medicine residents attended the FASE course and passed the exam . On the first screening performed by trainees immediately after the course , all suspected cases turned out to be false positives ( for specificity ) when the tutors ( radiologists ) rescanned the patients . The cost of having a specialist rescanning the patients to look for diagnostic errors was , however , lower to that of false negatives with ensuing complications due to an undiagnosed cyst . From 2001 to 2008 the trainees performed 22 , 793 ultrasound studies on children between 6 and 14 years of age and diagnosed CE in 87 cases ( 0 . 4% ) with an average of 8 . 9 years with 94 cysts identified ( 85 . 2% CE1 or CE2 and 6 . 4% CE4 or CE5 ) . Forty-nine ( 56 . 4% ) cases were enrolled in the ABZ protocol , 29 ( 33 . 3% ) were enrolled the expectant management ( watch and wait ) protocol and 9 ( 10 . 3% ) were treated with surgery ( Table 1 ) . Of the 22 , 793 ultrasounds performed , some may have been inadvertently repeated on healthy children despite being negative and this might explain the low prevalence of 0 . 4% . The last control used for our analysis occurred in 2009 . The average length of follow-up was 12 . 9 years for each patient . Of those scanned , 15 ( 24 . 2% ) cysts remained CE1 or CE2 , with their size unchanged and 37 ( 59 . 7% ) were CE4 and CE5 . Cysts disappeared in 6 ( 8 . 7% ) cases ( Table 1 ) . By 2009 , follow-up by general practitioners of cases first diagnosed in 1998–1999 reached 100% of cases at 5 years and 64% at 10 years from diagnosis [11] , [12] . Follow-up of patients first diagnosed in 2000–2008 reached 88 . 5% of cases by June 2010 . A detailed description of the evolution of diagnosed cases has been discussed in previous reports [12] , [13] .
Control programs for CE have been started in various endemic areas throughout the world . [14] . In the Province of Rio Negro it was started in 1980 , based on the deparasitization of dogs and epidemiological surveillance using serology from 1980–1997 and subsequently abdominopelvic US scans . Positive cases were subsequently studied with the imaging techniques available at the time ( X-Rays , scintigraphy , CT and US ) , and surgery was performed after determining the cyst location [5] , [6] . In 1997 , US was adopted as the method of choice for screening of CE due to its greater specificity and sensitivity compared with serologic tests [7] , [15] , [16] , [17] . Meanwhile , new treatment options had been made available with the introduction of benzimidazoles and of percutaneous treatments [18] . As the limitations of surgery ( morbidity , mortality , relapses depending on type of surgery and available medical facilities ) became apparent compared to new therapeutic options [19] , the US classifications of CE became a key element in clinical decision making and a more rational allocation of patients to treatment was implemented . For hepatic CE , this is based on cyst stage ( active , transitional , inactive , complicated or no complicated ) and size [20] , [21] , [22] , [23] . Experience has shown that small , active and uncomplicated cysts tend to respond well to ABZ , thus making surgery unnecessary [24] . Field studies in the Rio Negro Province showed that treatment of active cysts in young asymptomatic carriers with ABZ , in 56% to 88% of patients , resulted in a partial or total involution , the latter being a complete solidification , or stages CE4 and CE5 [12] , [13] . Due to the lack of ionizing radiation , US is eminently repeatable and therefore the best tool to monitor response to treatment , which in CE is a long term endeavor [18] . The cost of ABZ treatment ( for 4 months ) is far lower than that of surgery in Rio Negro , being estimated at around 1 , 350 USD against 4 , 596–5 , 936 USD for surgery [25] . Medical treatment of CE has reduced the need for hospitalization , the general healthcare costs and social cost by reducing the number of work days lost not only from treatment but also the time necessary for rural inhabitants to travel to tertiary care centers [25] . Moreover , recent clinical observations have confirmed findings from our previous studies that while selected cyst stages responded favorably to medical treatment , others needed no treatment at all [1] , [2] , [22] , a finding whose consequences include the lowering of health and social costs of CE . On the epidemiological side , US allows for the assessment of the effects of control programs on the human population of the targeted areas by evaluating cyst size and stage [13] , [26] . With successful control programs , fewer active small cysts are expected to be seen in the population surveyed over time [14] . US is no longer limited to radiologists and is now being utilized by many different specialties [27] . As an example , in Emergency Medicine , FAST ( Focused Assessment with Sonography for Trauma ) can be quickly taught to surgeons and emergency medicine specialists and has revolutionized the approach to blunt abdominal trauma and other acute surgical conditions [28] , [29] . Although the clinical management of many infectious diseases benefits from clinical US [30] , this tool is especially useful in CE , where its use spans epidemiology , diagnosis and staging , guide for percutaneous treatments and follow-up . Despite this , it has rarely [7] been used for active case finding , clinical decision-making and follow-up in rural areas . This is currently changing thanks to the increasing use of portable scanners [31] and by teaching non-radiologists a limited set of specific skills under the supervision of experienced specialists . This idea has been used in other diseases , for example extrapulmonary tuberculosis in rural areas of South Africa where HIV is highly endemic [32] and will be further extended with the continuing miniaturization of scanners [33] . While long term programs are needed for a more comprehensive on-site training [34] , [35] in low resource settings , short focused US courses on specific conditions should also be developed to address the lack of specialists in remote areas . With this experience , the training of general practitioners allows for an annual mass screening program by US and the allocation of patients to treatment with ABZ . Although all the trainees who participated in the screening for CE with US have left due to the high turnover of rural doctors ( moving to tertiary care hospitals ) we have found that having a short ultrasound teaching course allows for a smooth transition and ongoing continuing medical education of new physicians . We believe that this approach deserves to be extended to other rural areas in countries where CE is endemic . US is a wonderful technology that can be applied to the diagnosis and treatment of CE in areas with only rural hospitals and schools , including the use of health care workers with no previous experience in ultrasound in the first step of the diagnosis [8] , [36] . Portable US scanners and a short FASE course for general practitioners , together with the collaboration of veterinarians , surgeons , radiologists and paramedics and the availability of ABZ , helped prevent clinical problems resulting from unchecked cyst growth , minimized social and health care costs and avoided expenses related to time spent by rural patients travelling long distances to tertiary care centers . The course provided the local population the benefit of having US performed at no cost and with minimal disruption to their lives given the lack of transportation time to referral hospitals or clinics . Timely diagnosis and early treatment produced a marked decrease in morbidity , number of hospital admissions and the strain on medical services in patients with CE [4] , [11] . The need to travel to referral hospitals for the traditional surgical interventions was thus limited to a few cases and ultrasound was brought to rural areas instead , extending affordable healthcare to people who would otherwise not have access to it . Just as US revolutionized the clinical management and control of CE , focused US training has the potential to extend its benefits to underserved populations at an affordable cost for the community . | Cystic echinococcosis ( CE ) is an important and widespread disease that affects sheep , cattle , and humans living in areas where sheep and cattle are raised . CE is highly endemic in rural sections of Rio Negro , Argentina , where our group is based . However , it requires continuous monitoring of both populations with human disease best assessed by means of ultrasound ( US ) screening . This is challenging in remote rural areas due to the shortage of imaging specialists . To overcome this hurdle , we set up a two-day training program of Focused Assessment with Sonography for Echinococcosis ( FASE ) on CE for family medicine practitioners with no previous experience in US . After the course , they were equipped with portable US scanners and dispatched to remote rural areas in Rio Negro where they screened patients , located and staged the cysts and decided on the treatment with the help of surgeons and radiologists in local tertiary care centers . The need to travel to referral hospitals for traditional surgical interventions was therefore limited to a few cases . US was instead brought to rural areas thereby extending affordable healthcare to people who would otherwise not have access to it . | [
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] | [
"medicine",
"infectious",
"diseases",
"global",
"health"
] | 2012 | Early Diagnosis, Treatment and Follow-Up of Cystic Echinococcosis in Remote Rural Areas in Patagonia: Impact of Ultrasound Training of Non-Specialists |
Emerging evidence has shown microRNAs ( miRNAs ) play an important role in human disease research . Identifying potential association among them is significant for the development of pathology , diagnose and therapy . However , only a tiny portion of all miRNA-disease pairs in the current datasets are experimentally validated . This prompts the development of high-precision computational methods to predict real interaction pairs . In this paper , we propose a new model of Logistic Model Tree for predicting miRNA-Disease Association ( LMTRDA ) by fusing multi-source information including miRNA sequences , miRNA functional similarity , disease semantic similarity , and known miRNA-disease associations . In particular , we introduce miRNA sequence information and extract its features using natural language processing technique for the first time in the miRNA-disease prediction model . In the cross-validation experiment , LMTRDA obtained 90 . 51% prediction accuracy with 92 . 55% sensitivity at the AUC of 90 . 54% on the HMDD V3 . 0 dataset . To further evaluate the performance of LMTRDA , we compared it with different classifier and feature descriptor models . In addition , we also validate the predictive ability of LMTRDA in human diseases including Breast Neoplasms , Breast Neoplasms and Lymphoma . As a result , 28 , 27 and 26 out of the top 30 miRNAs associated with these diseases were verified by experiments in different kinds of case studies . These experimental results demonstrate that LMTRDA is a reliable model for predicting the association among miRNAs and diseases .
MicroRNAs ( miRNAs ) are a small class of endogenous non-coding RNAs with a length of about 20–24 nucleotides [1] . They bind to the 3'-untranslated region of target miRNA through sequence-specific base pairing , resulting in cleavage or translation inhibition of target miRNA , and thereby regulating gene expression at the post-transcriptional level [2] . A growing body of research has shown that miRNA plays an important role in many biological processes , and their mutations and dysfunctions may lead to a variety of diseases [3] . Therefore , it is very important to identify the relationship among miRNAs and diseases , which has become a research hotspot in recent years . Early studies often use biological experiments to determine the impact of a single factor on the results of the experiment and achieve higher accuracy . Lee et al . discovered the first miRNA in 1993 , that is , the presence of Lin-4 in C . elegans [4] . Since then , many miRNAs have been discovered and identified by using different biological experimental methods , thus giving new insights into the functions and regulatory mechanisms of miRNAs [5 , 6] . Furthermore , these studies have demonstrated that miRNAs are associated with many important biological processes , such as viral infection [7] , immune reaction [8] , tumor invasion [9] , signal transduction [10] , cell proliferation [11] , cell growth [12] , and cell death [13] . With the development of biotechnology , more and more miRNA-disease associations have been revealed . By studying the expression changes of cancer-associated miRNAs in the early stage of HBV-associated hepatocarcinogenesis , Gao et al . found that the deregulation of miRNAs is an early event and accumulates in various steps of HBV-associated hepatocarcinogenesis . At the same time , their results also indicate that miR-145 is a candidate tumor suppressor miRNA , which may play an important role in the development of HCC [14] . Bang et al . discovered that miR-23 , miR-27 and miR-24 cluster are involved in angiogenesis and endothelial apoptosis during cardiac ischemia and retinal vascular development , and plays an important role in cardiovascular angiogenesis [15] . However , the traditional experimental methods have the disadvantages of long experimental cycle , high cost , small scale and easy to be disturbed by the outside world . Therefore , researchers are committed to finding more efficient computational methods to achieve large-scale and credible predictions of the association among miRNAs and diseases . Based on the hypothesis that functionally similar miRNAs tend to be associated with diseases with similar phenotypes , many computational methods for predicting miRNA-disease association have been proposed [16–18] . These computational methods can be roughly divided into two categories: similarity-based measures methods and machine learning-based methods [19–21] . The former predicts miRNA-disease association by measuring the association strength between nodes in miRNA and disease network , while the latter applies the machine learning correlation algorithm to this problem [22–24] . Chen et al . proposed the RWRMDA method and applied it to the miRNA-miRNA functional similarity network , which starts at a given seed node and randomly simulates the transfer process of the pedestrian from the current node to its neighboring nodes in the network , thus predicting the relationship between miRNA and disease [25] . Liu et al . constructed a heterogeneous network by combining data from multiple sources and applied the random walk algorithm to predict miRNA-disease associations . In this method , the functional similarity information of miRNA , semantic similarity information of diseases and miRNA-disease association information are added to the network model , so that it can predict the potential association of new diseases with unknown miRNA related information [26] . Zeng et al . proposed a prediction method based on social network analysis , which combines social network analysis with machine learning to predict the relationship between miRNA and disease under the premise of known miRNA-disease association , miRNA-miRNA functional similarity , and disease-disease similarity [27] . Zou et al . used a supervised machine learning approach to predict miRNA-disease associations by training the biased SVM classifier with bootstrap aggregating algorithm [28] . In this study , we propose a new computational method of Logistic Model Tree for predicting miRNA-Disease Association ( LMTRDA ) based on the assumption that functionally similar miRNAs are often associated with phenotypically similar diseases , and vice versa . The LMTRDA combines multiple sources of data information , including miRNA sequence information , miRNA functional similarity information , disease semantic similarity information , and known miRNA-disease association information . In particular , LMTRDA incorporates biological sequence information of miRNAs extracted by natural language processing techniques . Specifically , LMTRDA first respectively calculates the similarity between miRNA and disease according to the miRNA functional similarity network and disease semantic similarity network , and combines them with the Gaussian interaction profile kernel similarity network to obtain the similarity descriptors of miRNA and disease . Secondly , the Natural Language Processing ( NLP ) technology is used to extract the feature information of the miRNA sequence , and the sequence information and the similarity information of each miRNA-disease pair are combined to form a complete feature descriptor according to the known miRNA and disease association . Finally , the reduced dimension feature descriptors are fed into the Logistic Model Tree ( LMT ) classifier to predict the associations among miRNAs and diseases . The flowchart of LMTRDA model to predict potential miRNA-disease associations is shown in Fig 1 . To evaluate the performance of LMTRDA , the five-fold cross-validation was implemented on the newly released HMDD V3 . 0 dataset . As a result , LMTRDA obtained 90 . 51% prediction accuracy with 92 . 55% sensitivity at the AUC of 90 . 54% . In comparison with different classifiers and feature descriptors , LMTRDA also achieved good results . Furthermore , we validated the proposed model against three human diseases including Breast Neoplasms , Colon Neoplasms and Lymphoma . Ultimately , most of the top 30 miRNA candidates associated with these three diseases ( 28 of 30 in Breast Neoplasms , 27 of 30 in Colon Neoplasms , 26 of 30 in Lymphoma ) predicted by LMTRDA were confirmed in some representative databases . These experimental results indicated that LMTRDA is well suitable for predicting miRNA-disease association .
In the experiment , we validate our model using the HMDD ( Human microRNA Disease Database ) dataset provided by Li et al . [29] . The HMDD dataset provides experiment-supported evidence for human miRNA and disease association , which collects miRNA and disease association data from the evidence of circulating miRNAs , epigenetics , genetics and miRNA-target interactions , and contains detailed and comprehensive annotations . Currently , the latest version of the HMDD dataset is V3 . 0 , which collects 32281 miRNA-disease association entries , including 1102 miRNAs and 850 diseases from 17412 papers . This dataset can be downloaded from the http://www . cuilab . cn/hmdd . When pre-processing the dataset , we removed some of the miRNAs because their information was judged to be unreliable by the public database miRBase . After screening , we chose 32226 miRNA-disease association pairs containing 1057 miRNAs and 850 diseases as positive samples in the experiment . Since HMDD does not provide unrelated miRNA-disease association entries , we randomly selected 32226 miRNA-disease pairs as negative samples from all possible miRNA-disease pairs that have removed the positive samples . In fact , the negative sample set thus constructed may contain positive samples that have not been confirmed by the experiment . However , from a statistical point of view , the proportion of negative samples we selected from all possible samples is only 32226÷ ( 850×1057 ) ≈0 . 0358 , and the number of samples with actually interactions as negative sample sets is very small . Ultimately , the dataset used in our experiment contained 64456 samples , of which positive and negative samples accounted for half . On this basis , we constructed the adjacency matrix AD of miRNA and disease , which consists of 850 rows and 1057 columns , corresponding to 850 diseases and 1057 miRNAs , respectively . When disease d ( i ) and miRNA m ( j ) are verified to be related by the HMDD V3 . 0 database , the element AD ( d ( i ) , m ( j ) ) of the adjacency matrix AD is assigned to 1 , otherwise it is assigned to 0 . Known human miRNA-disease associations and their names obtatined from HMDD V3 . 0 database can be seen in S1–S3 Tables . The disease semantic similarity information we use comes from the MeSH database , which can be downloaded from the National Library of Medicine database at https://www . nlm . nih . gov/ . The MeSH database gives a rigorous disease classification system of diseases , which provides great help for the study of disease semantic similarity [30] . In the system , the relationship among diseases is described as the Directed Acyclic Graph ( DAG ) , where node represents disease and edge represents their relationship [31] . If the disease d ( i ) is related to the disease d ( j ) , use the edge to connect them , indicating that the child node d ( i ) comes from the parent node d ( j ) . Thus , disease d ( i ) can be described as DAGd ( i ) = ( d ( i ) , Nd ( i ) , Ed ( i ) ) , where Nd ( i ) is the ancestor node set of d ( i ) including d ( i ) , and Ed ( i ) is the edge set containing the corresponding edges . We define the contribution of disease s in DAGd ( i ) to the semantic value of disease d ( i ) as follows: {Dd ( i ) ( s ) =1ifs=d ( i ) Dd ( i ) ( s ) =max{ε·Dd ( i ) ( s′ ) |s′∈childrenofs}ifs≠d ( i ) ( 1 ) Where ε is the semantic contribution factor linking disease s and its child disease s′ . In the DAG of disease d ( i ) , the contribution value of disease d ( i ) to its own semantic value is defined as 1 . Therefore , we can get the semantic value DV ( d ( i ) ) of disease d ( i ) , and its formula is as follows: DV ( d ( i ) ) =∑s∈Nd ( i ) Dd ( i ) ( s ) ( 2 ) Here , we assume that diseases sharing more parts of their DAGs will have higher semantic similarity . By considering the relative position of disease d ( i ) and disease d ( j ) in the MeSH disease DAG , the semantic similarity value SV1 ( d ( i ) , d ( j ) ) between them can be calculated , and the formula is as follows . In the SV1 model , we mainly consider the relationship between the layers of disease in DAG graph , that is , the contribution of different diseases in the same layer to the semantic value is the same . However , we observed that the number of different diseases appearing in the DAGs is different , and the contribution of disease less appearing in the DAGs should be higher than that of disease more appearing in the DAGs . Therefore , in order to distinguish this situation , we introduce the second calculation model [32] of contribution value of disease s , the formula is as follows: Dd ( i ) ′ ( s ) =−log ( num ( DAGs ( s ) ) num ( diseases ) ) ( 4 ) where num ( DAGs ( s ) ) indicates the number of DAGs containing disease s , and num ( diseases ) indicates the number of all diseases . Thus , the second model of semantic similarity value SV2 ( d ( i ) , d ( j ) ) of disease d ( i ) and disease d ( j ) is obtained , and the formula is as follows: SV2 ( d ( i ) , d ( j ) ) =∑s∈Nd ( i ) ∩Nd ( j ) ( Dd ( i ) ′ ( s ) +Dd ( j ) ′ ( s ) ) DV ( d ( i ) ) +DV ( d ( j ) ) ( 5 ) where the value of DV ( d ( i ) ) and DV ( d ( j ) ) are the same as model 1 , which can be calculated using formula 2 . The diseases used in disease similarity model 1 and model 2 are from the MeSH database , which accounts for only a part of the diseases we use . Therefore , the remaining disease similarity scores are calculated using Gaussian interaction profile kernel similarity . Under the hypothesis that functionally similar miRNAs are more likely to be associated with phenotypically similar diseases , Wang et al . proposed a functional similarity model to calculate the functional similarity between different miRNAs [31] , and placing its functional similarity score matrix at http://www . cuilab . cn/files/images/cuilab/misim . zip . In this article , we download it as the miRNA function similarity information . But similar to the case of the disease similarity model , the miRNAs provided in this matrix contains barely a portion of the miRNAs we use . Therefore , we combine it with Gaussian interaction profile kernel similarity to form a complete miRNA similarity matrix . The constructed miRNA functional similarity score matrix can be seen in S4 Table . Since the HMDD V3 . 0 dataset provides a greater number of diseases and miRNAs than the disease and the miRNA similarity models described above , we describe the remaining disease and miRNA similarity information using Gaussian interaction profile kernel similarity [33] . The calculation of Gaussian interaction profile kernel similarity for diseases is based on the hypothesis that similar diseases tend to be functionally similar miRNA , and vice versa . By observing whether disease d ( i ) is associated with each of the 1057 miRNAs we have compiled from the HMDD V3 . 0 dataset , we defined binary vector V ( d ( i ) ) to represent the interaction profiles of disease d ( i ) . Here , the binary vector V ( d ( i ) ) is the row vector of the adjacency matrix AD in which the disease d ( i ) is located . Gaussian interaction profile kernel similarity for diseases GD ( d ( i ) , d ( j ) ) between disease d ( i ) and disease d ( j ) can be calculated as follows: GD ( d ( i ) , d ( j ) ) =exp ( −θd‖V ( d ( i ) ) −V ( d ( j ) ) ‖2 ) ( 6 ) where θd is the width parameter of the function , which can be calculated by normalizing the original parameters . The formula is as follows: θd=1m∑i=1m‖V ( d ( i ) ) ‖2 ( 7 ) where m is the number of rows of the adjacency matrix AD . Similarly , Gaussian interaction profile kernel similarity for miRNA GR ( r ( i ) , r ( j ) ) between miRNA r ( i ) and miRNA r ( j ) can be calculated as follows: GR ( r ( i ) , r ( j ) ) =exp ( −θr‖V ( r ( i ) ) −V ( r ( j ) ) ‖2 ) ( 8 ) θr=1n∑i=1n‖V ( r ( i ) ) ‖2 ( 9 ) where the binary vector V ( r ( i ) ) is the column vector of the adjacency matrix AD in which the miRNA r ( i ) is located , n is the number of columns of the adjacency matrix AD . The sequence of miRNA contains abundant information . In order to describe the characteristics of miRNA more comprehensively , we transform them into numerical vectors and fuse them with the above similarity vectors to form the final descriptors . The usual approach to convert miRNA sequences into numerical vectors is to use k-mers [34] , which refers to the length of a subsequence of k . Given a miRNA sequence of length l , the number of possible k-mers is l−k+1 . For example , 6-mers sequence of miRNA can be represented as AAAAAA , AAAAAC , … , UUUUUU . However , this approach does not take into account the difference between the two k-mers because it treats the distance between any two k-mers as equal . But the difference between AAAAAA and UUUUUU is significantly larger than between AAAAAA and AAAAAC . Therefore , we introduce natural language processing technology to solve this problem [35–38] . It can not only transform the original high-dimensional data into low-dimensional continuous real-valued vector , but also learn its effective representation from miRNA sequences in an unsupervised manner . In this study , we use skip-gram in natural language processing's Word2vec algorithm to learn the distributed representation of miRNA for k-mers , which is a shallow two-layer neural network and represents an item by considering its context information from the nearby items . Given a sequence of words w1 , w2 , … , wn , skip-gram uses the co-occurrence information of words in the context window to learn the word representation , and look for the parameter set θ to maximize the product of the following conditional probabilities . argmaxθ∏w∈T[∏c∈C ( w ) p ( c|w;θ ) ] ( 10 ) where T is the text set; w is a word; c is a word in the context; C ( w ) is the set of words contained in the context in which the word w appears in the text set T; p is a conditional probability , which is defined as follows: p ( c|w;θ ) =exp ( vc∙vw ) ∑c′∈Cexp ( vc∙vw ) ( 11 ) where vc and vw are the column vectors of c and w , respectively; C is the set of words in all contexts , which is equivalent to vocabulary v; and parameter θ is the specific value of each dimension in vc and vw . In experiments , we use 6-mers to transform miRNA sequences , which ultimately get 46 = 4096 6-mers . Taking the AAGUCGUACGAU sequence as an example , 6-mers can convert it to {AAGUCG , AGUCGU , GUCGUA , UCGUAC , CGUACG , GUACGA , UACGAU} . After obtaining the 6-mers of all miRNAs in the HMDD V3 . 0 dataset , we trained the skip-gram word2vec algorithm using all the miRNAs downloaded from the public database miRBase as training sets . In the implementation of the algorithm , we use the following parameters: the minimum number of occurrences of the training words "min_count" is set to 5 , the maximum distance of the word vector context "window" is set to 5 , the dimension size of the word vector "size" is set to 64 , the maximum number of iterations in the stochastic gradient descent method "iter" is set to 10 , and the other parameters are set to default values . In this study , we ultimately used descriptors that fused multiple sources of data including disease similarity , miRNA similarity and miRNA sequence to predict the miRNA-disease association . The advantage is that it can reflect the characteristics of diseases and miRNAs from different perspectives , help to deeply dig out the potential relationship among miRNAs and diseases , and improve the performance of model prediction . For the similarity of diseases , we construct disease semantic similarity model SV1 , disease semantic similarity model SV2 and disease Gaussian interaction profile kernel similarity GD . The disease similarity matrix DSim ( d ( i ) , d ( j ) ) between disease d ( i ) and d ( j ) can be obtained by integrating the above disease similarities . The formula is as follows: DSim ( d ( i ) , d ( j ) ) ={SV1 ( d ( i ) , d ( j ) ) +SV2 ( d ( i ) , d ( j ) ) 2ifd ( i ) andd ( j ) hassemanticsimilarityGD ( d ( i ) , d ( j ) ) otherwise ( 12 ) For the similarity of miRNA , we combined miRNA functional similarity RF and miRNA Gaussian interaction profile kernel similarity GR to form miRNA similarity matrix RSim . The miRNA similarity matrix RSim ( r ( i ) , r ( j ) ) formula for miRNA r ( i ) and miRNA r ( j ) is as follows: RSim ( r ( i ) , r ( j ) ) ={RF ( r ( i ) , r ( j ) ) ifd ( i ) andd ( j ) hasfunctionalsimilarityGR ( r ( i ) , r ( j ) ) otherwise ( 13 ) For the final feature vector FV , we need to integrate the sequence information of miRNA RSeq . The feature vector FV ( d ( i ) , r ( j ) ) formed by diseases d ( i ) and miRNA r ( j ) can be described in the following formula: FV ( d ( i ) , r ( j ) ) =[DSim ( d ( i ) ) , RSim ( r ( j ) ) , RSeq ( r ( j ) ) ] ( 14 ) where DSim ( d ( i ) ) represents the i row vector of disease d ( i ) in the disease similarity matrix DSim; RSim ( r ( j ) ) represents the j column vector of miRNA r ( j ) in the miRNA similarity matrix RSim; RSeq ( r ( j ) ) represents the j row vector of miRNA r ( j ) in the miRNA sequence matrix RSeq . In this study , we use the Logical Model Tree ( LMT ) as a classifier to predict the associations among miRNAs and diseases . The basic idea of LMT originates from the combination of two complementary classification schemes: linear logistic regression and tree induction [39 , 40] . It uses the LogitBoost algorithm to establish the logistic regression function on the node of the tree , and uses the CART algorithm to prune . Specifically , LMT first constructs a basic "weak classifier" based on the existing sample dataset , and calls the "weak classifier" repeatedly . By giving more weight to the wrong samples in each round , it will pay more attention to the samples that are hard to judge . Then , after several rounds of cycles , the "weak classifiers" of each round are combined into the "strong classifier" by weighting method , thereby obtaining a higher precision prediction model . Finally , the tree grown in the training set is pruned using the CART algorithm to obtain the final classification model .
To have a comprehensive assessment of the performance of LMTRDA , we follow common evaluation criteria to evaluate the model , including accuracy ( Accu . ) , sensitivity ( Sen . ) , precision ( Prec . ) and Matthews Correlation Coefficient ( MCC ) . Their calculation formulas are defined as follows: Accu . =TP+TNTP+TN+FP+FN ( 15 ) Sen . =TPTP+FN ( 16 ) Prec . =TPTP+FP ( 17 ) MCC=TP×TN−FP×FN ( TP+FP ) ( TP+FN ) ( TN+FP ) ( TN+FN ) ( 18 ) where TP , TN , FP , and FN respectively indicate the number of correctly predicted positive samples , correctly predicted negative samples , incorrectly predicted positive samples , and incorrectly predicted negative samples by the model . In addition , the Receiver Operating Characteristic ( ROC ) curve and the area under the curve ( AUC ) that can comprehensively reflect the performance of the model are also used in the experiment [41] . To assess the ability of LMTRDA to predict miRNA-disease association , we validated it on HMDD V3 . 0 dataset using the five-fold cross-validation by LMT classifier . Firstly , we divided all 64452 miRNA-disease pairs into five subsets that were disjoint and roughly equal . Secondly , four of them are selected as training sets to train the LMT classifier , and the remaining one is used as a test set to obtain prediction results . Finally , take turns selecting different subsets as the test set and repeat step 2 until all subsets are treated as test set once and only once . We collected the results of these five experiments and used the mean and standard deviation as the final experimental results . Table 1 lists the experimental results of the five-fold cross-validation obtained by LMTRDA on the HMDD V3 . 0 dataset . We can see from the table that LMTRDA has achieved an average prediction accuracy of 90 . 51% . The accuracy of the five experiments is 90 . 99% , 90 . 29% , 90 . 74% , 90 . 22% and 90 . 30% respectively , while the standard deviation is only 0 . 34% . The LMTRDA model obtained the sensitivity , precision , Matthews correlation coefficient and area under ROC curve are 92 . 55% , 88 . 93% , 81 . 10% , and 90 . 54% , with standard deviations of 1 . 11% , 0 . 98% , 0 . 67% and 0 . 33% respectively . The ROC curves and PR curves generated by our proposed method on the HMDD V3 . 0 dataset are shown in Fig 2 and Fig 3 . Our proposed LMTRDA model has achieved satisfactory results on HMDD V3 . 0 dataset using the LMT classifier . In this part of the experiment , we select the state-of-the-art SVM classifier and random forest classifier to compare with it [42] . SVM is a supervised learning algorithm to solve classification problems . It can find the best separated hyperplane in the feature space to maximize the interval between positive and negative samples on the training set , and obtain the global optimization result [43 , 44] . Random forest is a classifier with multiple decision trees whose output is determined by the mode number of output categories of decision trees [45 , 46] . It can improve the prediction accuracy without significantly improving the amount of computation , so it is widely used in the field of pattern recognition and data mining . When classifying with SVM classifier , we optimized its parameters using grid search method and set the kernel function to radial basis function , c = 0 . 5 and g = 0 . 2 . We use radial basis as the kernel function for the SVM classifier , and the optimization results are stored in S5 Table . When classifying with random forest classifier , we also optimized its parameters , setting the maximum depth of the tree to 2 , and other parameters to the default values . Tables 2 and 3 summarize the five-fold cross-validation results performed by SVM and random forest classifier combined with the proposed feature descriptors on the HMDD V3 . 0 dataset . From Table 2 we can see that the accuracy , sensitivity , precision , MCC , and AUC obtained by the SVM model are 86 . 09% , 76 . 14% , 95 . 05% , 73 . 65% and 86 . 10% , and their standard deviations are 0 . 29% , 0 . 60% , 0 . 18% , 0 . 45% and 0 . 38% , respectively . It can be seen from Table 3 that the accuracy , sensitivity , precision , MCC , and AUC achieved by random forest model are 89 . 66% , 88 . 14% , 90 . 90% , 79 . 35% and 89 . 73% respectively . Their standard deviations are 0 . 50% , 0 . 57% , 0 . 50% , 1 . 01% , 0 . 58% , respectively . For convenience of comparison , we summarize the experimental results of the three models and present them in the form of the graph . From the Fig 4 we can visually observe that LMTRDA achieves the highest result among the five evaluation criteria of accuracy , sensitivity , MCC , and the third result in terms of precision . This indicates that LMTRDA does not perform as well as the other two models in terms of the precision , which representing the proportion of true positive samples in the positive samples predicted by the prediction model . But overall , the performance of LMTRDA is optimal , especially on the predictive accuracy and the MCC and AUC that represent the overall performance of the model . From Fig 4 , we also found that the RF model achieved higher results than that of the SVM model , but generally lower than LMTRDA . This shows that the RF classifier is more suitable for the proposed feature descriptors than the SVM classifier , but the LMT classifier is the most suitable one in this model . To evaluate the ability of our proposed descriptors to represent disease and miRNA feature information , we compare them with different descriptors . Since the descriptor we proposed consists of disease similarity information , miRNA similarity information , and miRNA sequence information , we constructed different descriptors to compare with them in this part of the experiment . That is , the descriptor ‘DescSeq’ consisting only of disease similarity information and miRNA sequence information , and the descriptor ‘DescSim’ consisting only of disease similarity information and miRNA similarity information . Tables 4 and 5 list the five-fold cross-validation results generated by the LMT classifier combined with these two descriptors respectively . It can be seen from the table that the accuracy of the ‘DescSeq’ and ‘DescSim’ descriptors generated on the dataset are 87 . 51% and 89 . 43% , the sensitivity are 87 . 25% and 92 . 46% and , the precision are 87 . 71% and 87 . 23% , the MCC are 75 . 03% and 79 . 03% , the AUC are 87 . 61% and 89 . 55% and , respectively . Fig 5 shows the five-fold cross-validation prediction results of three descriptors combined with LMT classifier on HMDD V3 . 0 dataset . As can be seen from the Fig 5 , our proposed descriptors have achieved the best prediction performance on the evaluation criteria accuracy , sensitivity , precision , MCC , and AUC , respectively . In particular , there is a significant improvement in the Accuracy indicating the average accuracy of the prediction model and the MCC and AUC indicating the overall performance of the prediction model . This suggests that the multi-source information descriptor which combines disease similarity , miRNA similarity and miRNA sequence can describe the miRNA-disease association from different aspects , so as to maximize the deeper meaning of miRNA-disease data hiding . To further evaluate the performance of LMTRDA , we implemented the case studies on three diseases including Breast Neoplasms , Colon Neoplasms and Lymphoma . In the experiment , we trained the classifier as the training set for all known miRNA-disease pairs in the HMDD V3 . 0 dataset . The test set is the miRNA-disease pairs consisting of these three diseases and all possible miRNAs . When LMTRDA obtained the predicted results , we took out the 30 miRNAs with the highest scores according to different diseases and verified them in dbDEMC V2 . 0 and miR2Disease databases [47] . Breast neoplasms are neoplasms that occur in breast tissue , accounting for about two-thirds of breast disease . Malignant breast neoplasms are commonly known as breast cancer , and 99% of them occur in women . The global incidence of breast cancer has been on the rise since the late 1970s , and one in eight women in the United States has breast cancer . At present , breast cancer has become a common neoplasm that threatens women's physical and mental health . A large number of experiments show that many miRNAs are related to breast neoplasms . So we selected breast neoplasms as the first case study and use LMTRDA to predict the miRNAs associate with them . The results are shown in Table 6 , 28 out of the top 30 predicted miRNAs are verified in the experimental data provided by the dbDEMC V2 . 0 and miR2Disease datasets . Colon neoplasms are common malignant neoplasms in the gastrointestinal tract , the incidence of which is second only to gastric and esophageal cancer . Lymphoma is a malignant tumor that originates in the lymphoid hematopoietic system . More and more literatures have reported that much miRNAs are closely related to these two diseases . Therefore , we also choose these two diseases as the case study to verify the predictive ability of LMTRDA . Tables 7 and 8 respectively list the top 30 miRNAs with the highest scores associated with the two diseases predicted by LMTRDA . After comparing with the dbDEMC V2 . 0 and miR2Disease database , 27 out of the top 30 miRNAs in the Colon neoplasms disease predictions can be validated , and 26 out of the top 30 miRNAs can be validated in the Lymphoma disease predictions can be validated .
In this study , we present a novel computational method LMTRDA for predicting miRNA-disease association base on fused multi-source data . An interesting aspect of LMTRDA is the use of natural language processing techniques to transform miRNA sequences into numerical vectors and merge them with miRNA functional similarity , disease semantic similarity , and known miRNA-disease association information to form feature descriptors . Cross-validation experiment results on HMDD V3 . 0 dataset demonstrated that this model can effectively predict the potential association among miRNAs and diseases . In comparison with different classifier and feature descriptor models , LMTRDA exhibits good performance . In addition , we validated it in human diseases including Breast Neoplasms , Breast Neoplasms and Lymphoma , and LMTRDA also achieved excellent results . These results indicated that LMTRDA is a reliable model for predicting miRNA-disease association . In future research , we will continue to study how to better apply natural language processing techniques to biological sequence data in anticipation of better performance of predictive mod . | Identification of miRNA-disease associations is considered as an important step for the development of diagnose and therapy . Computational methods contribute to discovering the potential disease-related miRNAs . Based on the assumption that functionally related miRNAs tend to be involved disease , the model of LMTRDA is proposed to prioritize the underlying miRNA-disease associations by fusing multi-source information including miRNA sequences , miRNA functional similarity , disease semantic similarity , and known miRNA-disease associations . Through cross validation , the promising results demonstrated the effectiveness of the proposed model . We further implemented the case studies of three important human complex diseases including Breast Neoplasms , Breast Neoplasms and Lymphoma , 28 , 27 and 26 of top-30 predicted miRNA-disease associations have been manually confirmed based on recent experimental reports . It is anticipated that LMTRDA model could prioritize the most potential miRNA-disease associations on a large scale for advancing the progress of biological experiment validation in the future , which could further contribute to the understanding of complex disease mechanisms . | [
"Abstract",
"Introduction",
"Materials",
"and",
"methods",
"Results",
"and",
"discussion",
"Conclusion"
] | [
"medicine",
"and",
"health",
"sciences",
"linguistics",
"statistics",
"natural",
"antisense",
"transcripts",
"gene",
"regulation",
"cancers",
"and",
"neoplasms",
"social",
"sciences",
"oncology",
"hematologic",
"cancers",
"and",
"related",
"disorders",
"micrornas",
"mathematics",
"forecasting",
"lymphomas",
"information",
"technology",
"research",
"and",
"analysis",
"methods",
"infectious",
"diseases",
"computer",
"and",
"information",
"sciences",
"mathematical",
"and",
"statistical",
"techniques",
"gene",
"expression",
"hematology",
"disease",
"vectors",
"biochemistry",
"rna",
"natural",
"language",
"processing",
"nucleic",
"acids",
"semantics",
"database",
"and",
"informatics",
"methods",
"genetics",
"biology",
"and",
"life",
"sciences",
"physical",
"sciences",
"species",
"interactions",
"non-coding",
"rna",
"statistical",
"methods",
"neoplasms"
] | 2019 | LMTRDA: Using logistic model tree to predict MiRNA-disease associations by fusing multi-source information of sequences and similarities |
The small amyloid-forming GNNQQNY fragment of the prion sequence has been the subject of extensive experimental and numerical studies over the last few years . Using unbiased molecular dynamics with the OPEP coarse-grained potential , we focus here on the onset of aggregation in a 20-mer system . With a total of 16 . 9 of simulations at 280 K and 300 K , we show that the GNNQQNY aggregation follows the classical nucleation theory ( CNT ) in that the number of monomers in the aggregate is a very reliable descriptor of aggregation . We find that the critical nucleus size in this finite-size system is between 4 and 5 monomers at 280 K and 5 and 6 at 300 K , in overall agreement with experiment . The kinetics of growth cannot be fully accounted for by the CNT , however . For example , we observe considerable rearrangements after the nucleus is formed , as the system attempts to optimize its organization . We also clearly identify two large families of structures that are selected at the onset of aggregation demonstrating the presence of well-defined polymorphism , a signature of amyloid growth , already in the 20-mer aggregate .
The aggregation of misfolded amyloid proteins into fibrils is a hallmark of many neurodegenerative diseases such as Alzheimer's , Parkinson's , and Huntington's diseases [1]–[5] and understanding amyloid aggregation mechanisms is crucial for controlling their destructive consequences . Fibrils are known to be ordered insoluble assemblies with a core cross- structure . They are not the only aggregated species involved , however , and oligomers , smaller intermediates on or off the fibril formation pathway , have been found to be responsible for amyloid cytotoxicity [6]–[8] . Their role in amyloid aggregation is still a matter of debate but significant efforts have gone into better understanding and characterizing their structure and dynamics both experimentally [9]–[11] and computationally [12]–[17] . Oligomers are often found to be precursors to amyloid fibrils . They could also , in some cases , appear as the product of a competition between the ordered fibrillar and amorphous globular morphologies , forming via different assembly pathways . This widespread characteristic of amyloid proteins is described as polymorphism [18]–[20] and is under kinetic control [21] . The presence of oligomers is therefore crucial for the fibrillisation process as well as the final morphology of fibrils [22] and understanding their kinetics of formation could be the key to controlling this polymorphism . The aggregation of amyloid proteins is a highly cooperative self-assembly mechanism , which is often described as a complex nucleation and growth process [23] . The nucleation step , in a supersaturation environment , consists of a series of stochastic events leading to the formation of metastable seeds for the oligomer or fibril to grow on [24] . Nucleation kinetics display two characteristic properties: the presence of 1 ) a lag time before aggregates can be detected and 2 ) a maximum growth rate after nucleation is triggered [25] , [26] . Direct experimental observations of nucleation and growth have been reported [27]–[30] and nucleation was always found to be the rate-limiting step of amyloid formation [26] . The aim of the present work is to investigate the dynamics of amyloid aggregation and the forces driving self-assembly for the 20-mer system of the amyloidogenic GNNQQNY peptide using molecular dynamics ( MD ) and a coarse-grained potential ( OPEP ) . The nucleation specificity of the N-terminal region ( 9–39 ) of the budding yeast prion protein Sup35 , GNNQQNY , is well understood . This small heptapeptide alone drives the entire Sup35 protein to self-assemble into amyloid fibrils [31] and , when isolated , displays the same amyloid properties and aggregation kinetics as the full-length Sup35 protein [32] . In addition , its cross- spine structure has been determined at the atomic level by X-ray crystallography [33] . It is therefore a very good candidate to the study of amyloid aggregation kinetics and numerous computations have been performed on the GNNQQNY sequence to characterize the onset of aggregation for this model [34]–[38] . This work expands on our previous multi-scale thermodynamic study of different sizes of GNNQQNY systems , where we identified the morphologies accessible to the trimer , dodecamer and 20-mer [39] . Now , we focus on the aggregation kinetics using long MD simulations of unbiased spontaneous self-assembly . We offer a full analysis of the onset of aggregation for GNNQQNY peptides at a refined coarse-grained level . A total of 16 . 9 of simulations have been collected to allow statistically relevant analyses . Altogether , our results indicate the presence of a nucleated-polymerization process intertwined with oligomer-involving mechanisms , thus leading to a certain degree of polymorphism that is already clearly established for the 20-mer .
As in our previous study , we perform implicit solvent coarse-grained molecular dynamics ( MD ) simulations using the OPEP potential version 3 . 0 [40] . OPEP is designed for efficient protein folding and structure prediction of large systems over long timescales and is also accurate for studying thermodynamics [41] . In OPEP , all heavy backbone atoms are fully represented ( N , H , , C and O ) . Side chains , for their part , are reduced to a single bead with appropriate geometrical properties and van der Waals radius . The implicit effects of the solvent are included in the interaction parameters of the potential energy function , as detailed elsewhere [40] , [42] . OPEP is a well tested potential and has been implemented with a palette of numerical methods such as Monte-Carlo [42]–[46] , the activation-relaxation technique ( ART nouveau ) [47]–[52] , MD [41] , [53]–[55] and REMD [39] , [56]–[59] . Here , two sets of single temperature MD are performed on a 20-mer of GNNQQNY , with both terminii of each peptide charged , in order to characterize in details the kinetics of aggregation . The first set consists of a total of 152 100 ns simulations ( 76 at 280 K and 76 at 300 K ) with configurations saved every 5000 steps . The choice of temperatures is motivated by the fact that 280 K and 300 K are temperatures below and above the transition temperature previously found for the 20-mer of GNNQQNY . As explained below , the initial atomic positions taken for this set are extracted from the simulations reported in Ref . [39] . An additional 10 30 ns simulations are then carried out from a subset of the starting atomic positions of the previous simulation set ( 5 at 280 K and 5 at 300 K ) with configurations saved every 50 steps to better monitor the detailed evolution of the system during the nucleation phase . All simulations are independent , starting with different random Boltzmann distributed velocities . In every case , we maintain simulation conditions as close as those of Ref . [39] , with a Berendsen thermostat for temperature control [60] , an integration time step of 1 . 5 fs and periodic boundary conditions applied to a box 200 Å in size containing 20 monomers of GNNQQNY , which represents a constant 4 . 15 mM concentration . For simplicity , the starting random structures for our simulations were extracted from the high-temperature set generated in our previous REMD OPEP runs of the GNNQQNY 20-mer [39] . A typical starting structure for our simulations is shown in Figure 1 with all 20 peptides isolated and in random coil conformations . At the start of each run , a minimization procedure is performed using a combination of the steepest descent algorithm and the conjugate gradient method [61] , followed by a thermalization of 50 000 steps ( 0 . 075 ns ) to ensure that all conformations are fully thermalized . Because of the implicit solvent treatment as well as the peptide's coarse-grained representation , that decrease the number of degrees of freedom , the aggregation kinetics is accelerated . It is therefore not possible to establish a direct connection between the aggregation time observed in the simulation and in experiments . However , as shown in Ref . [41] , the thermodynamical properties are , at least qualitatively , maintained . The simulations presented here , therefore , should provide the right qualitative picture for the first steps in the kinetics of aggregation . Most of the analysis on the nucleation and growth kinetics is carried out using a clustering tool [39] adapted to multiprotein assembly and designed to classify -sheet clusters based on strand attachment . For the purpose of this work , this procedure can also handle the calculation of kinetic association and dissociation rates . To assess strand attachment , the criterion used to define and calculate hydrogen bonds between strands is similar to the DSSP definition [62] . A peptide belongs to a cluster if it is attached to another strand of that cluster by at least two hydrogen bonds . An additional criterion is applied on dihedrals and angles to determine if a given strand in a cluster has enough amino acids in -conformation . For each amino acid the and angles are calculated and if they satisfy the region ( in degrees ) : [−180∶−150;0∶180] , ( in degrees ) : [−180∶0;150∶180] ( corresponding approximately to the region of the Ramachandran plot [63] ) , the amino acid is in a state . A GNNQQNY peptide is considered in a state if at least three of its residues are in the region . If a peptide is not found to be in a state , it is excluded from the cluster . This determination of secondary structure is solely used to determine cluster membership of the strands . The clustering analysis allows us to measure accurately the evolution of clusters over time based on local information and to monitor their properties such as the orientation of strands within -sheets ( i . e . , parallel or anti-parallel ) . For purposes other than cluster determination , secondary structure calculations are made using the STRIDE program [64] . In order to look at the aggregation process in more details , we also consider the association and dissociation rates of the clusters in the following way . With the concentration of , we consider aggregation as a dynamical process involving both association and dissociation that can occur either one monomer or more than one monomer at a time . The former is referred to as growth by monomer addition/monomer loss while the latter is described as being a mix of two processes , oligomer fusion/fragmentation and formation/destruction of oligomers from/into monomers , when involving more than one monomer at a time . We can then define the net rate of creation of as ( 1 ) where and are the creation rate of into and the destruction rate of into , and are the creation and destruction rates of either directly from/into monomers , or from the fusion/fragmentation of other sizes of oligomers . All the C and D rates are calculated from our clustering tool and allow us to gather statistics on the microscopic kinetic events and mechanisms .
At the lowest temperature of 280 K , all 76 100 ns simulations lead to ordered amyloid oligomers formation . In all cases , aggregation is accompanied by a sudden drop of the total potential energy of the system , by over 600 kcal/mol over less than 10 ns , and by an increase in the -sheet content of 30% , as calculated with the STRIDE program [64] . While the exact energy value is not significative , due to the implicit-solvent coarse-grained nature of our energy model , its drop corresponds to the formation of a more stable structure . The system then stays in a minimum of energy and both the number of hydrogen bonds and the amount of secondary structure stabilize . As shown in Figure 2 ( a ) , which presents a typical aggregation run , the -content in the structures fluctuates typically around 50% , near its maximum of 60% , as the glycines and tyrosines end residues of each 20 peptides do not get involved in the -sheet hydrogen bonding . Figure 2 ( a ) also shows the high correlation between the energy drop and the increase in the number of hydrogen bonds as a function of time , suggesting that the cooperativity between hydrogen bonds plays a crucial role in lowering the energy and stabilizing the system . Aggregation is slower at 300 K and only 68% of the 76 100 ns simulations lead to ordered amyloids . However , as shown in the typical aggregation run in Figure 2 ( b ) , the overall ordering follows a trend very similar to that at the lower temperature : a sudden potential energy drop of over 600 kcal/mol over less than 10 ns accompanied by correlated raises in both the number of hydrogen bonds and the -sheet content . If the final number of hydrogen bonds is very similar to that at 280 K , the secondary structure is less stable and tends to fluctuate around 40% rather than 50% . In order to describe the assembly process we represent the time evolution , the probability density and the orientation of strands in structures as a function of the number of hydrogen bonds and of the number of contacts between side chains as these two coordinates are the least correlated and are the best measure of how ordered the structures are . Figures 3 and S1 show these quantities for the trajectories plotted in Figure 2 at 280 K and 300 K , respectively . At 280 K , we observe three distinct kinetic stages over the course of a typical simulation ( Figure 3 ( a ) ) . The first phase is characterized as the nucleation phase , which lasts about 5 ns after the start of the simulation and leads to the formation of the metastable critical nucleus . During this phase , small oligomers form and break under stochastic collisions of the monomers . Seeds below the nucleus size fluctuate considerably , forming and disassembling at a high rate , forming a quasi-equilibrium perfectly reversible process . Once the metastable nucleus forms , the system can move into the aggregation ( or growth ) phase with a 50% probability , by definition . In this dynamical phase , almost all of the monomers rapidly assemble around the nucleus to form a partially disordered globular oligomer . In general , this stage is very rapid and typically lasts less than 10 ns . During the third phase , which extends over a timescale of up to 80 ns , the aggregate rearranges itself as monomers explore their local configuration environment within the confines of the oligomers , optimizing the energy and , as a consequence , the secondary structure and the number of side chain – side chain contacts ( see the last 75 ns in Figure 3 ( a ) and ( c ) ) . This process , which we describe as a stabilization phase , is the slowest of the three and accounts for the dense region in Figure 3 ( b ) . This aggregation process is consistent with the “condensation-ordering” mechanism previously observed experimentally [65] and computationally [12]–[14] , [66] . An interesting feature of the kinetics at 280 K is the increase and later dominance of parallel orientation in the structures over time during both the growth and stabilization phases while the structures are mostly antiparallel during the nucleation phase ( Figure 3 ( c ) ) . By looking at the color coding on the right axis , it appears as though the system is loosing some parallel orientation between region 1 and 2 from almost 100% to 80% . Instead our results indicate that the system continues to evolve and gain some secondary structure between region 1 and 2 of the graph . It is the newly formed -strands that adopt an antiparallel orientation while the parallel content formed during the growth process remains unchanged . As a whole , 91% of the MD simulations at 280 K lead to a final assembly dominated by parallel -sheets , in agreement with recent experimental findings [33] , [67] , [68] and computational studies [35] , [36] , [39] , [69] , [70] . At 300 K the kinetics globally display the same three phases for nucleation , growth and stabilization of oligomers observed at 280 K , and 95% of the final aggregated structures display a dominance of parallel orientation of the -strands ( Figure S1 ) . The main difference between the two temperatures ( Figure 2 ) is in the lag time associated with the nucleation phase: while the average lag time is found to be 13 ns at 280 K , it increased to 56 ns at 300 K , leading to a denser nucleation region on the probability map ( Figure S1 ( b ) ) . Mechanistically , this increase in nucleation time can be explained by the presence of bigger thermal fluctuations that destabilize the metastable aggregates , preventing nucleation . While most simulations at 280 K and 300 K generate a single aggregation event , we observe reversibility for 34% of aggregation events at 280 K against 40% at 300 K . In these cases , such as in the example shown in Figure 4 , monomers undergo a complete aggregation process up to and including the stabilization phase before the reverse reaction takes place , leading to a completely or partially random structure . For some simulations , this reversible transition was even observed to occur a few times during the 100 ns run . The presence of reversibility tells us that even though the free energy barrier for forming a 20-mer oligomer is high , the system is not completely biased towards the formation of an ordered oligomer . Thermal fluctuations for this 20-mer are sufficient to destabilize ordered oligomers on a relatively short time scale , a process that cannot be achieved in all coarse-grained aggregation simulations [16] , [71] but which is crucial in order to describe aggregation kinetics correctly . In this section we present the analysis of the 10 30 ns MD simulations , five at 280 K and five at 300 K , whose configurations are saved every 75 fs in order to describe the details of the kinetics during the final nucleation and full growth process . Because of the tremendous size of the resulting simulation data , we concentrated our analysis on a 10 ns window centered around the drop in energy ( Figure 7 ) . Panel ( a ) represents the average energy taken over all five simulations as a function of time at 280 K . Trajectories are aligned , in time , at the point at which they reach −80 kcal/mol , which is roughly the midpoint in the energy drop for all simulations . Most of the energy drop associated with oligomeric growth , on the order of 600 kcal/mol100 kcal/mol , takes place over 4 ns , in agreement with our earlier observations for a typical aggregation process at 280 K . The relatively small error bars along the energy curve indicate the good reproducibility of the properties over time at 280 K . At 300 K , the growth phase associated with the energy drop , of about 450 kcal/mol200 kcal/mol , also takes on the order of 4 ns ( Figure 7 ( b ) ) , similar to a 280 K energy drop . The standard deviation on the 300 K curve is , however , greater than at 280 K , demonstrating a greater variability associated with larger thermal fluctuations . At both temperatures 280 K and 300 K , aggregation is generally triggered by the formation of a small-sized metastable aggregate , which appears to be stable after a certain lag time . This suggests that we are in the presence of an assembly sequence that can be classified as a nucleated-growth process [24] , [26] , [84] , [86]–[90] , i . e . , that this small metastable aggregate , which we term nucleus , serves as a nucleation center of the aggregation process . The 152 100 ns MD simulations were divided in 3 sets at both 280 K and 300 K and we computed the free energy as a function of aggregate size and secondary structure for those 3 sets of simulations in order to determine the size and amount of secondary structure of the critical nucleus ( Figure 10 ) . Performing this task on different sets of data allows us to have an idea on the order of the fluctuations in the free-energy . At 280 K the nucleus size corresponding to the maximum of free energy is found to be between 4 ( Figure 10 ( a ) - green curve ) and 5 monomers ( Figure 10 ( a ) - red and blue curves ) and between 5 ( Figure 10 ( b ) - red and blue curves ) and 6 monomers ( Figure 10 ( b ) - green curve ) at 300 K . This result is expected since larger thermal fluctuations require a bigger aggregate to survive and lead to growth . The pentameric critical nucleus identified here is also near the critical size estimated by Nelson et al . [33] and by us , in a previous thermodynamic study [39] . As was shown recently [91] , [92] , the critical nucleus size in a finite-size system is systematically overestimated and it is necessary to correct for this artifact . From the classical nucleation theory ( CNT ) , Grossier et al . derive an expression for the total free energy of forming an aggregate of size monomers in an infinitely large system to be [91]: ( 2 ) where is the aggregate size , is the Boltzmann constant , is the temperature , is a dimensionless constant and represents the supersaturation and is the interfacial energy ( or surface tension ) taken to be a constant in the model . Due to our very small system size , a 20-mer , and the low critical nucleus , it is not possible to obtain a good fit to this continuous equation . However , the overestimation correction could explain the slight difference we observe with respect to the experimentally-derived critical nucleus of four monomers . Looking at the free energy barrier of forming a certain amount of secondary structure , we find that a viable nucleus requires between 24 and 28 residues in conformation at 280 K while it requires between 27 and 29 residues in conformation at 300 K ( Figure 10 ( c ) and ( d ) ) . The increase in free energy for 80 residues is due to the finite-size effects of our system . It becomes harder to have 80 residues in -conformation as no more monomers are available to the system to continue growth . Figure 10 ( e ) and ( f ) show the dominant pentamer nucleus structure having such amount of secondary structure at 280 K and 300 K . In both cases , the pentamer seed is partially ordered . In most cases , no more than a dimer is formed beside the nucleus . To assess the microscopic mechanisms involved in the kinetics , we first identify all types of association and dissociation: growth by monomer addition ( and , reversibly , loss by monomer subtraction ) , growth by fusing two oligomers together ( and , reversibly , fragmentation of one oligomer into two smaller oligomers at least 2 monomers in size ) and the direct formation/destruction of oligomers from/into monomers . In this section , we refer to any aggregate bigger than one monomer as an oligomer . It is important to point out that there is a wealth of “monomer addition” models for diverse polymer-forming proteins such as actin [93] , [94] , tubulin [82] , the sickle cell hemoglobin [25] , [74] and amyloid proteins such as A [27] , [95] , [96] , 2-microglobulin [79] and Sup35 [78] . There also exists numerous “oligomer fusion” models for A [9]–[11] , [97] , -synuclein [73] , [97] , [98] , the phosphoglycerate kinase protein [99] , the lysozyme protein [100] and Sup35 [65] , [101] , some of which have observed both processes happening at the same time . Association and dissociation rates were calculated , with our clustering code , every 75 fs over a 10 ns window ( centered around the energy drop ) for the 30 ns simulations and as described in Eq . ( 1 ) . Then , for each time interval , we calculated the total number of events , originating either from monomer addition/loss , from oligomer fusion/fragmentation or from monomersoligomers events across all species such as: ( 3 ) and ( 4 ) where and are the “monomer addition/loss” and “oligomer fusion/fragmentation” + monomersoligomers components of Eq . ( 1 ) . Figure 11 shows the evolution of these two quantities for both association and dissociation events at 280 K ( Figure 11 ( a ) ) and at 300 K ( Figure 11 ( b ) ) . We differentiate the fusion/fragmentation events from the formation/destruction of oligomers ( bigger than dimers ) directly from/into monomers . At both temperatures , the data clearly shows that the assembly mechanism is dominated by “monomer addition/loss” events . Then when nucleation and aggregation happen , we see a notable increase in the amount of monomer events and a trigger of “oligomer fusion/fragmentation” and “monomersoligomers” events . We notice a well-defined increase in the number of “monomer addition/loss” events just before the first “oligomer fusion/fragmentation” events appear . This increase corresponds to the start of nucleation and suggests that once nucleation is triggered and most of the monomers are recruited , they join different sites , or clusters , that will later on fuse together to form a larger oligomer . Later , when the aggregate stops growing in size , we observe no more “single monomer” or “monomersoligomers” events and observe , in some cases , the presence of only fusion and fragmentation of oligomers ( Figure 11 ( a ) ) . This means that further rearrangements in the structure during the stabilization phase are accomplished mainly through oligomer-involving events , if any . We presented here a detailed study of the onset of amyloid aggregation for 20-mers of GNNQQNY . Using molecular dynamics with the OPEP coarse-grained force field , we show that nucleation of this small amyloid peptide is dominated by monomer addition/loss events , with very small contributions from larger oligomers , following closely the classical nucleation theory . It is then meaningful to extract a critical nucleus , that can be obtained from the calculation of the free-energy as a function of nucleus size . We find that , at 280 K , this critical size is between 4 and 5 monomers , while it is between 5 and 6 at 300 K , in good agreement with the experimental estimate of 4 monomers [33] , especially when taking into account the finite-size bias that tends to overestimate the size of the critical nucleus [91] , [92] . Correspondence with CNT stops there , however , as the kinetic process associated with aggregation and growth differs in two majors from this theory . First , while most of the structural organization takes place during the 4 ns growth process , aggregates continue to mature by collective motions , slowly dropping in energy as hydrogen bonds and -sheet content evolve . Second , nucleation does not lead to a single structure , but shows clear polymorphism with a distribution of assemblies that can be classified into two distinct categories: a compact oligomer made of a number of relatively short -sheets , typically three , and a more extended fibril-compatible two-sheet structure . These structures represent well-separated local basins and the only way to move between them , in our simulations , was through a complete dissociation and reassociation of the monomers . The well-defined polymorphic nature of GNNQQNY is in line with experimental and numerical observations in other amyloid sequences , such as amyloid- . It was shown there that the protein could adopt multiple fibrillar structures [18] , [102] , but also off-pathway -barrel organizations that would be responsible for at least part of the toxicity . [103] For GNNQQNY , the two polymorph families observed here are close enough that they should lead to different fibrillar structures rather than on and off-pathway organizations . Only simulations with a larger number of peptides will be able to tell . How much of these results can be applied to experimental studies of GNNQQNY ? A previous stability study of the structures predicted with OPEP using explicit SPC solvent and all-atom GROMOS96 showed that our simulations are realistic , except for the most extended structures [39] . If the growth time is not directly extendable to all-atom systems , the thermodynamics and , therefore , the critical nucleus size but also the polymorphism , which is a signature of amyloid aggregates , should be valid . Our results suggest that the specific shape , out of a family of structures , is selected very early on and that moving from one to another requires going over a very high barrier , high enough that it was never observed in our simulations , the preferred being going first through a complete dissociation . Such behavior could change with larger aggregates , and the direct rearrangement become more favorable than complete dissociation . Only further work , on larger systems , will show whether new families of structures are possible for GNNQQNY and if the CNT applies when more monomers are in play . Our results on the 20-mer of GNNQQNY are at least compatible with experiments and offer a number of insights into the onset of aggregation and polymorphism for small amyloid peptides . | Protein aggregation plays an important pathological role in numerous neurodegenerative diseases such as Alzheimer's , Parkinson's , Creutzfeldt-Jakob , the Prion disease and diabetes mellitus . In most cases , misfolded proteins are involved and aggregate irreversibly to form highly ordered insoluble macrostructures , called amyloid fibrils , which deposit in the brain . Studies have revealed that all proteins are capable of forming amyloid fibrils that all share common structural features and therefore aggregation mechanisms . The toxicity of amyloid aggregates is however not attributed to the fibrils themselves but rather to smaller more disordered aggregates , oligomers , forming parallel to or prior to fibrils . Understanding the assembly process of these amyloid oligomers is key to understanding their toxicity mechanism in order to devise a possible treatment strategy targeting these toxic aggregates . Our approach here is to computationally study the aggregation dynamics of a 20-mer of an amyloid peptide GNNQQNY from a prion protein . Our findings suggest that the assembly is a spontaneous process that can be described as a complex nucleation and growth mechanism and which can lead to two classes of morphologies for the aggregates , one of which resembles a protofibril-like structure . Such numerical studies are crucial to understanding the details of fast biological processes and complement well experimental studies . | [
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results/Discussion"
] | [
"physics",
"biochemistry",
"biochemistry",
"simulations",
"proteins",
"biophysics",
"simulations",
"chemistry",
"biology",
"computational",
"chemistry",
"biophysics",
"thermodynamics",
"physical",
"chemistry"
] | 2012 | Kinetics of Amyloid Aggregation: A Study of the GNNQQNY Prion Sequence |
Vector-borne transmission of Trypanosoma cruzi is seen exclusively in the Americas where an estimated 8 million people are infected with the parasite . Significant research in southern Peru has been conducted to understand T . cruzi infection and vector control , however , much less is known about the burden of infection and epidemiology in northern Peru . A cross-sectional study was conducted to estimate the seroprevalence of T . cruzi infection in humans ( n=611 ) and domestic animals [dogs ( n=106 ) and guinea pigs ( n=206 ) ] in communities of Cutervo Province , Peru . Sampling and diagnostic strategies differed according to species . An entomological household study ( n=208 ) was conducted to identify the triatomine burden and species composition , as well as the prevalence of T . cruzi in vectors . Electrocardiograms ( EKG ) were performed on a subset of participants ( n=90 T . cruzi infected participants and 170 age and sex-matched controls ) . The seroprevalence of T . cruzi among humans , dogs , and guinea pigs was 14 . 9% ( 95% CI: 12 . 2 – 18 . 0% ) , 19 . 8% ( 95% CI: 12 . 7- 28 . 7% ) and 3 . 3% ( 95% CI: 1 . 4 – 6 . 9% ) respectively . In one community , the prevalence of T . cruzi infection was 17 . 2% ( 95% CI: 9 . 6 - 24 . 7% ) among participants < 15 years , suggesting recent transmission . Increasing age , positive triatomines in a participant's house , and ownership of a T . cruzi positive guinea pig were independent correlates of T . cruzi infection . Only one species of triatomine was found , Panstrongylus lignarius , formerly P . herreri . Approximately forty percent ( 39 . 9% , 95% CI: 33 . 2 - 46 . 9% ) of surveyed households were infested with this vector and 14 . 9% ( 95% CI: 10 . 4 - 20 . 5% ) had at least one triatomine positive for T . cruzi . The cardiac abnormality of right bundle branch block was rare , but only identified in seropositive individuals . Our research documents a substantial prevalence of T . cruzi infection in Cutervo and highlights a need for greater attention and vector control efforts in northern Peru .
Chagas disease is caused by the protozoan parasite Trypanosoma cruzi , and is primarily transmitted by triatomine vectors . Chagas disease is endemic to poor rural regions of Central and South America and is responsible for the largest public health burden of any parasitic infection in the Western Hemisphere [1] . An estimated 8 million people are infected with T . cruzi and millions more are at risk [2] . Trypanosoma cruzi is carried in the gut of the triatomine vector and transmitted through the insect’s feces . While the vector-borne route predominates , oral transmission , congenital transmission and infection through blood transfusion and organ transplantation also occur . Acute Chagas disease is asymptomatic or oligosymptomatic and if clinical manifests as fever and fatigue . The majority of individuals will survive this acute phase without treatment or even evaluation [2] . Approximately 20–30% of chronic infections advance to the chronic symptomatic form of the disease , characterized by cardiac , gastrointestinal or neurologic disease [2–4] . Heart disease is the most common clinical manifestation of chronic Chagas disease [2] . In Peru gastrointestinal and neurologic forms are extremely rare . Chagas heart disease is an irreversible fibrosing inflammatory cardiomyopathy characterized by conduction abnormalities , such as right bundle branch block , left anterior fascicular block , ventricular extra systoles and ventricular tachycardia [2] . As the disease progresses , manifestations include sinus node dysfunction , atrioventricular blocks , dilated cardiomyopathy and thromboemboli [2] . Chagas disease is understudied in northern Peru and little is known about the epidemiology of T . cruzi in the region [5] . Panstrongylus lignarius ( synonymous with Panstrongylus herreri ) [6] is known as the 'main domestic vector' of Chagas disease in northern Peru , specifically in the Marañon Valley , yet several other species have been described in northern Peru [7] . We conducted a series of cross-sectional surveys in several communities of Cutervo Province , in the Cajamarca region of Peru . The study aims were to ( 1 ) describe the seroprevalence of T . cruzi in humans , domestic dogs , and guinea pigs; ( 2 ) to describe the species and prevalence of vectors overall and with T . cruzi; ( 3 ) identify and characterize risk factors of T . cruzi infection in humans; and ( 4 ) characterize the extent and scope of cardiac abnormalities associated with T . cruzi infection in humans .
This study was conducted in December 2009 to October 2010 , in Cutervo Province of Cajamarca , Peru . Cutervo is located in the Huancabamba River Valley , near the Marañon Valley of the Andes ( altitude 850–1700 m ) , which ultimately drains into the Amazon River Basin ( Fig 1 ) . Six communities ( Campo Florido , Casa Blanca , La Esperanza , Pindoc , Nuevo Guayaquil and Rumiaco ) were included in the study based on government documented triatomine infestation and clinical reports of people with Chagas disease . All communities were located within an aerial distance of 15 km . They share the same ecoregion , known as the Peruvian Yungas or Selva Alta , which is characterized by neotropical forest , steep slopes and narrow valleys . Road infrastructure and access to these communities , however , was variable: Casa Blanca and La Esperanza were connected to the local highway via a gravel road; the community of Campo Florido , however , could only be reached by a poorly maintained dirt road that was impassable for several months during the rainy season . All six communities were included in the human serological survey and the electrocardiogram ( EKG ) study . A subset of four communities was sampled for domestic dog serology and for domiciliary and peridomestic vectors ( Campo Florido , Casa Blanca , La Esperanza , and Pindoc ) and one community ( Campo Florido ) was evaluated for guinea pig serology . Trained study nurses recruited participants both at the local health posts during a community-wide serological testing campaign and at people’s homes during house-to-house visits . All study participants provided informed written consent and a parent or guardian provided written consent on behalf of minors . A fingerprint , as a proxy of a written signature , was an acceptable alternative for individuals unable to write . With participant consent , all human seropositive individuals were referred to the Ministry of Health . Animal owners provided written consent for the participation of domestic animals . The methods of this study complied with federal and institutional regulations . The Institutional Review Board ( IRB ) of the Asociación Benéfica Proyectos en Informática , Salud , Medicina y Agricultura ( Lima , Peru ) approved the protocol ( file# CE0886 . 09 ) as did the IRB of the University of Pennsylvania ( file# 812713 ) . The animal protocol was reviewed and approved by the Institutional Animal Care and Use Committee of the Universidad Peruana Cayetano Heredia ( UPCH ) ( file# 52186 ) as well as the University of Pennsylvania ( file# 803364 ) . The animal protocol adhered to standards outlined by the National Research Council's Guide for the Care and Use of Laboratory Animals [8] . All residents of the six communities >2 years of age were eligible to participate in the serological survey . The age and sex of both survey participants and non-participants were recorded . Blood samples were collected from each participant , stored at 4°C and were transported on the same day to the field laboratory . Blood was separated by centrifugation and stored at -20°C . Serologic analysis was completed at the Universidad Peruana Cayetano Heredia Laboratory of Infectious Diseases ( LID-UPCH ) . All human serum specimens were tested by three assays: the Chagatek T . cruzi lysate ELISA ( bioMerieux , Marcy l’Etoile , France ) , the Wiener Recombinant ELISA ( Wiener , Rosario Argentina ) , and the trypomastigote excreted-secreted antigen ( TESA ) immunoblot [9] . T . cruzi infection in humans was considered confirmed if two or more tests yielded positive results [10] . Specimens with one or no tests positive were considered seronegative . The Chagatek and Wiener ELISA were completed according to manufacturer’s instructions and the threshold for positive results was 0 . 10 optical density ( OD ) units above the mean absorbance of two negative control specimens included on each plate . The TESA assay was completed according to specifications in Umezawa et al [9] . To understand the extent and scope of cardiac abnormalities in these communities and their association with chronic T . cruzi infection , an electrocardiographic study was conducted on 90 infected individuals and 170 controls . All participants of the serological survey were invited to the EKG study at the time of the serological survey recruitment . Controls were matched based on age and gender . A majority of infected individuals ( 80 ) were matched with two negative controls , and the remaining individuals ( 10 ) were matched with one . At the local health posts , participants underwent a structured medical history , a non-invasive physical exam ( PE ) by a study physician , and a 12-lead EKG in the 30° inclined position ( portable Welch Allyn CP100 ) . Parents were encouraged to be present for their children’s examinations . The duration of PEs and EKGs ranged from 15–30 minutes and all EKG data was subsequently read and coded by a board certified cardiologist . An EKG was considered to have abnormalities consistent with Chagas cardiomyopathy if one or more of the following were present: atrial fibrillation/flutter , junctional rhythm , ventricular tachycardia ( sustained or non-sustained ) , ventricular extrasystoles ( multiform , paired , or salvos ) , sinus node dysfunction , sinus bradycardia ( <50 bpm ) , second degree AV block ( type I or type II ) , third degree AV block , AV disassociation , left or right bundle branch block ( LBBB , RBBB ) , left anterior or left posterior fascicular block , or trifascicular block [2 , 11 , 12] . Incomplete RBBB was not considered consistent with Chagas cardiomyopathy . Four communities were evaluated in the household entomological survey: Campo Florido , Casa Blanca , La Esperanza , and Pindoc . With household member consent , two trained entomologic collectors , aided by a tetramethrin flushing-out agent ( Sapolio , Mata Moscas ) , searched domestic and peridomiciliary habitats including domestic animal enclosures for a total of one half-hour ( one person-hour ) . Captured triatomines were stored at 4°C until processing at the field laboratory and then examined for the presence of T . cruzi , following standard procedures [13 , 14] . Vector species was determined based on morphology . The species , quantity , sex and life stage of triatomine vectors was documented . Due to the specimen quality once the triatomines arrived at the field laboratory , not all of the collected triatomines were evaluated for sex , development stage , and intestinal contents . Second through fifth instar triatomines were evaluated for trypanosomatids . For each household the wall and roof construction material were documented; data on the total number and type of domestic animals were reported by the household representative . A serological survey of domestic animals was performed to document T . cruzi transmission through potential reservoir species . Domestic dogs ( Canis lupus familiaris ) from Campo Florido , Casa Blanca , La Esperanza , and Pindoc were evaluated , as were Guinea Pigs ( Cavia porcellus ) from Campo Florido . A household level census of all domestic species was conducted to estimate the domestic animal population . Canine age was reported by owners , and guinea pig age was approximated based on measured body length . Canine and guinea pig blood samples were collected by a veterinarian or trained phlebotomist , and , stray , pregnant , notably sick , and/or juvenile animals ( dogs <1 mo , and guinea pigs < 20 cm in length ) were not sampled . Transport and processing were identical to that of human blood samples , however , domestic animal serostatus was determined based on an enzyme-linked immunosorbent assay ( ELISA ) . At LID-UPCH , the domestic animal sera were tested for the presence of anti T . cruzi antibodies by epimastigote alkaline extract ( EAE ) ELISA using Arequipa strain epimastigote extracts ( 2 . 5 ug/ mL ) [15] . Each plate contained seven negative and one positive control . The positive control consisted of sera from either a Y strain experimentally infected guinea pig or from an Arequipa strain naturally infected dog . The sample was positive if the OD was greater than three standard deviations above the mean plate OD . A subset of canine and guinea pig samples ( n = 103 and n = 31 , respectively ) was evaluated by TESA-blot [9] . Descriptive statistics were first used to characterize the human study population and compare demographic information to the general population from which they were selected . The infection prevalence along with exact binomial 95% confidence intervals was ascertained for humans , domestic animals and triatomine vectors . Differences in EKG findings by T . cruzi serostatus were evaluated by chi-squared test . Among humans , differences in the frequency and distribution of demographic and household level variables by T . cruzi serostatus were evaluated by chi-squared test or nonparametric rank tests such as Wilcoxon ranksum . Vector count data was modeled using a negative binomial regression model to compare collections across communities . Adobe-housing material was used as the predictor of excess zeroes . A Vuong test was used to determine whether a zero-inflated negative binomial regression model was a better fit than a negative binomial regression model . Akaike’s information criterion ( AIC ) was used to determine that the zero-inflated negative binomial model was a better fit than the zero-inflated Poisson model . Using the model coefficient , an expected difference in vector count relative to the baseline community was calculated . Through univariate analysis , odds ratios were estimated for the association of demographic variables ( age and sex ) and household level variables ( presence of one or more vector , positive vector , guinea pig , positive guinea pig , dog , positive dog , or walls made of adobe ) with T . cruzi seropositivity . A mixed-effects modeling approach was used , clustered by household and using an exchangeable correlation structure and logit link . Variables that have previously been shown to have an association with the outcome of interest were initially included in a multivariable logistic mixed-effects model . Using an AICc selection process , a model was constructed that included community as a fixed-effect to adjust for heterogeneity in seropositivity between communities . Because certain combinations of variables in the model resulted in a decreased sample size , the researchers ensured that the model maintained a minimum sample size of 200 subjects . It was assumed that zero vectors were present if a house was entered for data collection and the number of vectors collected was not recorded . Cohen’s Kappa analysis was conducted to test the percent of agreement between the animal serologic diagnostic methods . Statistical tests were conducted using R 3 . 1 . 3 [16] , Stata 11 . 2 , and Stata 13 ( StatCorp ) .
The census enumerated 1134 people in six communities ( Table 1 ) . Of the 1093 residents older than 2 years , 612 ( 56 . 0% ) participated in the serological survey . There were more female than male participants ( 58 . 5% versus 41 . 5% ) and participants were younger than non-participants ( mean age = 27 . 4 versus 28 . 2 years ) . Ninety-one participants ( 14 . 9% , 95% CI: 12 . 2–18 . 0% ) had positive results by at least two serological assays . One participant had inconclusive results by both ELISAs and negative results by TESA-blot . His infection status therefore remained unresolved and his data were excluded from further analysis . The total study population was therefore 611 ( S1 Table ) . Females were more likely to have T . cruzi infection than males ( 16 . 2% versus 13 . 0% ) . The seropositive population was older than those without infection ( mean age 37 . 8 versus 25 . 6 years old ) . Overall and age-specific seroprevalence varied across the six communities ( Table 2 ) . In Pindoc , Nuevo Guayaquil and Casa Blanca seroprevalence increased with age . However , this trend was not seen in La Esperanza , Campo Florido , and Rumiaco ( Fig 2 and Table 2 ) . Among participants <15 years old seroprevalence differed significantly between communities ( ANOVA p < 0 . 02 ) , with a particularly high seroprevalence in Campo Florido ( 17 . 2% , 95% CI: 9 . 6–24 . 7% ) . Ninety T . cruzi infected and 170 uninfected participants underwent EKGs . Both adults and children >2yo enrolled in the EKG survey , and there were more female than male matched groups ( S2 Table ) . RBBB was rare , yet it was diagnosed in 2/90 seropositive participants and none of the 170 seronegative controls . Evaluation of the aforementioned Chagas associated EKG abnormalities showed no significant difference between seropositive and seronegative participants ( 4 . 4% of the seropositives had at least one of the EKG abnormalities versus 1 . 2% of seronegatives ) ( S3 Table ) . Vector searches were conducted in 208 ( 75 . 1% ) of the 277 houses in four communities . The search of these 208 houses was comprised of 1130 spaces: 858 rooms and 272 animal enclosures . A majority of rooms ( 551/858 ) were made of adobe ( 64 . 2% , 95% CI: 60 . 9–67 . 4% ) . Other less common room construction materials included brick , stone , plaster , wood , branches , and reed . The majority of roofs ( 560/858 ) were made of calamina , a corrugated roofing material of metal or plastic ( 65 . 3% , 95% CI: 62 . 0–68 . 5% ) . Other less common roof materials included wood or reed . Approximately half of the animal enclosures were outside of the household ( 128/272 ) and categorized as peridomestic ( 47 . 1% , 95% CI: 41 . 0–53 . 2% ) . Animal enclosures were most frequently made of adobe ( 113/272 ) and/or wood ( 100/272 ) ( 41 . 5% , 95% CI: 35 . 6–47 . 7%; and 36 . 8% , 95% CI: 31 . 0–42 . 8% respectively ) . Owned domestic animals included guinea pigs , dogs , cats , chickens , turkeys , geese , ducks , pigs , sheep and cows . Some of these animals were classified as intradomiciliary and others as peridomiciliary . The most common intradomiciliary animals were guinea pigs ( range 0–42 ) with at least one residing in 99 households ( 48 . 1% , 95% CI: 41 . 0–55 . 1% ) . The most common peridomicilliary animals were chickens ( range 0–93 ) , dogs ( range 0–6 ) and pigs ( range 0–11 ) with at least one owned by 113 ( 74 . 3% , 95% CI: 66 . 7–81 . 1% ) , 82 ( 53 . 9% , 95% CI: 45 . 7–62 . 0 ) , and 76 ( 50 . 0% , 95% CI: 41 . 8–58 . 2% ) households respectively . All vectors collected were identified as one species: Panstrongylus lignarius . Eighty-three houses ( 39 . 9% , 95% CI: 33 . 2–46 . 9% ) were infested , and 31 houses ( 14 . 9% , 95% CI: 10 . 4–20 . 5% ) had at least one T . cruzi-infected vector . Triatomines were more commonly found in rooms than animal enclosures , 105/858 rooms ( 12 . 2% , 95% CI: 10 . 1–14 . 6% ) and 11/272 animal enclosures ( 4 . 0% , 95% CI: 2 . 0–7 . 1% ) had at least one vector present . Triatomines were collected in kitchens , eating rooms , bedrooms , empty rooms , and storage rooms , however , of the 116 spaces where triatomines were found , 59/116 ( 50 . 9% , 95% CI: 41 . 4–60 . 2% ) were bedrooms and 39/116 ( 33 . 6% , 95% CI: 25 . 1–43 . 0% ) were kitchens . All five nymphal stages and both sexes were found in both rooms and animal enclosures , demonstrating colonization ( Table 3 ) . In total , there were 1963 triatomines collected . The intestinal contents of 1625 triatomines were evaluated and 315 of those were positive for T . cruzi ( 19 . 4% , 95% CI: 17 . 5–21 . 4% ) . No other trypanosomatids were identified . A median of 0 and a mean of 10 triatomines were found per household ( min 0 , max 236 ) . A zero-inflated negative binomial ( ZINB ) regression model examining the total household number of triatomines showed that Pindoc was significantly different from the other three communities . Pindoc had a coefficient of -1 . 66 ( 95% CI: -2 . 9–-0 . 4 , z = -2 . 64 , p<0 . 01 ) , and an expected vector count of 0 . 19 relative to the reference community of Casa Blanca . The estimated household numbers of triatomines in La Esperanza and Campo Florido were not significantly different from Casa Blanca . A similar ZINB regression model was run examining the total household number of T . cruzi positive triatomines . Pindoc , the community where no positive vectors were captured , was found to be different from the reference community , yet there was no difference in the estimated density of positive vectors in La Esperanza and Campo Florido compared to Casa Blanca ( Fig 3 ) . The number of infected vectors showed positive correlations with the number of T . cruzi-infected dogs overall and in Campo Florido ( ρ = 0 . 31 , p<0 . 02; and ρ = 0 . 72 , p < 0 . 01 respectively ) . There was a similar positive correlation in T . cruzi-infected guinea pigs ( ρ = 0 . 84 , p <0 . 01 ) . The serological survey included 108 dogs ( 75 . 5% ) and 207 guinea pigs ( 43 . 9% ) . Two dogs and one guinea pig were removed from the study due to missing age and size data , respectively . Study dogs had a mean age of 1 . 9 years ( min 1 mo , max 15 yr ) and guinea pig average length was 25 . 5 cm ( min 20 cm , max 32 cm ) . Based on EAE ELISA results , 21 dogs ( 19 . 8%; 95% CI: 12 . 7–28 . 7% ) and 7 guinea pigs ( 3 . 4%; 95% CI: 1 . 4–6 . 9% ) were positive for T . cruzi antibodies . There was a good agreement between ELISA and TESA-blot assays in canines ( K = 0 . 66 , 90 . 3% agreement , p < 0 . 01 ) and in guinea pigs ( K = 0 . 76 , 90 . 3% agreement p < 0 . 01 ) . In univariate analyses , risk factors for T . cruzi infection included older age and presence of infected triatomines in the house ( Table 4 ) . Owning a T . cruzi positive guinea pig showed borderline significance as a risk factor . In the multivariable model , only the presence of T . cruzi infected triatomines remained statistically significant once adjusted for community ( p<0 . 01 ) . People from 155/208 households in the entomological survey also participated in the human serosurvey and only these participants with corresponding household data were included in the multivariable model ( 477/611 ) . Consequently , the final multivariable model included 477 observations from 155 households ( Table 5 ) . A typical individual in a given community had 6 . 1 greater odds of testing positive for T . cruzi when living in the presence of T . cruzi infected triatomines compared to a typical individual in the same community without positive infestation ( 95% CI: 1 . 6–22 . 6 ) .
Our data show that this often-overlooked region in northern Peru has a significant Chagas disease burden and warrants additional investigation and control measures . Although Chagas disease has been documented within the range of P . lignarius in northern Peru [17] , very few studies to date have examined the extent of T . cruzi infection in humans and animals and its relationship to this vector . Evidence shows a high prevalence of T . cruzi infection , 14 . 9% , in human residents of these six rural communities in northern Peru . Human seroprevalence in this region had previously been reported between 1–5% [7 , 18–21] . In southern Peru , the human seroprevalence of T . cruzi has been documented at levels ranging from 1 . 4 to 13 . 4% in urban , periurban and rural sites [22–28] . This study illustrates that secondary vector species , such as P . lignarius , play an important role in the transmission of T . cruzi and are responsible for a significant burden of Chagas disease . Like other studies in endemic areas , our serological survey showed an increase in human seroprevalence with age [22] . Since infection is lifelong , in the absence of effective treatment , this pattern represents cumulative incidence over the residents’ lifetimes . An unusual pattern was seen in Campo Florido , Pindoc and La Esperanza . In Campo Florido in particular , the seroprevalence in children and adolescents was notably elevated , as high as 40 . 5% ( 95% CI: 24 . 8–57 . 9% ) between 11 and 20 year olds . This finding does not appear to be an aberration due to small sample size , as more than 100 residents 20 or younger were tested . Rather , it appears to show both recent transmission and possibly higher risk of exposure in younger individuals . A similar pattern was seen in communities on the outskirts of Arequipa , where a mathematical model estimated that transmission began less than 20 years earlier [22 , 25] . One explanation for apparent recent transmission in Campo Florido is that this community never received the household insecticide application that occurred in the other five communities . According to regional governmental documentation and communications with community leaders , household residual insecticide application was carried out in Casa Blanca , La Esperanza , Pindoc , Rumiaco and Nuevo Guayaquil to reduce malaria and Bartonellosis , two vector-borne diseases that affect the region . Several insecticide treatments were undertaken at different times in different communities over the 10–15 years preceding the study , with the most recent applications taking place in Cutervo Province in 2007 . Insecticide applications employed several synthetic pyrethroid compounds and may have sufficiently reduced triatomine populations to interrupt transmission over recent years . Triatomine reinfestation post spraying is the likely reason that vector density modeling showed no difference in the prevalence of household triatomines or T . cruzi infected triatomines in Campo Florido or La Esperanza compared to Casa Blanca . Clinically , the progression to cardiac disease is the most important determinant of prognosis in patients with Chagas disease [11] . In this study , the conduction abnormality of a right bundle branch block , while rare , was found to have an association with T . cruzi serostatus , similar to findings across the Americas [2 , 11 , 29] . The presence of a right bundle-branch block alone has been associated with an increased risk of mortality in T . cruzi positive individuals , as high as a seven-fold increase in Maguire et al [30] . Despite considerable research to understand domestic animals’ roles in maintaining and augmenting T . cruzi infection , domestic species’ infection rates have great geographic variability and many questions still remain [20 , 31–36] . Domestic dogs are believed to be important reservoirs of the parasite , however , depending on local circumstances , dog ownership may or may not increase risk of infection [37–41] . Data from our study does not implicate dog ownership for increasing T . cruzi risk for their owners . These dogs may serve as parasite reservoirs post-insecticide spraying , however , and may contribute to the reestablishment of T . cruzi in vector populations . Serial sampling of canine serology with concurrent entomologic data before , during and after insecticide treatments may give insight into their roles as reservoirs . Guinea pigs have historically been considered as potential T . cruzi reservoirs [7 , 20 , 36 , 42]; yet , evidence from this study does not implicate guinea pig ownership alone as a risk factor of human infection . Serological testing , however , may not be a reliable diagnostic in guinea pigs . Castro-Sesquen et al illustrate a slow rise of guinea pig immunoglobulin , which is only consistently detectable 40 days post T . cruzi inoculation . Considering the short life span of a domesticated guinea pig ( they are commonly slaughtered for food by 3 months of age ) , there exists only a narrow time window when antibody levels can be sufficiently detectable even if infection occurred at a very young age [43] . Sixteen triatomine species have been reported in northern Peru , nine of which are thought to have potential to be significant vectors for T . cruzi [7 , 44 , 45] . While the majority of Amazonian triatomines are reported to be sylvatic [46] , three species in northern Peru are known to be synanthropic , meaning ecologically associated with humans: Panstronylus lignarius , Rhodnius ecuadoriensis , and Triatoma dimidata [7] . Only one species was identified in our survey , Panstrongylus lignarius ( syn . P . herreri ) [6] . This vector has previously been called a 'domestic pest' in the Marañon River Valley [18] , but has also been documented as occupying niches in sylvatic ecotopes such as bird nests in Ecuador [5] . The species Triatoma carrioni , Rhodnius ecuadoriensis , and Panstrongylus geniculatus , which have also been documented in Cutervo Province , were not found in this study [7] . Triatoma infestans , the principal vector of southern Peru responsible for transmission of T . cruzi , has never been documented north of Lima and its surrounding communities [7] . In our entomological survey , Panstrongylus lignarius vectors in all five nymphal stages as well as adults were found , providing evidence that a complete life cycle within domestic and peridomestic habitats is possible . In Peru , the role of extradomiciliary triatomines in T . cruzi transmission remains poorly described , though is likely similar to that in geographically proximate regions of Ecuador [47] . For vector species that are capable of inhabiting both wild and domestic ecotopes , such as P . lignarius , reinfestation after insecticide treatment is expected and long-lasting surveillance and focal control may be necessary to permanently halt transmission . There are several limitations to our study . The serological analyses of humans and domestic animals are not directly comparable , as different sampling and diagnostic strategies were employed . The criteria for T . cruzi positivity in the human serosurvey was determined by a minimum of two out of three positive assays , whereas positivity in the domestic animal serosurveys was determined by the outcome of one ELISA assay . While there is a potential for serological misclassification in both the human and animal surveys , the misclassification rate in the human serosurvey is low on account of the three assay approach . Since vector-borne transmission was the primary focus of this study , children <2yo were excluded from the study , and consequently the role of congenital transmission was not examined . The low prevalence of infection among guinea pigs might suggest they are less relevant to T . cruzi transmission than dogs and other hosts . However , the life history of guinea pigs raised for consumption in Peru , and the time period of development of their immunological response to T . cruzi infection may obscure the interpretation of our serological tests . The prevalence of triatomine vectors and the prevalence of T . cruzi in this vector population are likely conservative estimates . The flushing out method ( one person-hour ) has a moderate sensitivity ( 76% ) but has the potential to be higher in areas with higher vector density [48 , 49] . The timed search approach to vector detection could have been improved with the use of traps . For parasite detection , diagnostic sensitivity for T . cruzi can vary according to vector species [50 , 51] . While limited diagnostic information exists for the sensitivity in Panstrongylus species specifically , in other genera , molecular techniques can offer greater sensitivity [52 , 53] . Lastly , it is difficult to ascertain temporal sequence of transmission between domestic animals , vectors and humans in a cross sectional survey . The prevalence of T . cruzi infection identified in these six communities of Cutervo Province , is equal to or higher than levels documented elsewhere in Peru , yet this region has few control measures in place , none of which targets T . cruzi and its vectors specifically . Furthermore , notably high T . cruzi seroprevalence was detected in the children and adolescents of Campo Florido . We also documented cardiac abnormalities in T . cruzi seropositive participants illustrating the potential health impacts of this protozoan to the people it infects . Prevention of Chagas related morbidity and mortality in this region may be possible with greater attention to T . cruzi infection , its vectors , and public health control strategies . | Chagas disease causes significant morbidity and mortality throughout Central and South America . The epidemiology and control of this disease is subject to unique regional particularities , including the behavior and ecology of the local insect vector species . Significant resources have been allocated towards research and control efforts in southern Peru , yet very little is known about the prevalence and epidemiology of Trypanosoma cruzi in northern Peru . Our study highlights significant T . cruzi infection in northern Peru and is one of the first to document substantial transmission by the insect Panstrongylus lignarius . Our results illustrate major gaps in knowledge and the need for public health interventions targeted at Chagas disease in the region of Cutervo Province of northern Peru . | [
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] | [] | 2015 | Prevalence and Transmission of Trypanosoma cruzi in People of Rural Communities of the High Jungle of Northern Peru |
Many mosquito species are naturally polymorphic for their abilities to transmit parasites , a feature which is of great interest for controlling vector-borne disease . Aedes aegypti , the primary vector of dengue and yellow fever and a laboratory model for studying lymphatic filariasis , is genetically variable for its capacity to harbor the filarial nematode Brugia malayi . The genome of Ae . aegypti is large and repetitive , making genome resequencing difficult and expensive . We designed exome captures to target protein-coding regions of the genome , and used association mapping in a wild Kenyan population to identify a single , dominant , sex-linked locus underlying resistance . This falls in a region of the genome where a resistance locus was previously mapped in a line established in 1936 , suggesting that this polymorphism has been maintained in the wild for the at least 80 years . We then crossed resistant and susceptible mosquitoes to place both alleles of the gene into a common genetic background , and used RNA-seq to measure the effect of this locus on gene expression . We found evidence for Toll , IMD , and JAK-STAT pathway activity in response to early stages of B . malayi infection when the parasites are beginning to die in the resistant genotype . We also found that resistant mosquitoes express anti-microbial peptides at the time of parasite-killing , and that this expression is suppressed in susceptible mosquitoes . Together , we have found that a single resistance locus leads to a higher immune response in resistant mosquitoes , and we identify genes in this region that may be responsible for this trait .
The rate at which parasites are transmitted by mosquitoes is an important determinant of the prevalence of vector-borne diseases in human populations . Alongside factors like the number of mosquitoes and their biting preferences , the rate of transmission depends on the ability of mosquitoes to acquire the parasite when feeding on an infected person and subsequently transmit it . This is referred to as their vector competence , and is affected by both environmental and genetic factors [1 , 2] . Even within a population of a single mosquito species there can be tremendous genetic variation in vector competence , often as a result of differences in the immune response of the mosquito to the parasites they are vectoring [2] . For example , variation exists in susceptibility of Anopheles gambiae to the malaria parasite Plasmodium falciparum [3 , 4] and in Aedes aegypti to dengue and filarial nematodes [5 , 6] . This has attracted much attention as it could one day lead to be better disease control by manipulating mosquito populations to reduce vector competence . For example , field trials are underway that are releasing Ae . aegypti mosquitoes carrying the bacterial symbiont Wolbachia , which reduces the mosquitoes’ ability to transmit dengue virus [7] . The tropical disease lymphatic filariasis , or elephantiasis , is a leading cause of morbidity and disability worldwide , with especially high parasite burdens in Africa and south and south-east Asia [8] . It is estimated to affect 120 million people worldwide , and symptoms include lymphedema and swelling of the extremities [9] . In humans , the disease is caused by the filarial nematodes Wuchereria bancrofti , Brugia malayi and Brugia timori , and is vectored by a range of mosquitoes , including species of Culex , Mansonia , Anopheles and Aedes [9] . W . bancrofti is the major cause of filariasis worldwide , leading to 90% of the cases of lymphatic filariasis , and Brugia species , which are only found in Asia , cause the remaining 10% [8] . B . malayi is the main laboratory model for studying lymphatic filariasis , and it grows readily in some strains of the mosquito Ae . aegypti . Despite having overlapping ranges , Ae . aegypti does not naturally vector any of the nematodes that cause lymphatic filariasis in humans . It is however a natural vector of Dirofilaria , which causes filariasis in dogs [10] . B . malayi , along with other filarial nematodes , are heteroxenous , requiring both a vertebrate host and a mosquito vector for their life cycle [8 , 9] . Humans , cats , and monkeys can all serve as vertebrate hosts for B . malayi [10] . Male and female worms reproduce sexually in the vertebrate , producing microfilariae which circulate in the bloodstream and are ingested by mosquitoes during blood feeding . After penetrating the mosquito midgut , the filarial nematodes develop inside various tissues within the mosquito . In the case of B . malayi , the microfilariae migrate to the thoracic muscles of the mosquito , where they undergo successive molts until they become L3 larvae [11] . They then migrate to the mosquito proboscis , where they are transferred to the vertebrate host during blood feeding . Beginning in the 1960’s , mosquito strains and species have been identified that are naturally refractory ( resistant ) to infection by filarial nematodes [12] . Proposed mechanisms of resistance include reduced ingestion of parasites , physical killing of parasites in the foregut , barriers to penetration of the midgut , and hemolymph factors that kill the parasite in the thoracic cavity and lead to melanotic encapsulation [13] . Some species such as Armigeres subalbatus , a natural vector of Brugia pahangi , are completely refractory to infection by B . malayi while being highly susceptible to B . pahangi [14] . Others , such as Ae . aegypti , are polymorphic within species for resistance [15] . In laboratory lines of Ae . aegypti , genetic variation in resistance to B . malayi has a simple genetic basis , and is primarily determined by a single dominant locus on the first chromosome [16] . This genetic resistance extends to some other species of nematodes , such as B . pahangi and W . bancrofti , but not to Dirofilaria , for which Ae . aegypti is a natural vector [16] . In this mosquito , sex is also determined by a region on the first chromosome , and the resistance locus is tightly linked to the sex-determining region [17–19] . The immune responses of Ae . aegypti have been extensively studied , but it remains unknown which factors are important in killing filarial nematodes and whether genetic differences in susceptibility are caused by differences in immune responses . Despite the mechanisms being unclear , the mosquito immune response does appear to control filarial nematodes . Fewer parasites reached the L3 stage when the immune system was upregulated by inoculating mosquitoes with bacteria before they fed on blood carrying microfilariae [20] . Similarly , parasite numbers were reduced when the mosquito was infected with the bacterium Wolbachia , which also upregulated the immune response [21] . Anti-microbial peptide ( AMP ) production may be responsible for these effects as cecropin negatively affects worm motility [22] . However , activation of the two main immune signaling pathways , Toll and IMD , by RNAi depletion of their negative regulators , Cactus and Caspar , produced no measurable effect on resistance to B . malayi [11] . Genetic mapping of parasite resistance in mosquitoes has so far been done by individually testing markers [3 , 6 , 18] or with high-density genotyping using SNP arrays or RAD-sequencing [19 , 23] . These approaches often utilize randomly selected markers sparsely interspersed in the genome and rely on markers being in linkage disequilibrium with the causative polymorphism , which itself is unlikely to be sampled . In species like An . gambiae , linkage disequilibrium extends very short distances in wild populations [24] , and it is preferable to concentrate efforts on regions that are likely to be involved in the trait of interest . In humans , the solution has been to use exome capture to sequence only protein-coding regions of the genome , which has been met with much success in identifying the mutations that cause Mendelian diseases [25] . This is especially desirable in species like humans and Ae . aegypti , where the large and repetitive genomes mean that whole genome sequencing is prohibitively costly and that much of the non-coding sequence cannot be investigated because relatively short sequence reads cannot be uniquely mapped to the genome . We have investigated the genetic and mechanistic basis of resistance to B . malayi in Ae . aegypti using a combination of genomic and transcriptomic approaches . First , we resequenced the exome using probes we designed for Ae . aegypti and performed an association study to map the locus causing resistance with unprecedented precision . Using RNA-seq , we then measured gene expression in resistant and susceptible genotypes of the mosquito to understand how this locus alters the transcriptional response to filarial nematode infection . To minimize the contribution of random genetic differences between the resistant and susceptible lines , we performed genetic crosses to isolate the resistance locus in a common genetic background . This allowed us to identify differences in immune and non-immune response gene expression that will facilitate our understanding of mechanisms of resistance .
A wild outcrossed population was established for association mapping . Mosquito eggs were collected in July 2010 from a 120 km stretch between Kilifi , Malindi , and Mombasa in coastal Kenya using oviposition traps [26] . Each trap consisted of a black plastic cup , hay infused water ( 4 g dried grass in 1 L of water for 4 days ) and a strip of creped cardboard paper . Eggs from each collection site ( median of 42 eggs/trap with 1–16 traps used per collection site ) were hatched in the laboratory and reared separately . Strains were established from two collection sites near Kilifi ( St . Thomas and Mabarikani ) and one site each near Malindi ( Muthangani ) and Mombasa ( Mtwapa ) . At the F2 generation all strains were reciprocally crossed to each other and to themselves , with similar numbers of males and females in each group . Fifteen males and fifteen females from each cross ( 480 individuals total ) were used to start an outcrossed population , where they were allowed to mate randomly for six generations . Each generation was maintained at a minimum population size of 900 adults and was not allowed to overlap with the previous generation . We measured the effect of the resistance locus on gene expression by taking advantage of sex linkage to generate susceptible and resistant mosquitoes that are genetically equivalent across most of their genome . Resistance has previously been mapped to approximately 4-21cM from the sex-determination locus [17 , 19] and is dominant in action . The Liverpool IB12 ( LVP-IB12R ) strain of Ae . aegypti is a highly inbred line that was used for the genome sequencing project [27] and was previously found to be resistant to infection[19] . It is derived from the Liverpool strain which has been maintained in culture since 1936 and was originally collected from West Africa [12] . A strain of Liverpool susceptible to infection by B . malayi ( LVP-FR3S ) [19] was obtained from the NIAID/NIH Filariasis Research Reagent Resource Center ( FR3 , Atlanta , Georgia , USA ) . We refer to the strains as LVPR or LVPS from this point on . To obtain resistant progeny , we crossed LVPR virgin females to LVPS males and backcrossed F1 males to LVPS virgin females . To obtain susceptible progeny , we crossed LVPS virgin females to LVPR males and backcrossed F1 males to LVPS virgin females . All mosquitoes were reared at a larval density of 200 individuals in 1 . 8 L of water . They were fed liver powder as larvae and 10% w/v fructose with 0 . 1% para-aminobenzoic acid ( PABA ) as adults and kept at 28°C ( ± 1°C ) with 75% ( ±5% ) humidity and a 12 hour light:dark cycle . Females were blood fed using an artificial membrane feeder ( Hemotek Limited , UK ) with donated human blood obtained from Blood Transfusion Services at Addenbrooke’s Hospital , Cambridge , UK . The temperature of the blood was maintained at 37°C in the feeders . To infect mosquitoes for association mapping , B . malayi was obtained from Darren Cook and Mark Taylor at the Liverpool School of Tropical Medicine ( LSTM ) , where they were reared in gerbils . Microfilariae were harvested into RPMI medium , which was then centrifuged at 700 rpm for 5 minutes and 0 . 5 mL of the pellet was transferred to 40 mL of blood . Microfilariae were incubated in the blood at 37°C for at least one hour prior to feeding . Outcrossed and control LVPS mosquitoes were fed on blood containing parasites at a concentration of 457 microfilariae per 20 μl of blood . Female mosquitoes were 6 to 9 days old on the day of infection . Unfed mosquitoes were discarded , and infected mosquitoes were maintained on a 10% fructose solution with 0 . 1% PABA for 10–11 days post-infection . To check for infection , individual mosquitoes were separated at the head and thorax at 10 or 11 days after infection and incubated in 100 μl of 1X phosphate buffered saline ( PBS ) for one hour at 37°C . We found this caused L3 larvae to migrate into the PBS and gave similar estimates of infection as individually dissecting mosquitoes . The supernatant was transferred to a microscope slide , the number of L3 parasites was counted , and the mosquito carcasses were stored at -80°C until DNA extraction could be performed . Mosquitoes were classified as susceptible to infection if they had one or more L3 parasites and were classified as resistant if they had none . For measuring gene expression , resistant and susceptible progeny from the crosses described in the previous section were collected from the following treatments: immediately prior to blood feeding and 12 and 48 hours post-feeding with either a control blood meal or a blood meal containing microfilariae . Microfilariae were harvested into 50 mL RPMI medium and incubated overnight with 0 . 5 mL gentamicin ( 10 mg/ml in water ) at 28°C , and 0 . 5 mL of the pellet formed overnight was transferred to 16 mL of blood . The infective blood meal contained 160 microfilariae per 20 ul of blood . A non-infective control of 50 mL RPMI with 0 . 5 mL gentamicin was also incubated in the same manner , and 0 . 5 mL of solution was transferred to 16 mL of blood . Both blood vials were then incubated at 37°C for at least one hour prior to feeding . Female mosquitoes were 4 to 8 days old on the day of blood feeding . Three to four replicate cages were maintained for each treatment and all time points were collected from the same cages . After blood feeding , mosquitoes were maintained in paper cups in groups of 8 individuals and were given 10% fructose with 0 . 1% PABA after collection of the 12 hour time point . We dissected five individual mosquitoes of each genotype at 24 , 48 , and 72 hours after infection to follow the progression of B . malayi development in resistant versus susceptible mosquitoes . Pools of 8 individuals for each treatment were snap frozen at each time point and stored at -80°C prior to RNA extraction . DNA was extracted from single mosquitoes using QiaAmp MicroDNA kit ( Qiagen ) with the following modifications . Tissues were incubated with RNAse post-homogenization and no carrier RNA was used . DNA was eluted in 50 μl AE buffer and 1 μl of eluate was quantified with a Qubit 2 . 0 fluorimeter ( Invitrogen ) . Total RNA was extracted using Trizol ( Invitrogen ) and was treated with Turbo DNAse ( Ambion ) prior to library preparation . RNA integrity was assessed using a Bioanalyzer ( Agilent ) . We sequenced the exomes of individual mosquitoes . DNA sequencing libraries were made using TruSeq DNA Sample Preparation kits ( Illumina ) . Genomic DNA ( 600ng to 1ug of starting material ) was sheared to 500bp fragment sizes via sonication , and libraries were prepared following the instructions from the manufacturer . Exome capture was then performed to enrich for coding sequences using custom SeqCap EZ Developer probes ( Nimblegen ) . Overlapping probes covering the protein coding sequence ( not including UTRs ) in the AaegL1 . 3 gene annotations [27] were produced by Nimblegen based on exonic coordinates specified by us . In total , 26 . 7Mb of the genome ( 2% ) was targeted for enrichment . Exome capture coordinates are available at https://www . jiggins . gen . cam . ac . uk/data/Aaegypti_exome . bed . Captures were performed on pools of 24 uniquely barcoded individuals , and the target enriched libraries were sequenced with either 100bp paired-end HiSeq2000 or 150bp paired-end MiSeq ( see S1 Table ) . Library preparation , exome capture , and sequencing were performed by the High-Throughput Genomics Group at the Wellcome Trust Centre for Human Genetics ( Oxford , UK ) . In addition to the exome sequencing , we also produced low coverage whole genome sequences from some mosquitoes ( these were largely different individuals but were from the same experiment , see S1 Table ) . For production of these libraries , DNA was sheared and PCR adapters were added in a single transposase mediated ligation step using the Nextera Library kit ( Illumina ) . Fifty ng of genomic DNA was used per individual and libraries were prepared following the instructions from the manufacturer . Libraries were pooled in groups of 21–25 uniquely barcoded individuals and sequenced with 100bp paired-end HiSeq2000 by the Biosciences Core Laboratory at King Abdullah University of Science of Technology ( KAUST ) ( Thuwal , Saudi Arabia ) . RNA sequencing libraries were made using the TruSeq RNA Sample Preparation kit version 1 ( Illumina ) starting with 3 ug of total RNA per library . Libraries from different treatments were balanced between lanes ( see S2 Table ) , pooled in groups of 8–10 libraries per lane , and sequenced with four lanes of 100 bp paired-end HiSeq2000 by the Eastern Sequence and Informatics Hub ( EASIH ) ( Addenbrooke’s , Cambridge , UK ) . Sequences from DNA sequencing libraries were quality trimmed from the 3’ end using Trimmomatic version 0 . 30 [28] when average quality scores in sliding windows of 4 base pairs dropped below 20 or when the quality score at the end of the read dropped below 20 . Sequences less than 50 base pairs in length and unpaired reads were discarded . Sequences were aligned to the reference genome ( AaegL1 , Oct 2005 ) [27] with BWA version 0 . 6 . 1-r104 [29] using the default parameters . Alignments for individuals sequenced across different lanes were merged into single BAM files using Picard version 1 . 93 . Alignments were sorted , indexed , and assigned read groups using SAMtools version 0 . 1 . 18 [30] and Picard . Indels were realigned using GATK version 2 . 3 [31] , and PCR and optical duplicates were removed using Picard . We have deposited the raw sequencing reads to the Short Read Archive with Accession Number SRP044393 . We performed association mapping using a combination of high and low coverage sequences . Average exome coverage from whole genome sequenced libraries was 0 . 73X per sample while average exome coverage from exome captured libraries was 32X for HiSeq sequencing and 2 . 3X for MiSeq sequencing . For this reason , we tested for associations with infection status using genotype posterior probabilities , which incorporate uncertainty in genotype calls , rather than calling individual genotypes prior to mapping [32] using the doAsso function in ANGSD version 0 . 539 [33] . BAM files were used as input for ANGSD . All SNPs called with a LRT statistic greater than 24 ( P<10–6 ) were tested for association with susceptibility to Brugia . Only bases with a minimum base quality greater than 20 and only reads that were uniquely mapped and with a mapping quality greater than 20 were included . Major and minor alleles were inferred from genotype likelihoods using the genotype likelihood model implemented in SAMtools [34] , and allele frequencies were estimated assuming a known minor allele using an EM algorithm [35] . Associations were tested under an additive model with logistic regression , a dominance model or a recessive model . The dominance and recessive models test for associations with infection status assuming the minor allele is dominant or recessive respectively . In addition , the additive model was reimplemented setting the most significant marker from the original test as a covariate ( supercont1 . 398 , position 175496 ) to test for the presence of a second locus . Only individuals with full genotypic information at this SNP with a posterior probability of 0 . 7 were included ( 73 of 140 individuals ) , and the covariate was coded under the dominant model . At least 15 individuals were required to have each genotypic class for the additive , dominance , and covariate models , and at least 10 individuals were required to have each genotypic class for the recessive model . To obtain a genome-wide significance threshold for each model that is corrected for multiple tests we permuted the phenotypes and repeated the analysis 200 times , each time retaining the lowest P-value across all variants to generate a null distribution . This was used to set a genome-wide significance cutoff of P<0 . 01 and P<0 . 05 . We also tested whether any indels were associated with resistance . ANGSD can only test SNPs for associations directly from BAM files , so we provided indel genotype probabilities , which are used in an intermediate step in ANGSD , to test for associations . Genotype probabilities for indels were produced using GATK’s UnifiedGenotyper and ProduceBeagleInput . Only the additive model was tested using this method , and significance was assessed by permutation as described for SNPs . The variant effect predictor [36] was used to assign variants to genes and classify their effects ( non-synonymous , synonymous , etc ) using gene annotation set AaegL1 . 3 . We found that multiple SNPs in linkage disequilibrium were associated with resistance , so we excluded variants that explained the infection data significantly less well than our top hit ( supercont1 . 398 , position 175496 ) . Using only the HiSeq exome sequenced individuals , we fitted a generalized linear model with a logit link function , where the response was the probability of a mosquito being infected with an L3 worm , and the predictor variables were the ‘top hit’ SNP and the SNP in question . A SNP was rejected if it was not significant but the ‘top hit’ SNP was . Sequences from RNA sequencing libraries were quality trimmed using the same method as used for DNA sequencing libraries , except that sequences less than 25 base pairs in length were discarded . An average of 42 million paired-end reads were obtained from each of the 36 libraries ( S2 Table ) . Reads were aligned to predicted transcripts in the Ae . aegypti transcriptome ( gene annotation set AaegL2 . 0 ) with Bowtie2 version 2 . 1 . 0 [37] using TopHat2 version 2 . 0 . 9 [38] with 10 mismatches allowed , read gap length and read edit distance set to 5 , and no novel junctions allowed . Reads were mapped to the B . malayi genome ( Ensembl version 3 . 0 . 19 ) [39] using TopHat2 as described above , but gene expression was not analyzed further due to low coverage . We have deposited the raw sequencing reads to the Short Read Archive with Accession Number SRP044393 . Differential expression analysis was performed using edgeR [40] after enumerating the number of reads per transcript with HTSeq [41] . We made the following comparisons: 1 ) Differential expression in response to infection , performed separately for each genotype and time point; 2 ) Differential expression between genotypes prior to infection to measure constitutive expression differences; 3 ) Difference in response to infection between genotypes ( interaction model ) , performed separately for each time point . In all cases , we filtered out lowly expressed genes by requiring that each gene included in our comparison have at least 0 . 1 count per million ( 0 . 1 cpm ) in enough samples to equal our smallest replicate size for that comparison ( n = 2–4 ) . All pairwise comparisons were made using exact tests , and the interaction models were fit using general linear models that accounted for genotype and infection status . Significance was assessed either as having an experiment-wide FDR<0 . 20 ( pairwise comparisons ) or an individual gene significance of P<0 . 01 ( interaction model ) . The biological coefficient of variation ( BCV ) , a measure of biological variability between replicates , ranged from 0 . 179 to 0 . 455 ( S3 Table ) . After excluding the four libraries with the lowest library amplifications ( S2 Table ) , the BCV ranged from 0 . 179 to 0 . 353 . The number of genes meeting our filtering criteria ranged from 12 , 549 to 13 , 594 ( of 17 , 165 ) after excluding poor libraries ( S3 Table ) . We used the RNA-seq data and Popoolation2 [42] to measure differentiation ( FST ) between resistant and susceptible progeny on a per SNP and per gene basis . BAM alignments from all treatments from the same cross ( yielding either resistant or susceptible progeny ) were merged prior to analysis . We compared our data on gene expression patterns with previously published microarray data for the Toll and IMD pathways [43] and JAK-STAT pathway [44] . To determine which pathways were activated by infection , we examined gene expression patterns in response to B . malayi in those genes that were previously shown to be differentially expressed as a result of perturbation of each pathway . We also compared expression patterns with the response to infection by Wolbachia strain wMelPop-CLA [45] . Data for this comparison was downloaded from VectorBase [46] , and only genes that were significant at P<0 . 01 were used for comparison . We classified immunity genes using ImmunoDB ( http://cegg . unige . ch/Insecta/immunodb/ ) and manual curation by ourselves of more recently identified immune genes .
To create a population that varied in susceptibility to B . malayi , we collected Ae . aegypti eggs from the coastal region of Kenya where there is known to be a mixture of genetically resistant and susceptible individuals [15] . These eggs were used to create a large outcrossed population that was maintained in the laboratory for 6 generations . The mosquitoes were then fed on human blood containing B . malayi microfilariae , which resulted in 23% ( 88 of 388 ) becoming infected with L3 larvae , with an average of 2 . 4 L3’s in each infected mosquito . This is a considerably lower infection rate than in the susceptible control line ( 86% of mosquitoes infected , 19 of 22 , with an average of 2 . 5 L3’s per infected mosquito ) , suggesting that there is genetic variation in susceptibility within our population . To identify genes associated with resistance , we used a combination of exome sequencing or low coverage whole-genome sequencing . The Ae . aegypti genome is large and repetitive , so exome sequencing provided us with far higher coverage of the exonic regions than was possible with the whole genome sequencing . The exome capture was highly efficient , resulting in 100 times greater coverage of the exome regions ( 26 . 7 MB , 2% of the genome ) compared with the non-exome regions . So that we could combine the high and low coverage data , we performed our association mapping using an approach based on genotype probabilities at each site ( as opposed to calling genotypes and then testing each site for an association with infection ) . In total we sequenced 67 L3-infected and 73 uninfected mosquitoes , which we classified as susceptible and resistant respectively . We found that susceptibility to Brugia has a simple genetic basis in our population , with a small number of sites highly significantly associated with infection ( Fig . 1A ) . Of the sites that have a known position in the genome , all of those with a genome-wide significance of P<0 . 01 were clustered together at 0 cM on chromosome 1 ( Fig . 1A ) . To test whether there were multiple genes affecting susceptibility to Brugia , we repeated the association study including the most significant variant from the first analysis as a covariate . This resulted in no significant associations ( Fig . 1B ) . Furthermore , quantile-quantile ( qq ) plots comparing expected and observed P-values in the analysis with the top SNP as a covariate confirm that there are no additional associations ( S1 Fig ) . Therefore we can conclude that there is a single variant causing the differences in susceptibility , and all the significant associations are caused by sites in linkage disequilibrium with this variant . To test whether resistance is dominant , we repeated the association study using a model that assumed the minor allele to be either dominant or recessive . The dominant model resulted in a cluster of significant associations on chromosome 1 , and the top associations were more significant than the previous analysis that assumed additive effects ( Fig . 1C ) . In contrast , the recessive model generated only two marginally significant associations ( Fig . 1D ) , and inspection of these showed that they were caused by linkage disequilibrium with the highly significant dominant variants . The minor allele of the most significant site was associated with increased resistance , so we can conclude that resistance is caused by a single dominant locus at 0 cM on chromosome 1 ( positions 1p23 and 1p25 on the physical map [47] ) . This is within the same region that we previously genetically mapped in crosses between laboratory lines from West Africa ( 0 to 12 cM ) [19] , and in the same physical region where a sex-linked , dominant resistance locus was approximately mapped ( resistance: 1p31 , sex determining region: 1q21 ) [18 , 48] , so the associations we detected are likely to be caused by the same resistance region . Therefore , the previously identified laboratory QTL is present at an appreciable frequency in East Africa in the wild . We next examined whether the variant causing resistance could be identified in our dataset . We found 26 SNPs associated with infection below a genome-wide significance threshold of 1% and an additional 27 below 5% ( Dominance Model; Fig . 1C and S1 Dataset ) . No indels were significantly associated with infection . Because of our much higher coverage of the exome , all of the significant associations were in or near to genes . We used two criteria to exclude SNPs from the list of candidate loci . First , our data only provides evidence for a single causative variant , so we are able to reduce this list from 53 SNPs to 19 SNPs by excluding variants that explain the phenotypic data significantly less well than our top hit ( see methods for details; based on HiSeq exome sequences alone , S1 Dataset ) . Second , in the experiments described below we find that worms never develop in mosquitoes carrying the resistance allele , and only 7 of the 19 SNPs follow this pattern ( allowing a 10% phenotyping error and using only HiSeq exome data , S1 Dataset ) . None of these SNPs alter the protein sequence ( 6 synonymous and 1 intronic , S1 Dataset ) . Furthermore , these SNPs occur in five different genes , none of which have known immune functions ( gamma-tubulin complex component 3 [AAEL008465] , cysteine synthase [AAEL008467] , putative latent nuclear antigen [AAEL002082] , conserved hypothetical protein [AAEL008350] , and chaoptin [AAEL008940] ) . Therefore , we cannot identify a single variant that causes resistance . A combination of two factors has likely prevented us from identifying the gene that is causing mosquitoes to be resistant . First , linkage disequilibrium in our mapping population means that significant associations are found across multiple scaffolds , sometimes with identical segregation patterns . This is visible in quantile-quantile plots where there is a large excess of sites with differing frequencies in resistant and susceptible chromosomes ( S1 Fig ) . Second , low sequencing coverage means we have limited power to detect associations in non-coding sequence . Therefore , if the variant causing resistance is in a low coverage region we may not detect it , but it will generate significant associations in nearby genes in linkage disequilibrium . While we did not identify a causal variant , the list of sites and genes associated with resistance are strong candidates . Even if none of these genes prove to be causing resistance , we have narrowed the region down to the 14 scaffolds that either map to the 0 cM genetic position on chromosome 1 ( scaffolds 1 . 48b , 1 . 97b , 1 . 166 , 1 . 222a , 1 . 267 , 1 . 272 , 1 . 296 , 1 . 327 , 1 . 398 , 1 . 461 , 1 . 487 , 1 . 512 , 1 . 696 , and 1 . 970 , containing 175 genes and 15 . 0 Mb ) or the 6 scaffolds with SNPs in the top 1% of hits whose location in the genome is uncertain ( scaffolds 1 . 226 , 1 . 360 , 1 . 676 , 1 . 1219 , 1 . 257 , 1 . 389 , containing at most 94 genes and 6 . 6 Mb ) . FST data obtained from our RNA-seq experiment ( described in the following section ) supports four of these six scaffolds as being located on chromosome 1 ( 1 . 360 , 1 . 676 , 1 . 257 , and 1 . 389 ) . This conflicts with a previous analysis showing that two of these scaffolds ( 1 . 360 and 1 . 257 ) are on chromosome 2 [19] , but this is not unexpected as the genome has a high misassembly rate , and many scaffolds are chimeras of sequences found in different places in the genome [19] . To examine the effects of the B . malayi resistance locus on mosquito gene expression and worm development we used two laboratory mosquito lines , one resistant ( LVPR ) and one susceptible ( LVPS ) . When comparing these lines it is desirable to homogenize the genetic background in which the two alleles of the locus are found . To do this we took advantage of the locus being dominant and sex-linked . By crossing resistant ( LVPR ) males with susceptible ( LVPS ) females or vice-versa , and then backcrossing the F1 males to susceptible female ( LVPS ) mosquitoes , we generated resistant and susceptible mosquitoes that are genetically similar except for the effects of other dominant sex-linked genes . We checked that this was the case by calling SNPs from the RNA-seq data ( described below ) and measuring differentiation between the backcross progeny using FST , which ranges from 0 ( complete allele sharing ) to 1 ( complete differentiation ) . As expected , we found that the first chromosome was differentiated between the backcross progeny from the resistant grandmother compared to progeny from the susceptible grandmother ( mean FST = 0 . 081 ) . The second and third chromosomes were not differentiated ( mean FST = 0 . 011 and 0 . 009 respectively ) ( S2 Fig ) . The region of differentiation extended across the length of the first chromosome ( S2 Fig ) , which is likely to be due to the low rate of recombination in Ae . aegypti . We found that resistance closely segregated with the sex-determining locus ( Fig . 2 A-B ) , with no L3 worms ( 0 of 48 ) developing in the backcross progeny from the resistant ( LVPR ) grandmother . In contrast , in the backcross progeny from the susceptible ( LVPS ) grandmother , 89% ( 24 of 27 ) of the mosquitoes were infected , and these had an average load of 5 . 1 L3’s per mosquito ( Fig . 2 A-B ) . The infection rates of these backcrosses were similar to those of the parental lines ( LVPS and LVPR; Fig . 2 A-B ) . The sex determining locus is at approximately 21 cM [19] , and we observe no recombinants between the sex and resistance loci ( 0 of 48 , Fig . 2 A-B ) , although some recombinants may not be detected because not all susceptible individuals develop an infection ( 83% , 10 of 12 ) . These results suggest that the difference in resistance between these lines is being controlled by the same dominant sex-linked locus described in the Kenyan population in the previous section . We will subsequently refer to these as the resistant and susceptible genotypes . Using these genetically similar mosquitoes we examined the stage in B . malayi development being targeted by resistant mosquitoes . In susceptible Ae . aegypti , microfilariae penetrate the midgut and migrate to the thorax from 1–24 hours following infection , enter a non-feeding L1 stage of development from 6–72 hours , and eventually exit the mosquito as L3 larvae after approximately ten to twelve days [11 , 49] . Microfilariae are approximately 200–300 um in size at the time of ingestion and L3’s can reach 1–3 mm in size [11] , and the numbers of worms decrease from an average of 17 microfilariae per mosquito at the time of ingestion to 6 at the L3 stage [11] . We find microfilariae and L1’s in the thorax of both resistant and susceptible mosquitoes for several days after infection ( Fig . 2 C-H ) . At 24 hours we found that microfilariae have migrated from the midgut to the thoracic tissues in both genotypes ( Fig . 2 C-D ) . We continue to find live microfilariae and L1’s in both genotypes at 48 and 96 hours ( Fig . 2 E-H ) , which is consistent ongoing migration until blood digestion ends at around 72 hours after feeding [50] . By 48 hours and especially by 96 hours after infection there are clear differences between the genotypes , with microfilariae molting into the non-feeding L1 larvae in susceptible hosts , whereas growth is arrested in resistant hosts ( Fig . 2 E-H ) . The microfilariae and L1’s in the resistant genotype are phenotypically distinguishable from those in susceptible genotype , and appear to be smaller and thinner ( Fig . 2 G-H ) . We could not accurately compare microfilariae numbers using our dissections , so we instead compared gene expression ( see next paragraph ) . We sometimes see worms being melanized in both genotypes , but usually they are not . We can also follow worm development from the number of RNA-seq reads mapping to the B . malayi genome . At 12 hours , B . malayi gene expression levels are comparable in resistant and susceptible genotypes , suggesting similar growth and numbers at this time point . By 48 hours , gene expression is much higher in the susceptible genotypes , suggesting greater parasite growth in the susceptible host [51] ( Fig . 2I ) . This suggests that parasite development is aborted in resistant mosquitoes between 12 and 48 hours after infection , although languishing microfilariae and L1’s can be found in resistant mosquitoes for several more days . Our results corroborate previous studies which show that parasites in refractory Ae . aegypti decline in numbers by 48 hours after infection [49 , 51 , 52] . Therefore , parasites are able to reach thorax of resistant mosquitoes , but are killed in the microfilariae and L1 stages . This suggests that the resistance mechanism plays an important role after invasion of the thorax . We used RNA-seq to investigate the transcriptional response of Ae . aegypti to B . malayi ( using mosquitoes from the same experiment as shown in Fig . 2 ) . We chose to measure gene expression early in the infection when differences in worm development in susceptible and resistant mosquitoes appear . For each genotype separately , we measured differential expression between infected and uninfected mosquitoes . We found a strong correlation between genotypes in the induced responses at both time points ( Pearson’s R2 at 12hr: 0 . 51 , 48hr: 0 . 58; Fig . 3 ) , suggesting a similar response to B . malayi infection in resistant and susceptible mosquitoes . Our first time-point was 12 hours post-infection , when microfilariae are migrating from the midgut to the thorax and beginning to enter the L1 stage . B . malayi infection significantly regulated 459 genes in the resistant line ( 238 induced and 221 repressed ) and 636 genes in the susceptible line ( 535 induced and 101 repressed ) ( S2 Dataset ) . A total of 97 genes were significantly upregulated in both resistant and susceptible mosquitoes in response to infection , and 26 were significantly downregulated in both genotypes . Our second time point was 48 hours post-infection , when over half of the parasites will be L1’s in susceptible mosquitoes [11] and parasites are beginning to die in resistant mosquitoes . At this time , Brugia infection significantly regulated 1 , 029 genes ( 725 genes induced and 304 genes repressed ) in the resistant line and 609 genes ( 467 genes induced and 142 genes repressed ) in the susceptible line ( S2 Dataset ) . A total of 274 genes were significantly upregulated in both genotypes and 91 genes were downregulated in both genotypes . There is a clear transcriptional response of immune-related genes to B . malayi at both 12 and 48 hours after infection ( Figs . 3 , 4 ) . The large majority of the immune-related genes are upregulated in the infected relative to the uninfected mosquitoes , indicating that the immune system is being activated ( Fig . 4 ) . Furthermore , this upregulation of immune genes is apparent at 12 hours post-infection , which conflicts with an earlier report that very little transcriptional response is seen until 5 days after infection in susceptible mosquitoes [11] . The previous study used microarrays and pooled across several time points so may be less sensitive than our approach . There is a striking concordance between the transcriptional response of immune genes in the resistant and susceptible genotypes ( Fig . 4 ) . Among the immune genes significantly differentially regulated 12 hours post-infection in at least one of the genotypes , almost invariably the gene is differentially regulated in the same direction in the other genotype ( even if this is not statistically significant ) ( Fig . 4 ) . At 48 hours there is still a similar response in the two genotypes , although some differences are emerging ( see below ) . The genes that are differentially regulated in response to Brugia infection in both genotypes have a range of immune functions ( Fig . 4 ) . These include melanization ( CLIPB8 , CLIPB29 , and CLIPB35 [53] ) , hemocyte-mediated immunity ( CLIPD1 , FREP3 , FREP5 , FREP10 , and a Nimrod homolog [54] ) , antioxidant protection ( GPXH1 , HPX3 ) , the Toll pathway ( CACT , GNBPB3 ) , IMD pathway ( IKK2 ) , and JAK-STAT pathway ( STAT ) . Melanization of B . malayi was occasionally observed at early stages of infection in both genotypes and is controlled by clip serine protease cascades and phenoloxidases which are released by hemocytes [53] . It is associated with the production of reactive oxygen species , which is expected to trigger the production of antioxidants . The activation of the Toll , IMD and JAK-STAT pathway by Brugia infection has never been previously described . In the resistant and susceptible mosquitoes a subset of genes responded to B . malayi infection in different directions or with a significantly different magnitude , and these may give clues as to the mechanism of resistance ( Figs . 3 , S3 ) . At 12 hours post-infection we detected 86 such genes ( N = 13 , 156 ) by testing for a significant interaction between genotype and infection ( S2 Dataset ) . Of these , 62 had sufficient data to be tested for differential expression separately within each genotype . Contrary to our expectation , we found that most ( 53 of 62 ) of these genes were more highly expressed in the susceptible line , with 48 of these being upregulated in the susceptible line and downregulated in the resistant ( Fig . 3 ) . There is little evidence to suggest that resistance is due to immune gene expression at the early 12 hour time point , as the susceptible mosquitoes upregulated more immune-related genes than the resistant ones ( Figs . 3 , 4 ) . Immune genes upregulated in susceptible and downregulated in resistant mosquitoes include a cecropin ( CECN ) , a lectin ( CTL14 ) , a gram-negative binding protein ( GNBP4 ) , the IMD signaling pathway molecule IKK1A , and a leucine-rich repeat protein ( LRIM15 ) , all of which may be related to the immune response ( Fig . 4 ) . In contrast , the resistant line only upregulated expression of a single clip-domain serine protease ( CLIPD9 ) ( Fig . 4 ) . Furthermore , 10 genes that are induced by infection by the Wolbachia strain wMelPop-CLA , which is known to make mosquitoes more resistant to filarial nematodes [21 , 45] , are upregulated in the susceptible line but downregulated in the resistant line ( S5 Fig ) . At the early time point , the non-immunity genes that had a different transcriptional response to B . malayi in the two genotypes had a diversity of functions , but many were related to metabolism , with the susceptible line upregulating transport and digestion-related genes and the resistant line downregulating digestion and upregulating vitellogenin ( egg yolk protein ) ( Figs . 3 , S4 ) . Of the 48 genes upregulated in the susceptible genotype and downregulated in the resistant genotype , a number were transporter , transmembrane , or gut structural genes ( amino acid transporter PAT1 ( AAEL007191 ) , innexin ( AAEL014846 ) , integrin ( AAEL014846 ) , an insect major allergen/G12 family member ( AAEL009166 ) , and a monocarboxylate transporter ( AAEL013915 ) ) . A translational repressor that is important in starvation responses was also induced only in the susceptible line ( eukaryotic initiation factor 4E binding protein , AAEL001864 ) [55] . Only nine genes had higher induction in the resistant genotype . Two were paralogs of a vitellogenin A1 precursor ( AAEL006126 and AAEL006138 ) and two were paralogs of trypsin modulating oostatic factor ( TMOF ) ( AAEL006670 and AAEL014561 ) . TMOF is expressed in the ovaries and binds to receptors in the midgut where it turns off trypsin production [56] and downregulates the production of vitellogenin . The other genes were an immune modulator ( CLIPD9 ) , an AMP dependent ligase ( AAEL006823 ) , a putative pupal cuticle protein ( AAEL010127 ) and two genes with unknown functions ( AAEL008449 and AAEL000681 ) . At 48 hours post-infection , the time point where parasites are being killed in resistant but not susceptible mosquitoes , 152 genes ( out of 13 , 594 ) responded to B . malayi infection differently in the susceptible and resistant mosquitoes ( Figs . 3 , S3 and S2 Dataset ) . Of the 122 genes with sufficient data to be tested for differential expression in response to infection within genotype , 75 had higher expression in the susceptible line and 47 had higher expression in the resistant line . In contrast to the earlier time-point , by 48 hours there were differences in the induced immune response between resistant and susceptible mosquitoes . The 152 genes that behaved differently between the two genotypes were significantly enriched for immune response categories ( S4 Fig ) and these immune response genes tended to have higher expression in the infected resistant mosquitoes ( Figs . 3 , 4 , S3 ) . These included antimicrobial peptides ( CECE: AAEL000611 , CECN: AAEL000621 , DEFA: AAEL003841 , DEFC: AAEL003832 , and DEFD: AAEL003857 ) , a prophenoloxidase ( PPO5: AAEL013492 ) , JNK ( AAEL008622 ) , CLIPD2 ( AAEL004979 ) , FREP28 ( AAEL003156 ) , a Niemann Pick-type C gene ML15B ( AAEL009556 ) , and the heme peroxidase HPX5 ( AAEL002354 ) . The only immune gene that had higher expression in susceptible mosquitoes was DSCAM ( AAEL011284 ) . Unexpectedly , many immune genes were downregulated in susceptible mosquitoes in response to B . malayi infection while being upregulated in the resistant ones ( Fig . 4 ) . Two Cecropin genes were found to be suppressed in susceptible mosquitoes at 48–72 hours post-infection in a previous study [11] , supporting our results . This pattern therefore suggests parasite suppression of these immune response genes or dysregulation within susceptible mosquitoes . At 48 hours after infection , the susceptible genotype appears to have higher upregulation of genes that are highly induced by blood feeding [50] and are potentially related to metabolism and digestion . This include trypsins ( AAEL013703 , AAEL013707 , AAEL013712 , and AAEL013715 ) [57] , several chymotrypsin-related serine proteases expressed in the midgut ( AAEL001674 , AAEL001690 , AAEL001693 , AAEL001701 , AAEL008769 , and AAEL008780 ) [57] , a putative serine collagenase ( AAEL007432 ) [57] , a sphingomyelin phosphodiesterase ( AAEL013717 ) [27] , and methylenetetrahydrofolate dehydrogenases ( AAEL006085 and AAEL014871 ) [27] . This is consistent with trypsin modulating oostatic factor ( TMOF ) paralogs , which inhibit late trypsin production , being downregulated in the susceptible line but not the resistant line at 12 hours , leading to a longer period of expression of trypsins [56] . The Toll , IMD , and JAK-STAT pathways play a central role in controlling the immune response of insects , so we investigated whether they were activated by B . malayi , and whether patterns of activation differ between resistant and susceptible mosquitoes . To do this we compared our results to microarray datasets from mosquitoes where REL1 or REL2 had been overexpressed ( the transcription factors activated by the Toll and IMD pathways respectively ) [43] or PIAS , a negative regulator of the JAK-STAT pathway , had been knocked down [44] . At 12 hours post-infection , the transcriptional response of both resistant and susceptible genotypes suggests they have activated their Toll and/or IMD pathways ( Figs . 5A , S5 ) . Consistent with our earlier findings of an induced immune response , the susceptible genotype appears to be activating these pathways to a higher degree ( Figs . 5A , S5 ) . As has been previously reported , overexpression of REL1 and REL2 elicits similar responses in an overlapping set of genes ( S6 Fig ) [43] , so these patterns could be generated by B . malayi activating the Toll , IMD or both pathways . The susceptible genotype appears to activate the JAK-STAT pathway at 12 hours post infection , while the resistant genotype does not show evidence for this ( Figs . 5A , S5 ) . In particular , the insect major allergen/G12 gene AAEL009166 appears to be induced in the susceptible line but strongly downregulated by the resistant line . This gene is induced by the JAK-STAT pathway , and is highly expressed in the midgut following a bloodmeal and is thought to be involved in digestion [58 , 59] . This JAK-STAT-like response is distinct from the other pathways we investigated , as of the 10 genes that overlap between the Toll and JAK-STAT datasets , 9 go in opposite directions ( S6 Fig ) . These 9 genes include genes that respond differently to B . malayi in the two genotypes: the antimicrobial peptides CECE and DEFC ( upregulated by the Toll pathway and downregulated by the JAK-STAT pathway [44] ) and four insect major allergen genes ( AAEL009166 , AAEL010429 , AAEL010436 , and AAEL013118; downregulated by the Toll pathway and upregulated by the JAK-STAT pathway ) . The pattern of immune pathway activation 48 hours post-infection is less clear . In both the resistant and susceptible mosquitoes , B . malayi infection increases the expression of genes normally upregulated by active Toll , IMD and JAK-STAT pathways ( Figs . 5A , S5 ) . This is reflected in our earlier result that both genotypes show signs of an induced immune response . However , many genes that are normally downregulated by these pathways are upregulated in both genotypes . This upregulation appears to be higher in the susceptible genotype , especially in the case of the IMD pathway ( Figs . 5A , S5 ) , including trypsins , a trypsin 5G1 precursor , and serine-type endopeptidases ( Fig . 3 ) . Ae . aegypti has been artificially infected with a bacterial symbiont called Wolbachia [45] , and the wMelPop-CLA strain increases resistance to B . malayi [21] . Additionally , B . malayi harbors its own obligate strain of Wolbachia wBm [60] , which may also interact with the Ae . aegypti immune response . We therefore compared the transcriptional responses to Wolbachia and B . malayi using previously published data for Wolbachia ( see Methods ) . At the early time point there are striking differences between the resistant and susceptible genotypes . In the resistant genotype there is a strong negative correlation in the responses to B . malayi and Wolbachia wMelPop-CLA , with genes tending to respond in opposite directions ( Figs . 5A , S5 ) . In contrast , in the susceptible genotype there is a positive correlation ( Figs . 5A , S5 ) . At 48 hours post-infection , the transcriptional response to B . malayi in both the susceptible and resistant mosquitoes is strongly correlated to the response to the wMelPop-CLA ( Figs . 5A , S5 ) . A subset of these genes behaved differently in the two genotypes , including the downregulation of antimicrobial peptides in susceptible mosquitoes that is described above ( S5 Fig ) . Until now we have only investigated induced responses to B . malayi , but it is possible that constitutive gene expression differences might contribute to resistance . As in the previous section , we identified genes that were upregulated or downregulated by stimulation of the Toll , IMD , or JAK-STAT pathways , or by Wolbachia infection ( Fig . 5B ) . We found that the resistant line had significantly higher constitutive expression of genes under the control of the Toll pathway and genes induced by the Wolbachia strain wMelPop-CLA ( Fig . 5B ) . Overall , it appears that some immune pathways are upregulated in the resistant genotype prior to infection , suggesting that constitutive differences in gene expression could potentially play a role in resistance . We confirmed that these results are robust to the confounding effects of biases in mapping sequence reads . Reads will not map to the reference genome if SNPs cause a significant number of mismatches to the reference genome . Despite homogenizing the genetic background of our mosquitoes , the resistant and susceptible genotypes are still genetically different in sex-linked regions . We found that these differences are biasing our estimates of gene expression , as genetic differentiation ( FST ) between the genotypes correlated with differences in expression levels ( details in S7 Fig ) . We therefore repeated our comparison of constitutive expression using only genes found on the second or third chromosomes which are not affected by this bias , and found that the results of our pathway comparisons were qualitatively unchanged . It is important to note that this bias does not affect the analysis of induced responses to B . malayi as these rely on comparisons within genotype rather than between genotypes . The polymorphism that confers resistance could be in a gene whose expression changes in response to B . malayi infection . Therefore , we examined our gene expression data for differentially expressed genes in the genomic region controlling resistance ( Fig . 6 ) . Specifically , we chose genes that were either differentially expressed in response to B . malayi infection or constitutively differentially expressed between the genotypes ( note the caveat above that this comparison can be confounded by mapping biases ) . This generated a list of 41 genes ( Fig . 6 ) . The most significant SNP in the association study is within the intron of the gene AAEL009467 , and this gene is induced in response to B . malayi infection by both genotypes at the early time point ( non-significant in the resistant genotype ) and by the resistant genotype only at the later time point ( Fig . 6 ) . This gene encodes a protein that contains an actin-depolymerizing factor homology ( ADF-H ) domain , which mediates actin binding . It has 1:1 orthologs in other mosquito vectors including Anopheles gambiae and Culex quinquefasciatus . This SNP is unlikely to be causing resistance as it does not have the expected segregation pattern ( see above ) . We had high sequence coverage and thus high power for this gene across the entire coding sequence so it is unlikely that the casual mutation is protein coding . It is possible that a non-coding variant affecting gene expression could be causal . An unnamed gene ( AAEL009459 ) encoding a hypothetical protein is located 80kb away from the most significant hit and is strongly induced by infection in resistant mosquitoes at the early time point and in both genotypes at the later time point ( Fig . 6 ) . It also has higher constitutive expression in the resistant genotype , although high FST suggests that this may be an artifact of mapping biases . This gene is also strongly upregulated by Wolbachia wMelPop-CLA and has no known orthologs in other species . This gene appears to have a neighboring unannotated paralog that is expressed in our RNA-seq data , and the duplicate has two distinct alleles in the resistant and susceptible genotypes from the laboratory crosses . Unfortunately , because it is unannotated , sequence coverage is extremely low in the association study samples , so it is unclear if these alleles are associated with resistance . Two immune-related genes , a Niemann Pick-type C1 ( NPC1 ) gene ( AAEL009531 ) and Toll1A ( AAEL007613 ) , are also found in the region associated with resistance and were upregulated in response to infection in one or both genotypes . The Niemann Pick-type C1 gene is expressed in the mosquito midgut [61] and here is induced in response to B . malayi infection . Knockdown of this gene inhibits dengue infection , and it has been proposed to be a negative regulator of the immune response [61] . Toll1A is a homolog to the Drosophila Toll [62] , and the scaffold 1 . 267 , which contains Toll1A , also contains several other Toll genes .
Resistance of Ae . aegypti to B . malayi has a simple genetic basis . By combining exome and whole-genome sequencing , we found that resistance to B . malayi in a Kenyan population of Ae . aegypti is controlled by a single dominant locus . This locus is located in the same genomic region as a dominant QTL controlling B . malayi resistance that was mapped in crosses between laboratory lines from West Africa [12 , 18] . In quantitative genetics it is commonly seen that major-effect QTL from lab crosses are an artifact caused by multiple closely linked loci of smaller effect [63] , and these are broken up using high-resolution approaches like ours . We found that a model accounting for the most significant SNP gave no evidence for any additional associations , which is most parsimoniously explained by resistance to B . malayi being controlled by a single causative locus . An alternative explanation is that multiple resistance genes are being held in tight linkage disequilibrium by an unknown chromosomal inversion . It is also normal in QTL studies to identify different loci when using different crosses [64] , and for QTL that are important in the laboratory to be unimportant in the wild [65 , 66] . Again that appears not to be the case here , as we found that the locus identified from West Africa is common in an East African population . The simple genetics of B . malayi resistance is striking as most quantitative traits tend to be affected by many genes , each of which only has a small phenotypic effect [63] . However , there are a number of studies suggesting that susceptibility to infection in insects often has a very simple genetic basis . For example , in Drosophila a small number of major effect loci control resistance to viruses [67] , and the same is probably true for resistance to parasitoid wasps [68 , 69] . Similarly , it has been suggested that in humans susceptibility to infection has a simpler underlying genetic basis than other quantitative traits due to strong selection by pathogens [70] . It has been suggested that this simple genetics is a result of directional selection driving major effect resistance alleles through populations [67] . The polymorphism we have investigated in African mosquitoes is not the result of B . malayi selecting for resistance , because the parasite is only found in south and south-east Asia and is not thought to be naturally vectored by Ae . aegypti . It is known that this locus also confers resistance to the filarial nematodes B . pahangi and Wuchereria bancrofti , but not to Dirofilaria immitis , the cause of dog heartworm [16] . It is possible that selection for resistance to another closely related but unidentified nematode is responsible for the maintenance of this polymorphism . Any selection for resistance is likely to vary among populations , as genotypes that are susceptible to B . pahangi are only known to be common in East Africa [15 , 71] . Thus it remains to be seen what has maintained this polymorphism in the wild since at least 1936 , but understanding this experimentally tractable system may yield insights into natural mosquito-parasite associations . In mosquito-filarial systems , it is commonly seen that resistance is sex-linked and dominant , including a separate locus that confers resistance of Ae . aegypti to D . immitis and to the bullfrog filarial nematode Waltonella flexicauda [72 , 73] . It is tempting to speculate that this could be the result of sexually-antagonistic selection favoring resistance only in females . Our association study mapped the locus causing resistance with a high precision , but we did not identify the causative gene . Due to large effective population sizes , linkage disequilibrium in insect populations can sometimes extend only a few tens of base pairs [74] . This means that association studies like ours would often be expected to identify the exact genes involved . However , we found high levels of linkage disequilibrium in our mapping population and multiple loci with identical segregation patterns . This could be due to difficulties in establishing African Ae . aegypti in the laboratory resulting in a relatively small number of individuals founding our mapping population . This could also occur if the resistance locus falls within a chromosomal inversion , which may result in long-distance linkage disequilibrium in natural populations and might prevent mapping to the level of a single gene . Despite this , we localized the locus with a very high resolution to a single genetic position ( 0 cM ) and identified a small number of strong candidate genes , one of which may cause resistance . The availability of detailed genetic maps combined with the density of genetic markers allowed by whole-exome sequencing has enabled us to compile the first complete list of candidate resistance genes . Several steps are needed to identify the exact variant ( s ) causing resistance . First , additional association studies on populations with less linkage disequilibrium may narrow identify far fewer genes . Second , gene knockouts , RNAi and other functional studies can confirm the role of these genes in defense against filarial worms . In our association study , we identified polymorphisms in five different genes which best explained resistance and had the segregation pattern expected for single dominant resistance locus . None of these genes have any known involvement in the immune response , but it is still possible that one of them is responsible for resistance . We also identified candidates in the correct region of the genome with expression differences between the resistant and susceptible genotypes or in response to infection . These candidates did not always have significant associations , but it is possible that the true causative site was untested if it is non-coding or lacked sufficient coverage . Two genes identified this way , a niemann-pick C1 gene and Toll1A , are thought to be involved in the immune response . The niemann-pick C1 gene , which is upregulated in response to Brugia infection , has recently been shown to be required for successful dengue infection [61] . Toll1A is one of four Ae . aegypti genes homologous to Drosophila Toll ( Toll-1 ) [62 , 75] , and it is induced following infection . In Drosophila Toll is crucial for signaling through the Toll pathway , and while Toll1A does not play the same role in Ae . aegypti , there is some evidence to suggest it may play a role in modulating the immune response [62] . Lastly , the top hit in our association study is an unnamed gene that is induced by infection ( AAEL009467 ) . This gene encodes a protein with an actin-binding ADF-H domain and Brugia develops in the thoracic muscle fibers of Ae . aegypti , where actin dynamics are expected to be important . We found that Ae . aegypti upregulates immune-related genes in response to B . malayi as early as 12 hours after infection . Many of these immune genes are known to be under the control of the Toll and/or IMD signaling pathways , so it is likely that B . malayi is activating one or both of these pathways ( although the pattern is less clear-cut if genes downregulated by these pathways are considered; see Results ) . The role of the immune response in resistance to filarial infection is unclear , as it has only been studied in susceptible mosquitoes and the outcome of those studies is not entirely clear-cut . Activation of the immune system prior to infection with B . malayi confers a protective effect , with susceptible Ae . aegypti mosquitoes that are given sham infections or non-lethal bacterial infections sustaining a significantly lower intensity of infection [20] . Similarly , infection with the bacterium Wolbachia wMelPop-CLA also induces an immune response and provides some protection to susceptible mosquitoes [21] . However , RNAi knockdown of the Toll pathway and over-activation of the Toll or IMD pathways in susceptible mosquitoes did not influence the response to infection [11] . This could simply be that using RNAi to induce an immune response is less efficient than using bacteria , or it may mean Toll- and IMD-independent aspects of immunity are controlling filarial infections . Only by repeating these assays in resistant mosquitoes will it be possible to determine which pathways underlie protection in genetically resistant individuals . At 48 hours after infection , the genes most responsive to B . malayi infection are those that are induced by Wolbachia wMelPop-CLA infection . It is possible that the mosquito immune response may be launched not solely against the worm itself , but also in response to the Wolbachia harbored by the nematode , as is the case in the vertebrate host [76] , or to midgut bacteria that enter the hemocoel after B . malayi penetration . By comparing gene expression in both resistant and susceptible mosquitoes , we find that resistant mosquitoes are upregulating a number of immune effector genes at the time point where the parasites are dying , whereas the susceptible mosquitoes are not . At 12 hours post infection the immune response of the two genotypes is similar , but by 48 hours post infection , when parasites are dying in resistant mosquitoes , differences in the immune response are clear . At 48 hours , effector molecules of the immune response , including anti-microbial peptides ( AMPs ) and prophenoloxidases ( PPOs ) , are upregulated in resistant mosquitoes and either downregulated in susceptible mosquitoes or induced to a much lower degree . Expression of the AMP cecropin was previously seen to be downregulated 48–72 hours after infection in susceptible mosquitoes , supporting what we’ve seen here [11] . Although AMPs are often thought to be primarily antibacterial and antifungal , cecropins have been shown to kill B . panhangi in vitro and in vivo [22] , and cecropin and defensin overexpression kills Plasmodium in Ae . aegypti [77] . The downregulation of immune effectors like antimicrobial peptides in susceptible mosquitoes suggests that the differences in the immune response may be driven by immune suppression by the parasite . Parasitic nematodes have mechanisms that circumvent the humoral immune responses of their hosts , including proteinases that directly inhibit cecropins and cuticle factors that sequester host hemolymph proteins and prevent the activation of the immune response ( reviewed in [78] ) . This is likely to be the case for B . malayi , as the closely related parasite B . pahangi reduces the melanization response of Ae . aegypti [79] . If parasite suppression of the immune response is causing the weaker immune response seen in susceptible mosquitoes , then this pattern could either be a cause or consequence of resistance . The resistance locus that we mapped might cause the mosquitoes to mount a stronger immune response to B . malayi , perhaps by circumventing the parasite suppression , and this in turn is the cause of resistance . Alternatively nematodes in the resistant mosquitoes may be unable to suppress the immune response because they are dying for some other reason , such as if resistant mosquitoes are metabolically inhospitable or lack a receptor necessary for nematode migration . In this case , the stronger reduction in immune response seen in susceptible mosquitoes may be a consequence of having more and healthier nematodes . Resistant mosquitoes also constitutively express immune genes to a higher level , which supports a direct role for the immune system in resistance . We found that resistant mosquitoes have higher constitutive expression of genes under the control of the Toll pathway as well as genes expressed in response to infection with Wolbachia wMelPop-CLA . As described earlier , both experimental stimulation of the immune response and infection with Wolbachia strain wMelPop-CLA provide some protection to Brugia infection , although not complete resistance [20 , 21] . Genes involved in digestion , nutrient transport and egg production are expressed differently between the resistant and susceptible mosquitoes after infection by B . malayi . Susceptible mosquitoes appear to be upregulating genes involved in digestion and nutrient transport , and downregulating the egg yolk precursor protein vitellogenin . Many of the genes that are differently induced are genes that respond to blood feeding [58] . The target of rapamycin ( TOR ) pathway is central to coordinating this blood-feeding response , and is thought to detect the influx of amino acids and upregulate the expression of genes related to digestion , nutrient transport , and vitellogenesis [80] . Several of the genes that respond differently to infection in the two genotypes are in or under the control of the TOR pathway , including a known translational repressor of the pathway 4E-BP [55] , the amino acid transporter PAT1 [81] , and Vitellogenin A1 precursors [80] . Amino acids and other nutrients that are ingested in a blood meal are normally used for egg production . In infected mosquitoes , competition between the host and parasite for nutrients can result in a reduction of egg production [82] , such as seen with Wolbachia [83] and Plasmodium [84] . There can also be a conflict between the immune system and egg production , where the same resources are needed for both processes [85] . Altered expression of these genes could reflect active or passive manipulation of the host resources by the parasite , or competition within the mosquito for resources . Vitellogenin expression is strongly downregulated by B . malayi in susceptible mosquitoes at 12 hours post-infection and strongly upregulated in resistant mosquitoes at 48 hours-post infection . Despite this , no difference in the number of eggs laid is seen in either genotype in response to infection ( Cristina Ariani , personal communication ) , although other mosquito-filarial worm combinations do lead to a reduction [85–87] . In An . gambiae mosquitoes , vitellogenin production is negatively regulated downstream of Cactus/REL1/REL2 , leading to the arrest of oogenesis when the immune system in activated [88] . Vitellogenin interferes with TEP1 binding to the surface of developing Plasmodium ookinetes , so downregulation of vitellogenin production boosted the efficiency of parasite killing [88] . A key molecule underlying the different responses to B . malayi in resistant and susceptible mosquitoes might be trypsin modulation oostatic factor ( TMOF ) . TMOF is produced in the ovaries and turns off egg production and late trypsin production by binding to midgut receptors [56] . Susceptible mosquitoes have decreased TMOF production at 12 hours , and are expressing trypsins and serine proteases at higher levels at 48 hours . Trypsin production in the midgut is highly upregulated by blood feeding to allow quick digestion of proteins in the blood meal [57] , and tryptic activity has long been speculated to be a source of conflict with the parasite . Trypsin activity has been suggested to change nutrient availability , or to either digest or activate parasite proteins as they pass through the midgut [89 , 90] . Indeed , inhibition of trypsin in Ae . aegypti increases dengue virus titres [89] and decreases Plasmodium load [91] . In conclusion , by developing exome sequencing we were able to conduct the first large-scale association study in Ae . aegypti , a species with a large and repetitive genome . This allowed us to identify a major-effect locus conferring resistance to B . malayi . This locus alters the transcriptional response of mosquitoes to B . malayi , with resistant individuals expressing immune effectors to a higher degree . Our results suggest that this pattern may be driven by B . malayi suppressing the immune response in susceptible mosquitoes . | Within mosquito populations , genetic differences between individuals affect their ability to transmit human diseases such as malaria , dengue fever , and lymphatic filariasis . In the mosquito Aedes aegypti , some individuals are genetically resistant to Brugia malayi , a mosquito-vectored parasite that causes a debilitating tropical disease called lymphatic filariasis . To characterize the genetic basis of resistance , we identified resistant and susceptible mosquitoes from a wild Kenyan population , and sequenced the protein-coding region of their genomes ( the exome ) . This allowed us to locate a single region of the mosquito genome that is causing resistance and to identify genes that may be controlling the trait . To understand the mechanisms of resistance , we measured gene expression . The susceptible mosquitoes have reduced expression of immunity genes after they are infected with B . malayi , including genes known to kill this group of parasites . This is possibly because their immune response is being suppressed by the parasites . We conclude that resistance is controlled by a single locus and show that resistance results in an increased immune response . | [
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] | [] | 2015 | Exome and Transcriptome Sequencing of Aedes aegypti Identifies a Locus That Confers Resistance to Brugia malayi and Alters the Immune Response |
The late onset of neurodegeneration in humans indicates that the survival and function of cells in the nervous system must be maintained throughout adulthood . In the optic lamina of the adult Drosophila , the photoreceptor axons are surrounded by multiple types of glia . We demonstrated that the adult photoreceptors actively contribute to glia maintenance in their target field within the optic lamina . This effect is dependent on the epidermal growth factor receptor ( EGFR ) ligands produced by the R1-6 photoreceptors and transported to the optic lamina to act on EGFR in the lamina glia . EGFR signaling is necessary and sufficient to act in a cell-autonomous manner in the lamina glia . Our results suggest that EGFR signaling is required for the trafficking of the autophagosome/endosome to the lysosome . The loss of EGFR signaling results in cell degeneration most likely because of the accumulation of autophagosomes . Our findings provide in vivo evidence for the role of adult neurons in the maintenance of glia and a novel role for EGFR signaling in the autophagic flux .
The degeneration of the nervous system can be viewed as a failure to maintain cell survival and function within the nervous system . In mammals , the survival of neurons during development and adulthood is actively maintained by the neurotrophic factors produced by other neurons or glias [1 , 2] . In Drosophila , neurotrophin-like proteins are secreted by neuron , muscles , and glia to maintain the survival of specific subsets of neurons during development [3–6] . The survival of glia during development can be reciprocally dependent on the trophic support from neurons . For example , in mammals , the neuregulin NRG1 , neurotrophins , transforming growth factor alpha ( TGFα ) , and purines can act on various types of glia to maintain their survival [7–10] . In the Drosophila embryonic central nervous system ( CNS ) , the survival of the longitudinal glia ( LG ) and midline glia ( MG ) are dependent on the neuregulin-like epidermal growth factor receptor ( EGFR ) ligands Vein ( Vn ) and Spitz ( Spi ) , respectively [6 , 11 , 12] . The PVR ligand PVF1 is also required for MG survival [13] . However , it is unclear whether glia survival is actively maintained in adult flies . We hypothesized that glia survival is actively maintained in the adult visual system via the gliotrophic factors secreted by the closely associated cells . Because endocytosis , which is involved in the internalization of many activated receptors , strongly affects cellular signaling outcomes [14 , 15] , blocking endocytosis should perturb these signaling events . Therefore , we expressed temperature-sensitive Shibire ( Shits1 ) , driven by the repo-GAL4 , which is expressed in most glia [16] . The shi gene is the fly homolog of mammalian dynamin [17] , which is required for multiple forms of endocytosis [18–20] , as well as vesicle recycling , which indirectly affects exocytosis [21] . Shits1 is dominant-negative at non-permissive temperatures , which thereby blocks endocytosis [19] . The use of this approach in the fly visual system enabled us to examine the gliotrophic requirements during the adult stage and precisely determine the specific cell types involved . EGFR signaling is highly conserved evolutionarily and is involved in many developmental processes [22 , 23] and pathological conditions in vertebrates [24–26] . The ligand-bound EGFR can be internalized by endocytosis . In the endosome , the EGFR can either recycle back to the cell surface or undergo lysosomal degradation [27] . The activated EGFR can signal from the cell surface and continues to signal from the early endosome before it is eventually ubiquitinated and degraded in the lysosome [28–30] . Five EGFR ligands exist in Drosophila: four agonists ( Spi , Keren ( Krn ) , Gurken ( Grk ) and Vn ) and one antagonist ( Argos ) [22] . During eye development , EGFR signaling , which is mediated by Spi and Krn , drives the progressive differentiation of multiple retinal cell types [31] . Spi is subsequently expressed in the photoreceptors and transported to the axon termini in the lamina to regulate EGFR on the lamina neurons and the differentiation of cartridge neurons [32] . The regulation and function of the EGFR ligands sent through the photoreceptor axon to their target field during eye development is well characterized [22 , 32–37] . However , the role of the EGFR ligands in the adult visual system has not been studied . Spi and Vn exert a gliotrophic function for glia in the embryonic CNS [6 , 11 , 12]; thus , we investigated whether EGFR signaling is also important in the adult visual system . Tissue degeneration may be a result of excessive cell death . The EGFR/Ras/Raf/MAPK signaling pathway can protect cells from apoptosis via direct inhibition of the pro-apoptotic protein Hid [38 , 39] . The ligand-activated EGFR can bind to the autophagy protein Beclin-1 [40] and suppress autophagy in mammals [41] . Therefore , the loss of EGFR signaling can cause either apoptosis or autophagy , which most likely depends on the cell type and cellular context [42] . We demonstrated that the adult R1-6 photoreceptor-secreted Spi acts on the lamina glia EGFR to maintain glial integrity . In the absence of the EGFR signaling , the lamina glia undergoes degeneration . Our results suggest that the primary defect caused by a lack of EGFR signaling is not apoptosis but the accumulation of autophagosomes , which subsequently leads to cell degeneration without cell loss . Therefore , our results demonstrate that the adult photoreceptors actively maintain the functional integrity of the glia in their target field . In addition , our findings indicate a novel role for EGFR signaling in the promotion of late endosome/autophagosome trafficking to lysosomes and identify a novel form of degeneration that does not involve cell loss .
We inhibited the endocytic function specifically in the glia of adult flies using a targeted expression of Shits1 , which was driven by the glia-expressing repo-GAL4 ( abbreviated repo>Shits1 ) . At the non-permissive temperature , the repo>GFP . nls and repo>H2B-RFP flies exhibited normal retina and optic lobe structures ( Fig 1A and 1C ) . The lamina in the repo>Shits1 adults were normal when cultured at the permissive temperature ( 21°C ) ( Fig 1L ) ; however , they exhibited vacuoles in the optic lamina two days after a shift to the non-permissive temperature ( 28°C ) ( Fig 1B and 1D ) . The phenotype progressively worsened , and 5% of the lamina volume became vacuolated at 14 days ( Fig 1K ) . When the repo>Shits1 flies were shifted to 28°C for 12 days and then shifted to 17°C for 9 days , the vacuolization phenotype was not reversed ( Fig 1L ) . Thus , blocking Shi function in the glia causes an irreversible and progressive degeneration of the optic lamina . We next examined the specific cell types that were affected by vacuolization . The optic lamina possesses six distinct glia cell types , namely , fenestrated glia , distal satellite glia , proximal satellite glia , epithelial glia , marginal glia , and chiasm glia [43] . The location of the vacuoles correlated with the location of the epithelial glia and , to a lesser extent , the marginal glia . Shits1 expression driven by an epithelial glia-specific HisCl-Gal4 ( Fig 1E ) or a marginal glia-specific NP2109-Gal4 ( Fig 1F ) also caused a weak lamina vacuolization ( Fig 1G and 1H ) . We used the MARCM method [44] to clonally express Shits1 and GFP in glial cells . At 21°C , the MARCM clones did not exhibit defects ( Fig 1I ) . At 29°C , of 70 MARCM clones , 28 clones exhibited vacuoles , which can be detected within a single cell clone ( Fig 1J ) . We further examined the phenotype using electron microscopy ( EM ) . In the wild type adult optic lamina , one lamina cartridge contains five lamina neurons , with the L1/L2 terminals in the center , surrounded by six photoreceptor terminals , which are then surrounded by epithelial glia ( Fig 2A ) . In the repo>Shits1 lamina , small and large vacuoles were identified within the electron-dense glial cytoplasm , and the R cell axons were enlarged but contained no vacuole ( Fig 2B ) . Most vacuoles appeared empty , with only a few vacuoles that contained double membrane structures ( Fig 2C and 2C’ ) . We also observed double membrane autophagosome-like structures [45] within the cytoplasm ( Fig 2D ) . These results suggest that blocking Shi function in the lamina glia caused a cell-autonomous vacuolization . The neural response to a light pulse was measured by electroretinogram ( ERG ) , which is composed of an “ON” transient , a depolarization , and an “OFF” transient ( S1A Fig ) . The depolarization measures the transmission within the photoreceptor axons , whereas the ON and OFF transients measure the synaptic transmission from the photoreceptor neurons to the lamina neurons [46 , 47] . We demonstrated that the repo>Shits1 flies exhibit a normal depolarization but a loss of the ON and OFF transients on the ERG on day 3 ( S1B–S1D Fig ) . This result suggests that while the neural transmission along the photoreceptor axon is normal , the synaptic transmission from the photoreceptor neurons to the lamina neurons is defective . Because the lamina synaptic region is wrapped by epithelial glia , which is known to recycle the neurotransmitters from the photoreceptors [48–50] , the synaptic transmission defect is most likely a result of an epithelial glia dysfunction . Because endocytosis is involved in many signaling pathways in the receiving cells , the lamina glia may receive a gliotrophic signal via endocytosis . One potential source for the gliotrophic factor may be the photoreceptors , since their axons form synaptic contacts with both the monopolar lamina neurons and the epithelial glia in the lamina cartridge [51] . We demonstrated that the expression of Shits1 using a R1-6 photoreceptor-specific Rh1-GAL4 ( Fig 3A ) caused a lamina vacuolization ( Fig 3B and 3L ) similar to the repo>Shits1 flies . Dynamin is also required for vesicle recycling [21]; thus , the loss of Shi function could affect the vesicle recycling , which leads to the loss of ligand secretion , as demonstrated for Wg secretion [52] . In the Rh1>Shits1 flies , the structure of the lamina cartridge of the photoreceptor axons was disorganized , and the lamina neuropile contained vacuoles in the epithelial glia layer ( Fig 3I and 3K ) . In the glial nuclei layer , the vacuoles formed near the nuclei ( Fig 3I ) . A glial nucleus is squeezed by a large vacuole to become adjacent to another glial nucleus ( arrow in Fig 3K compared with 3J ) . When the expression was driven by the R7/8-specific Pan-Rh7-Gal4 ( Fig 3C ) , no lamina vacuolization was identified ( Fig 3D and 3L ) . A specific lamina L2-5 neuron Ln-GAL4-driven expression , combined with a repo-GAL80 to block the Ln-GAL4 activity in the satellite glia ( Fig 3E ) , did not cause lamina vacuolization ( Fig 3F and 3L ) . Furthermore , when the R1-6 photoreceptors were killed via the expression of the apoptotic gene hid , lamina vacuolization was induced ( Fig 3G and 3L ) . We also ablated the photoreceptors in a different manner . The rhodopsin protein phosphatase RdgC is expressed in the retina and ocelli , and the rdgC306 mutant exhibits normal lamina morphology at birth but a light-dependent retinal degeneration [53 , 54] . The rdgC306 mutant exhibited degeneration in the lamina and retina after constant illumination for 14 days ( Fig 3H and 3L ) . These results indicate that R1-6 photoreceptors are required for lamina glia vacuolization . EGFR , which is internalized by endocytosis and continues to signal from the early endosome , is required for glia survival in the embryonic CNS [6 , 11 , 12]; thus , we investigated whether EGFR signaling acts in the adult lamina glia to maintain the glia . To specifically drive expression in adult glia , we combined the repo-GAL4 with tub-Gal80ts ( abbreviated as repots ) . In these flies , GAL4 activity is suppressed by the GAL80ts at the permissive temperature , and a shift to the non-permissive temperature after eclosion induces GAL4 activity . The coexpression of a constitutively active form of EGFR ( repots>Shits1+Egfrλtop4 . 2 ) suppressed the repots>Shits1 vacuolization phenotype ( Fig 4A and 4B and 4G ) . The phenotype could also be rescued via the coexpression of an active form of the fly MAPK Rolled ( repots>Shits1+RlSem ) ( Fig 4C and 4G ) or a heterozygotic combination with the gain-of-function allele rlSem ( Fig 4H ) . These results suggest that EGFR/MAPK signaling is sufficient to maintain glial integrity , and the vacuolization phenotype was not a result of the EGFR trapped at the cell surface , but rather a loss of signaling . Conversely , the expression of a dominant-negative Drosophila EGFR ( DERDN ) in the glia ( repots>DERDN ) caused a similar lamina vacuolization as in the repots>Shits1 flies ( Fig 4D and 4G ) . In the Egfrco mutant flies , vacuoles could be identified within the clones ( 70/183 in Fig 4F compared with 45/48 in Fig 4E ) . These results suggest that EGFR signaling is cell-autonomously required in the lamina glia to maintain their integrity . What is the gliotrophic signal produced by the photoreceptors ? Based on RNA microarray data , spitz , Keren , and vein , but not gurken , are expressed in the adult eye [55] . The Spi protein can be predominantly detected in the adult retina and as puncta in the lamina ( Fig 5A ) . The targeted expression of full length Spi ( mSpi-GFP ) [56] in photoreceptors ( GMR>mSpi-GFP ) exhibited a strong GFP signal in the retina and a weak signal in the lamina neuropile , where the photoreceptor axons terminate ( Fig 5B ) . These results indicate that Spitz expressed from the photoreceptors can be transported from the retina to the lamina . The knockdown of both Spi and Krn in the photoreceptor cells also caused lamina vacuolization ( Fig 5C and 5G ) . Although the severity of the repots>Shits1 fly phenotype was not affected by a reduction in the dosage of Egfr , spi or Krn , it was strongly enhanced in spi and Krn double-heterozygous mutants ( Fig 4H ) . These results suggest that the EGFR ligands Spi and Krn are redundantly required in the photoreceptors to prevent lamina vacuolization . The EGFR ligands Spi , Krn and Grk are synthesized as membrane-bound precursors and must be transported by the chaperone Star and cleaved in the ER by the intramembranous protease Rhomboid ( Rhom ) to acquire their active secreted form [34] . We generated whole-eye rho7M43 ru1 clones that have double null mutations for rhom-1 ( rho ) and rhom-3 ( also referred to as roughoid , ru ) [57] and identified lamina vacuolization in these mutants ( Fig 5E ) . iRhom is an inactive Rhomboid-like pseudoprotease that promotes the degradation of EGFR ligands in the ER [58] . We expressed iRhom in the retina to promote the degradation of EGFR ligands in the signal-producing cells . Lamina vacuolization was identified in the Rh1>iRhom flies ( Fig 5D and 5G ) . Rab11 is required for Spitz secretion in the larval photoreceptors [35] . The expression of a dominant-negative Rab11S25N in the R1-6 photoreceptors caused a mild lamina vacuolization ( Fig 5F and 5G ) . These data indicate that the transport , processing and secretion of the EGFR ligands is required in the R1-6 photoreceptors to maintain lamina glial integrity , which suggests that the R1-6 photoreceptor neurons are the source of EGFR ligands . The previous results suggested that the EGFR ligand Spi secreted by the photoreceptors can be transported to the lamina and activate EGFR in the lamina glia . Spi can be found in the photoreceptor axons in the lamina and colocalizes , in part , with Black , an aspartate decarboxylase specific for the cytoplasm of epithelial glia cells [50] ( Fig 6A ) . This result suggests that Spitz can be secreted from the photoreceptor axons and internalized in the epithelial glia . The EGFR target pointed-lacZ can be used as a reporter for EGFR signaling [59–61] and was expressed in the lamina epithelial and marginal glial cells ( Fig 6B ) . The pnt-lacZ expression in the lamina was lost after the blockade of EGFR signaling ( in repots>DERDN; Fig 6C ) or endocytosis ( repots>Shits1; Fig 6D ) in the glia . When the Spitz expression was knocked down in the R1-6 photoreceptors ( in Rh1>Dcr2+Spi-RNAi ) , the pnt-lacZ expression was lost in the lamina glia ( Fig 6E ) . These results demonstrate that EGFR signaling is active in the lamina glia and is dependent on Spi produced by the R1-6 photoreceptors . We next addressed the cellular basis of lamina glia vacuolization . Apoptosis , which was assessed by activated Caspase-3 and TUNEL assays , was not identified in the repots>Shits1 and repots>DERDN flies ( Fig 7A–7C ) compared with the control experiments ( S2A and S2B Fig ) . We also used Apoliner , which is an in vivo fluorescent sensor for activated caspases [62] that contains a caspase cleavage site flanked by a membrane-targeted RFP and a nuclear-targeted GFP . The nuclear GFP is typically retained at the cell membrane by tethering to the mRFP ( Fig 7D ) ; however , it relocalizes into the nucleus following the caspase site cleavage . The coexpression of Apoliner with Shits1 or DERDN resulted in a perinuclear distribution of the GFP that was colocalized with mRFP ( Fig 7E and 7F ) , which indicates that Caspase-3 was not activated . The repots>Shits1 vacuolization phenotype was not rescued by the coexpression of the anti-apoptotic proteins P35 [63] or Diap1 [64–66] , even at 21 days ( Fig 7G ) . Consequently , the vacuolization is most likely not a result of apoptosis . The vacuolization does not involve a reduction in the number of epithelial glia cells , as demonstrated by the nuclear RFP signal in the repo>Shits1+H2B-RFP flies ( Fig 7H ) . This finding suggests that the vacuolization affects the glial cell body without causing cell loss . This finding is consistent with our EM results that demonstrated the nuclei in the vacuolated glia are intact ( Fig 2E ) . A defect in lipid metabolism homeostasis can be involved in neuronal or glial degeneration [67] . However , we found no apparent change in lipid accumulation in the repots>Shits1 lamina ( S3B Fig compared with S3A Fig ) . Autophagy-like vesicles that encapsulated bulk cargo and organelles were identified in the glia of repo>Shits1 and repo>DERDN flies ( Fig 2D’ and 2E’ ) . We subsequently assessed the levels of the autophagy markers GFP-LC3 [68] and Ref ( 2 ) P , the Drosophila ortholog of p62 [69] . Atg8/LC3 requires activation via proteolytic cleavage by Atg4 and is subsequently conjugated to phosphatidylethanolamine by Atg7 and Atg3 . Therefore , Atg8/LC3 overexpression in the fly does not enhance autophagy [70] and is generally used as an inconspicuous marker of autophagy . In the repots>GFP-LC3+DERDN adult flies shifted to 28°C , the GFP-LC3 puncta became detectable on day 2 ( Fig 8B compared with Fig 8A ) . Ref ( 2 ) P typically binds to LC3 and is degraded in the autolysosomes; however , it accumulates in the presence of autophagosome-lysosomal trafficking defects and neurodegenerative diseases [71–74] . In the repots>DERDN flies , the Ref ( 2 ) P signal was weak on day 1 ( Fig 8A ) ; however , it accumulated in the glia and colocalized with the LC3-GFP puncta between days 2 and 3 ( Fig 8B and 8C ) . The accumulation of Ref ( 2 ) P was also identified cell-autonomously in the Egfrco mutant glial clone , which suggests that the repo>DERDN effect is a result of the loss of EGFR signaling rather than an effect of DERDN ( Fig 8D ) . We further examined the specific compartment in which the autophagosomal cargo accumulated . The double-tagged GFP-mCherry-Atg8a contains mCherry , which is resistant to the low pH in the lysosome , and GFP , which is quenched in the lysosome . Therefore , this tag can be used to distinguish the autophagosomes ( GFP and mCherry , yellow ) from the autolysosomes ( mCherry , red ) during autophagic flux ( Fig 8E ) [75 , 76] . In the repots>Shits1 and repots>DERDN flies , the induced puncta signal predominately appeared in the autophagosomes versus the autolysosomes ( Fig 8G and 8H compared with 8F ) . These results indicate that EGFR signaling in the glia promotes the fusion of autophagosomes to lysosomes . The absence of EGFR signaling caused a failure in Atg8 and Ref ( 2 ) P degradation and resulted in their accumulation in the autophagosomes . These results suggest that autophagy may contribute to glial vacuolization . Autophagy gene dAtg1 expression in the glia ( repots>dAtg1 ) induced a similar lamina vacuolization phenotype compared with the repots>Shits1 flies ( S4A Fig ) . When repots>DERDN adults were treated with the autophagy inhibitor 3-methyladenine ( 3-MA ) , the vacuolization was partially rescued ( S4B Fig ) . The repots>Shits1 and repots>DERDN vacuolization phenotypes , which were repressed by the coexpression of the autophagy induction blocker dTOR [77] , were repressed by reducing Atg1 and Atg13 ( S4C and S4D Fig ) , and enhanced by coexpressing an activated form of the autophagy-promoting S6K ( S6KSTDETE ) ( S4E and S4F Fig ) . A knockdown of the autophagy proteins Atg5 , Atg7 , and Atg12 alleviated the vacuolization phenotype of the repots>Shits1 flies; however , it was not sufficient to rescue the stronger phenotype of the repots>DERDN flies ( S4E and S4F Fig ) . These results indicate that autophagy is , at least in part , responsible for the glia vacuolization phenotype . We next examined the effect on GFP-LAMP1 , which is targeted to the membrane of the late endosome/lysosome and subsequently degraded in the mature lysosomes [78 , 79] . When GFP-LAMP1 was expressed in the glia ( repots>GFP-LAMP1 ) , the GFP signal was weak ( Fig 9A–9C ) . In the repots>Shits1 and repots>DERDN flies , the GFP-LAMP1 signal was significantly increased in the lamina after only 12 h at 28°C and was strongly accumulated on day 2 ( Fig 9D–9I ) . The increased GFP-LAMP1 accumulation identified in the repots>Shits1 flies was reduced when the EGFR signaling was enhanced using a gain-of-function allele rlSem/+ and was enhanced when the doses of the EGFR ligands Spi and Krn were reduced ( S5D–S5G Fig ) . In all conditions , the severity of lamina vacuolization correlated with the GFP-LAMP1 intensity ( Figs 4G , 4H , S5G , S5H ) . The early and strong accumulation of GFP-LAMP1 also suggests that the impairment of the lysosomal system may be the primary cause of the glia vacuolization . In the repots>Shits1 and repots>DERDN flies , the late endosome marker Rab7-mCherry also accumulated as puncta in the lamina ( Fig 9K and 9L compared with 9J ) . Taken together with the accumulation of GFP-LC3 , Ref ( 2 ) P and LAMP1-GFP , these results suggest that the trafficking or the fusion of the late endosome and autophagosome to the lysosome is blocked . The accumulation of the autophagosomal proteins GFP-LC3 and Ref ( 2 ) P may be a result of a failure in lysosomal degradation or autophagosome-lysosomal trafficking . Feeding the repots>Shits1 and repots>DERDN flies with chloroquine , which inhibits lysosomal acidification and degradation [80 , 81] , did not affect the LAMP1-GFP phenotype ( S6 Fig ) . This result suggests that the GFP-LAMP1 accumulation could be because of a block at a step upstream of lysosomal degradation . In this case , a block downstream of lysosomal degradation would not affect the upstream blockage . The overexpression of the apoptotic protein Hid did not induce GFP-LAMP1 accumulation , vacuolization , or autophagy accumulation in the glia ( S5G and S2C and S2D Figs ) , which suggests that the lysosomal defect in the glia is not a response to apoptosis . While the overexpression of the autophagy gene dAtg1 in the glia caused lamina vacuolization ( S4E Fig ) , it did not cause GFP-LAMP1 accumulation ( S5C and S5G Fig ) , which suggests that autophagy is not induced upstream of the lysosomal defect . Because the autophagy marker GFP-LC3 was increased only 2 days after blocking EGFR signaling ( Fig 8B ) , these results suggest that autophagy is a late event in glia vacuolization and may be a secondary response or independent of the lysosomal impairment . Considering that these results demonstrated that blocking an early step of the endocytic pathway in the repo>shits1 flies caused vacuolization , we investigated other components of the vesicle trafficking pathways . Rab5 is required for the fusion of the endocytic vesicles with the early endosome [82] . The expression of a dominant-negative Rab5 ( Rab5S43N ) [83] in the adult glia ( repots>Rab5S43N ) caused lamina vacuolization ( Fig 10A and 10K ) and enhanced the GFP-LAMP1 signal ( S5A and S5G Fig ) via similar effects as the phenotype observed in the repo>Shits1 flies ( Figs 4A and 9F ) . Rab52 mutant MARCM clones also exhibited lamina glia vacuolization ( Fig 10E ) . α-Adaptin ( α-Ada ) is a subunit of the AP-2 complex , which is required for endocytosis [84] . Vacuoles could be identified in the lamina glia of the α-Ada3 mutant clones ( Fig 10F ) . These data suggest that the early steps of endocytosis , which involve Shi , Rab5 and Ada , are required for lamina glia maintenance . Activated EGFR is endocytosed and continues to signal from the early endosome [29]; thus , these results suggest that EGFR signaling from the early endosome is important to prevent vacuolization of the lamina glia . We also examined other steps involved in vesicle trafficking . Hrs is required for the transition from the early endosome to the late endosomes or multivesicular bodies ( MVB ) [29] . The HrsD28 homozygous mutant clones did not exhibit vacuolization ( Fig 10G ) . Rab7 is required for the docking of the early endosome to the late endosome , as well as the fusion of the late endosome and autophagosome with the lysosome [85 , 86] . The expression of a dominant negative form of Rab7 ( Rab7T22N ) did not cause lamina vacuolization ( Fig 10K ) or GFP-LAMP1 accumulation ( S5B and S5G Fig ) [87] . Because the endolysosomal conversion was not affected by Rab7T22N , which suggested that this mutant could not be a dominant-negative form [88] , a Rab7KO mutant clone was generated and did not exhibit vacuolization in the lamina ( Fig 10H ) . Rab11 is required in recycling endosomes and promotes the fusion of late endosomes or MVBs with autophagosomes [89 , 90] . The expression of the dominant-negative Rab11S25N in the glia caused vacuolization [91] ( Fig 10C and 10K ) , which indicates that either recycling endosomes or autophagosome maturation may also be involved in the maintenance of cell integrity . Our results suggest that the vesicle trafficking steps that involve Hrs and Rab7 are not required to prevent lamina glia vacuolization . This finding was consistent with the lack of EGFR signaling from the late endosomes [29] . The class C vacuolar protein-sorting ( Vps ) complex plays a role in vesicle sorting and trafficking between different vesicular compartments . Deep orange ( Dor ) and Carnation ( Car ) are subunits of the Vps-C complex and are involved in the trafficking between late endosomes and lysosomes [78 , 92 , 93] . The depletion of Dor and Car in the fat body caused autophagosome accumulation [94 , 95] . Therefore , we assessed whether dor and car were involved in glia vacuolization . We identified a high frequency of vacuolization in the dor or car mutant glial clones ( Fig 10I and 10J ) . Although knockdown of Dor or Car alone in the glia did not cause vacuolization ( Fig 11J and 11K ) , it enhanced lamina vacuolization in the repots>DERDN flies ( Fig 11B , 11E , 11J , 11K compared with 11A and 11D ) . Surprisingly , glial vacuolization and GFP-LAMP1 accumulation in the repots>DERDN flies were also slightly enhanced by the coexpression of wild-type Dor or Car ( Figs 11C and 11F and S7 ) , although the expression of Dor or Car in the wild-type did not cause a defect ( Fig 11J and 11K ) . Both a reduction and increase in the dosage of Dor or Car enhanced the repots>DERDN flies vacuolization phenotype; thus , these results suggest that a proper balance in the expression of the Vps-C complex components is essential for glia maintenance . The knockdown of both Car and Dor strongly enhanced lamina vacuolization and Ref ( 2 ) P accumulation in the repots>DERDN flies ( Fig 11H and 11L compared with 11G ) . By coexpressing both Car and Dor in the repots>DERDN flies , both vacuolization and Ref ( 2 ) P accumulation were rescued ( Fig 11I and 11L ) . Although we cannot exclude the possibility that EGFR signaling may act in parallel to the Vps-C complex , these genetic interactions suggest that EGFR signaling acts at a step upstream of Dor/Car in the promotion of the autophagic flux .
Presynaptic and postsynaptic neurons can mutually maintain the survival of their synaptic partners . During development , neurons can also provide gliotrophic factors to maintain glia survival . The majority of human neural degeneration exhibits a late onset and progresses over time; thus , the major concern is the maintenance of cell survival or function . For hereditary neural degenerations or genetically manipulated animal models of neural degeneration , it is typically difficult to separate the developmental effects from the true maintenance requirement in adults . Our experimental approach specifically bypassed the development and examined the events at the adult stage , which therefore addresses the maintenance of the adult visual system in a manner more relevant to human nervous system degeneration . Our results demonstrate for the first time that the adult photoreceptor neurons actively maintain the integrity of glia within their target field in the optic lamina . We demonstrated that in the adult visual system , the R1-6 photoreceptors produce and transport the EGFR ligand Spi , and presumably Krn , to the axon termini in the optic lamina to act on the EGFR in the lamina epithelial and marginal glia to maintain integrity . Spi and Krn are the first gliotrophic factors demonstrated to act in the adult nervous system . Because of the advantages offered by the fly visual system , we were able to clearly define the source and recipient cell types for the gliotrophic signal . Photoreceptors secrete gliotrophic factors most likely to maintain the functional integrity of their microenvironment and its synapses . The epithelial glia is involved in the reuptake of neurotransmitters from the synaptic cleft and their metabolism . In the absence of EGFR signaling , the lamina glia undergoes a progressive and irreversible vacuolization , which is accompanied by a defect in photoreceptor synaptic transmission . Interestingly , this degeneration is not because of apoptosis and does not involve cellular losses . This conclusion is based on the following observations: ( 1 ) there was no apparent loss of Repo+/DAPI+ nuclei number in the epithelial glia layer in the degenerating lamina , ( 2 ) there was no apoptotic signal ( assessed by anti-activated caspase 3 , TUNEL assay , and Apoliner ) in the degenerating lamina , ( 3 ) the glia nuclei in the degenerating lamina appeared intact ( assessed by Repo , DAPI staining and EM ) , ( 4 ) the coexpression of the anti-apoptotic P35 or Diap1 failed to rescue the phenotype , and ( 5 ) the repots>hid flies did exhibit lamina vacuolization or autophagy ( GFP-Cherry-Atg8a ) . Thus , adult lamina glia degeneration represents a new type of cellular degeneration with the loss of cellular integrity and function , but without the loss of cell number . Most studies have focused on neurons in these degenerative conditions . We now provide a model system in which the glial cells are the primary degenerating cells . It would be interesting and important to determine whether the gliotrophic maintenance is also at play when the nervous system is damaged by trauma or other pathological conditions , as demonstrated for the response to injury in the larval ventral nerve cord [96] . EGFR ligand-binding on the cell surface activates the receptor and results in the transduction of a signal into the nucleus . The ligand-bound receptor becomes internalized by endocytosis . Internalized EGFR can exhibit a sustained level of activation and signaling from the early endosome [28–30] . In our study , endocytosis is blocked in the repo>Shits1 flies , which presumably results in more activated EGFR at the cell surface . This effect caused glia degeneration , which suggests that the cell surface EGFR signaling is not sufficient to maintain glial integrity . However , the repo>Shits1 phenotype could be rescued by the coexpression of activated EGFR , which would remain on the cell surface because endocytosis is blocked by Shits . This rescue indicates that increased cell surface EGFR signaling can replace the missing EGFR signaling from the early endosomes . Therefore , EGFR signaling from the two compartments , namely , the cell surface and early endosome , are qualitatively the same and may only be different in terms of signaling intensity ( S6 Fig ) . EGFR can signal via multiple mechanisms [97] . The membrane-bound EGFR can signal via its tyrosine kinase activity through the Ras-Ref-MEK-MAPK , PI3K- Akt-mTOR , PLC-γ-PKC , and Jak2-STAT3 pathways . EGFR can also signal via kinase-independent mechanisms most likely through interactions with other proteins [97] . Our results demonstrated that in the lamina glia , EGFR signals through the MAPK pathway . Ligand-activated EGFR can also enter the cells and exert certain functions in the nucleus and mitochondria [97] . The nuclear transport of EGFR requires endocytosis [98 , 99] . Whether the mitochondrial transport of EGFR requires endocytosis is controversial [100 , 101] . The nuclear and mitochondrial transport of EGFR has not been reported in Drosophila . We demonstrated that the early endocytic steps that involved Shi , Rab5 , and α-Ada were required to prevent lamina glia degeneration , which suggests that the internalized EGFR signals from the early endosome . However , we cannot exclude the possibility that EGFR signals from the nucleus or mitochondria because blocking endocytosis would also block the nuclear transport , and possibly the mitochondrial transport , of activated EGFR . In mouse cortical astrocytes and Drosophila embryonic CNS glia , the absence of EGFR signaling leads to glia apoptosis [6 , 12 , 102] . Our findings demonstrate that in the adult lamina , the absence of EGFR signaling triggers a different type of cellular degeneration , which is independent of apoptosis . The same Spi signal from the same photoreceptors is transported to the lamina and exerts different functions in each developmental stage . Spi acts on the lamina neurons during the larval stage for the differentiation of cartridge neurons [32] , whereas it acts on the lamina glia in the adult for their maintenance . There is no report that links EGFR signaling and autophagy in Drosophila . Our results suggest that the vacuolization is , at least in part , a result of autophagy . In cancer treatment with anti-EGFR antibodies and small molecule drugs that inhibit EGFR tyrosine kinase activity , autophagy is often induced [42] . This finding suggests that EGFR signaling can inhibit autophagy in the lamina glia . The mammalian EGFR can bind directly to the autophagy regulator Beclin-1 and inhibit autophagy [41] . It is unknown whether , in the fly , EGFR can also bind to and phosphorylate Atg6 , the Drosophila Beclin-1 homolog . EGFR can also prevent autophagy via interaction with the sodium/glucose cotransporter 1 ( SGLT1 ) in a kinase-independent manner to maintain the intracellular glucose level [103] . It is unknown whether a similar mechanism also operates in Drosophila . Our findings may be the first to link EGFR signaling to autophagy in Drosophila . Blocking EGFR signaling in the glia caused several phenotypes . The accumulation of GFP-LAMP1 occurred 12 h after shifting to the non-permissive temperature . The ERG was normal on day 1; however , the ON/OFF transients were completely absent on day 3 . The lamina vacuoles were noticeable on day 2 and became progressively more apparent . The autophagy marker GFP-LC3 increased on day 2 . Because the accumulation of GFP-LAMP1 was the earliest and strongest effect , we suppose that this finding reflects the primary cause of the degeneration . Our results suggest that EGFR signaling is required for proper vesicle trafficking from the late endosome and autophagosome to the lysosome ( S6 Fig ) . A failure at the fusion step of the late endosome or autophagosome to the lysosome caused the accumulation of autophagosomes and increased GFP-LC3 in the fly [92 , 94 , 95 , 104] , as well as in certain mammalian lysosomal storage diseases [105 , 106] . The accumulation of autophagosomes may cause cellular degeneration perhaps because of the accumulation of certain proteins , typically destined for degradation , that become toxic to the cell and trigger autophagy [107] . Although we propose that the autophagy is a secondary cause of the failure in the autophagosome-lysosome fusion , we do not exclude the possibility that the loss of EGFR signaling could independently enhance autophagy . Our findings are the first study to link EGFR signaling with the trafficking from the late endosome and autophagosome to the lysosome . EGFR signaling is increased in many cancers . Fifty to sixty percent of primary glioblastoma tumors exhibit increased EGFR signaling [108] . The EGFR signaling pathway has been a major therapeutic target for various types of cancer , including glioblastoma [109 , 110] . The level of EGFR signaling must be well balanced because too much signaling can lead to oncogenic growth , whereas too little signaling may lead to glia degeneration , as demonstrated by our study . Therefore , our study highlights the caution needed in the therapeutic treatments that act via a reduction of EGFR signaling .
Fly culture and crosses were performed according to standard procedures at 25°C unless otherwise noted . The fly stocks ( repo-Gal4 , GMR-Gal4 , point1277-lacZ , longGMR-Gal4 , ey3 . 5-FLP , tubGAL80ts , UAS-GFP . nls , UAS-lacZ , UASp-GFP-mCherry-Atg8a , UAS-Apoliner , UAS-DsRed , UAS-Hid , FRT19A tubP-GAL80 hs-FLP; UAS-mCD8-GFP , FRTG13 tubGal80 , FRTG13 UAS-GFP , FRT80B tubGAL80 , FRT42D tubGAL80 , FRT40A tub-GAL80 , UAS-Rab5S43N , UAS-Rab7T22N , UAS-Rab11S25N , UAS-dTorWT , UAS-S6KSTDETE , GMR-wIR and rdgC306 ) were obtained from the Bloomington Stock Center . The rlSem was obtained from the Drosophila Genetic Resource Center . The UAS-Spitz-RNAi ( KK103817 ) and UAS-Keren-RNAi ( GD27110 ) were obtained from the Vienna Drosophila Research Center . The UAS-Dor-RNAi ( 3093R-4 ) and UAS-Car-RNAi ( 12230R-1 ) were obtained from the NIG-FLY . The Rh1-GAL4 UAS-lacZ was provided by Larry Zipursky . The UAS-mCherry-Rab7 was provided by Jui-Chou Hsu . The repo-GAL4 , UAS-mRFP was provided by Yuh Nung Jan , and the UAS-P35 was provided by Bruce Hay . The following stocks were provided by the original authors: Ln-GAL4 [111] , repo-FLP [112] , repo-GAL4 UAS-CD4-mtdTomato [113] , repo-GAL80 [114] , UAS-H2B-RFP [115] , UAS-Shits1 [19] , UAS-Egfrλtop4 . 2 [116] , UAS-DERDN [117] , UAS-RlSem [118] , UAS-mSpiGFP [56] , UAS-iRhom [58] , UAS-dAtg1 and UAS-Atg1-RNAi [119] , UAS-Atg5-RANi , UAS-Atg7-RNAi and UAS-Atg12-RNAi [77] , Egfrco [120] , spiOE92 [121] , Krn27-7-B [122] , rho17M43 ru1 [123] , dor8 and UAS-Dorwt [124] , carΔ146 , UAS-GFP-LAMP1 and UAS-Car [92] , atg13Δ81 [125] , Rab52 [82] , Rab7KO [126] , α-Adaptin3 [84] , HrsD28 [29] , UAS-GFP-LC3 [68] , repo-FLP repo-GAL4 UAS-actGFP; FRT82B tubGAL80 [112] . The repo-Gal4 and tubGAL80ts were recombined into repo-GAL4 tubGAL80ts ( repots-GAL4 ) on the third chromosome . The recombinant lines were selected by crossing with UAS-Hid . The repo>Hid is larva-lethal at room temperature; however , it is viable with tubGal80ts . The recombinant of repo-GAL4 UAS-Shits1 was selected by the lethality feature at 30°C for 7 days . The genotypes for the MARCM clone generation were as follows: FRT42D tubGAL80/FRT42D Egfrco; repo-GAL4 UAS-mtdTomato/repo-FLP , hs-FLP/+; FRTG13 tub-Gal80/FRTG13 UAS-mCD8GFP; repo-GAL4/UAS-Shits1 , FRT19A dor8/FRT19A tubGAL80 hs-FLP; UAS-mCD8GFP; repoGal4/+ , FRT19A carΔ146/FRT19A tubGAL80 hs-FLP; UAS-mCD8GFP; repoGal4/+ . Forty-eight h after egg laying , the animals were heat-shocked for 90 min at 37°C . The whole eye rho7M43 ru1 double mutant clones were generated from ey3 . 5-FLP/UAS-DsRed; GMR-GAL4/+ FRT80B rho7M43 ru1 /FRT80B tubGAL80 for 12 days . The crosses and flies were maintained at 17 or 21°C ( permissive temperature ) until adult eclosion . The adults ( 3–7 days old ) were shifted to a restrictive temperature ( 28 or 29°C ) to enable transgene expression for the indicated time . The fixed fly heads were dehydrated in series of ethanol/ddH2O steps , embedded in wax , and sectioned in paraffin blocks at 5–7 μm thickness . The sectioned head slides were deparaffinized with xylene and rehydrated in a series of ethanol/ddH2O . The slides were immersed in hematoxylin ( Thermo Fisher Scientific ) for 2 min and eosin ( Thermo Fisher Scientific ) for 5 min . Permount was added on the slides , which were imaged on a Zeiss AxioImager-Z1 microscope equipped with Plan Apo 20X DIC II and Plan Apo 40X DIC III immersion objectives . GMR-wIR is a White-RNAi driven by a GMR enhancer to reduce the autofluorescence from the retinal pigments . For cryosectioning , adult flies were fixed in 4% paraformaldehyde for 3 h at room temperature . The fly heads with proboscis were removed and incubated in 1x PBS that contained 25% sucrose at 4°C for 24 h and embedded in OCT compound ( Tissue-Tek , Sakura ) . The solidified samples were sliced at a 100-μm thickness using a Leica LX2501 cryostat . The slices were incubated with the following primary antibodies: mouse anti-Repo ( 1:100 ) , rat anti-Spitz ( 1:50 ) ( Developmental Studies Hybridoma Bank ) , rabbit anti-β-Gal ( 1:500; Cappel ) , rabbit anti-Cleaved Caspase-3 ( Asp175 , 1:200 , Cell Signaling ) , rabbit anti-full length Ref ( 2 ) P ( 1:300 , a gift from Tor Erik Rusten ) , and guinea pig anti-Black ( 1:500 , a gift from Bernhard Hovemann ) [50] . The fluorescent secondary antibodies ( 1:200 ) were obtained from Jackson ImmunoResearch . DAPI ( 25 ng/ml , Sigma ) was used to stain the DNA and tissue background . Immunolabeled slices were mounted in FocusClear ( CelExplorer Labs ) and imaged on a Zeiss LSM 510 Meta confocal microscope . The severity of glial vacuolization in the lamina was quantified by outlining the vacuoles in the lamina . The area of the vacuole and lamina of each brain hemisphere was scored using Metamorph software ( Molecular Devices ) . The measurement of GFP-LAMP1 fluorescence by image analysis generates intensity values that range from 1 to 255 using Metamorph software . The intensity of the collected images was assessed below the saturation level . The GFP intensity of each pixel in the lamina neuropile that was greater than the lower threshold ( intensity value ≥25 ) , as defined by the background , was averaged and expressed as the percentage of the mean values of the control genotype . For counting glial cell numbers , we used only females to avoid the differences in body size and sexual dimorphism in the brain . The lamina of 4°C cold-shocked adults were dissected , fixed in 4% paraformaldehyde at 4°C for 30 min , and imaged by Z-stacks of confocal images . The number of epithelial glial nuclei was examined by manually counting the nuclear RFP in the epithelial layer using Metamorph software . All data are presented as the means ± sem . The P-values of the multiple comparisons were obtained by one-way ANOVA for the normally distributed data and Kruskal-Wallis tests for the non-normally distributed data . The P-values of the two data sets were tested by unpaired Student t-tests for the normally distributed data and Mann—Whitney tests for the non-normally distributed data using GraphPad Prism software v5 . Values of P<0 . 05 compared with the control group were considered statistically significant . *P<0 . 05 , **P<0 . 01 , ***P<0 . 001 . n . s . : not significant . The N is indicated in the figures . The In Situ Cell Death Detection Kit ( TMR red ) was performed according to the user manual ( Roche ) . Adult head sections for TEM were prepared as previously described [127] . Adult flies ( 3–7 days old ) were pretreated with 5 mM of 3-Methyladenine ( 3-MA ) or 1 mg/ml of Chloroquine ( CQ ) in 2% sucrose on tissue papers for 1 day at 17°C , followed by a temperature shift to 29°C for 4 and 2 days , respectively . During the incubation , the papers were kept moist and replaced once every 2 days . Seven to eight adults of each genotype at the indicated age were placed in yellow tips , which were fixed by nail oil on the tip and left eye . The recording electrode touched on the surface of the right eye , and the ground electrode was on the head capsule . The flies were adapted in the dark for 30 s and stimulated by a 1-s 5000 Lux light pulse ( Apex Monochromator Illuminator , 150 W Xenon Arc , Newport ) . The electrophysiological data were recorded via a microelectrode amplifier ( Axonclamp 900A , Molecular Devices ) . The results were acquired using a data acquisition system ( Digidata1440A , Molecular Devices ) and analyzed using pClamp 10 software ( Molecular Devices ) . Cryosectioned fly heads were post-fixed in formal calcium ( 0 . 01 mg/ml CaCl2 in 4% paraformaldehyde pH 4 . 0 ) for 1 h and rinsed in deionized H2O and 50% isopropanol for 5 min . The slides were stained in an Oil Red O working solution ( 3 mg/ml Oil Red O in 60% isopropanol ) for 6 min and rinsed in deionized H2O and 50% isopropanol for several seconds . The slides were stained by hematoxylin for 3 min ( for nuclei staining ) , and the images were captured on an AxioImager-Z1 microscope ( Zeiss ) equipped with Plan Apo 20X DIC II and Plan Apo 40X DIC III immersion objectives . | Degeneration of the nervous system can be viewed as a failure to maintain cell survival or function in the nervous system . The late onset of neurodegeneration in humans indicates that the cell survival in the nervous system must be maintained throughout our lives . Neuronal survival is maintained by neurotrophic factors in adults; however , it is unclear whether glia survival is also maintained throughout adulthood . Here , we use the Drosophila visual system as a model to address the role played by adult neurons for the active maintenance of glia . We demonstrated that the adult photoreceptors secrete a signaling molecule , which is transported to the brain to act on the lamina glia and maintain its integrity . When this signaling pathway is blocked , the lamina glia undergoes a progressive and irreversible degeneration . The primary defect occurs in the trafficking from the late endosome and autophagosome to the lysosome . This defect leads to an accumulation of autophagosomes and subsequent cell degeneration as a result of autophagy . Our findings provide in vivo evidence for a novel aspect of the neuron-glia interaction and a novel role for EGFR signaling in regulating the maintenance and degeneration of the nervous system . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [] | 2015 | Maintenance of Glia in the Optic Lamina Is Mediated by EGFR Signaling by Photoreceptors in Adult Drosophila |
Measles remains a significant childhood disease , and is associated with a transient immune suppression . Paradoxically , measles virus ( MV ) infection also induces robust MV-specific immune responses . Current hypotheses for the mechanism underlying measles immune suppression focus on functional impairment of lymphocytes or antigen-presenting cells , caused by infection with or exposure to MV . We have generated stable recombinant MVs that express enhanced green fluorescent protein , and remain virulent in non-human primates . By performing a comprehensive study of virological , immunological , hematological and histopathological observations made in animals euthanized at different time points after MV infection , we developed a model explaining measles immune suppression which fits with the “measles paradox” . Here we show that MV preferentially infects CD45RA− memory T-lymphocytes and follicular B-lymphocytes , resulting in high infection levels in these populations . After the peak of viremia MV-infected lymphocytes were cleared within days , followed by immune activation and lymph node enlargement . During this period tuberculin-specific T-lymphocyte responses disappeared , whilst strong MV-specific T-lymphocyte responses emerged . Histopathological analysis of lymphoid tissues showed lymphocyte depletion in the B- and T-cell areas in the absence of apoptotic cells , paralleled by infiltration of T-lymphocytes into B-cell follicles and reappearance of proliferating cells . Our findings indicate an immune-mediated clearance of MV-infected CD45RA− memory T-lymphocytes and follicular B-lymphocytes , which causes temporary immunological amnesia . The rapid oligoclonal expansion of MV-specific lymphocytes and bystander cells masks this depletion , explaining the short duration of measles lymphopenia yet long duration of immune suppression .
Measles is associated with a transient but profound immune suppression , which may last for several weeks to months after the acute stage of the disease . The clinical importance of this immune suppression is illustrated by the observation that measles mortality is typically caused by secondary infections in the respiratory or digestive tract [1]–[3] . However , the mechanism by which measles virus ( MV ) infection causes immune suppression is not completely understood . Multiple in vivo correlates of immune suppression have been described , including disappearance of Mantoux responses [4] , [5] , lymphopenia [6] , [7] and impaired responses to vaccination [8] , [9] . Decreased lymphoproliferative responses [10] , [11] , altered cytokine response profiles [12] and impairment of antigen-presenting cell function [13]–[15] have been described in vitro . The relevance of these observations to immune suppression and enhanced susceptibility to opportunistic infections remains unclear . The paradox of measles is that the acute phase of the disease is not only associated with immune suppression , but also with immune activation [16] and induction of robust MV-specific humoral and cellular immune responses that result in lifelong immunity . MV infection is initiated in the respiratory tract . It has long been thought that the initial target cells of the virus were epithelial cells of the upper respiratory tract , but recent studies have demonstrated a major role for alveolar macrophages and dendritic cells ( DC ) in this process [17] , [18] . These first MV-infected cells transmit the virus to the bronchus-associated lymphoid tissue ( BALT ) and/or the draining lymph nodes , where the infection is further amplified in lymphocytes and viremia is initiated [18] , [19] . MV infects both T- and B-lymphocytes by binding of the MV-hemagglutinin ( H ) glycoprotein to the cellular receptor CD150 [20] . Recently , the adherens junction protein PVRL4 was identified as cellular receptor on epithelial cells [21] , [22] . However , as this receptor is exclusively expressed on the basolateral surface of epithelial cells , it does not facilitate MV infection of epithelial cells during the early stages of the disease , but instead is thought to play a role in transmission during the late stages of measles pathogenesis [23] . It has been proposed that direct infection of lymphocytes and the subsequent lymphopenia could explain measles-associated immune suppression [24] , [25] . However , this hypothesis has often been dismissed based on the observation that lymphopenia only lasts about a week , whilst immune suppression persists for several weeks to months . In addition , during the peak of viremia no more than 1 to 5% of the total lymphocyte population in peripheral blood is infected [26] , [27] . Recent observations that MV infects high percentages of cells in lymphoid tissues [27] and preferentially targets CD45RA− or CD45R0+ memory T-lymphocytes [27] , [28] led us to revisit the lymphocyte depletion hypothesis for immune suppression using the macaque model . Analysis of virological , immunological and histopathological parameters has demonstrated a remarkable similarity between measles in macaques and humans [29] . Here we present a comprehensive overview of a number of in vivo studies performed in macaques which provides a unifying model for the etiology of measles immune suppression that is both compatible with the measles paradox and with historical in vitro and in vivo correlates of measles immune suppression .
We have analyzed data from rhesus or cynomolgus macaques ( n = 40 ) infected with recombinant ( r ) MV strains ( rMVIC323 or rMVKS ) expressing enhanced green fluorescent protein ( EGFP ) , spanning the early , intermediate and late stages of MV infection ( Table S1 ) . Data from previous studies [18] , [27] , [30] and from additional experimentally infected macaques ( n = 14 ) were combined . The rMVs expressed EGFP from an additional transcription unit , and MV replication results in the host cell becoming EGFP+ . At the peak of viremia , EGFP fluorescence was macroscopically detected in all lymphoid tissues ( Figure 1A–D ) . The percentages MV-infected cells in lymphocyte subsets in PBMC or lymphoid tissues collected 9 or 11 days post-infection ( d . p . i . ) were determined by flow cytometry . Lymphocytes were subtyped as CD4+ or CD8+ naive ( CD45RA+ , Tn ) , central memory ( CD45RA−CCR7+ , TCM ) or effector memory ( CD45RA−CCR7− , TEM ) T-lymphocytes or as naive ( IgD+CD27− , Bn ) or memory ( IgD−CD27+ , BM ) CD20+HLA-DR+ B-lymphocytes ( Figure S1 ) . TCM and TEM were infected at a significantly higher level than Tn . In contrast , BM were not preferentially infected ( Figure 1E–H ) . To determine whether the increased susceptibility of CD45RA− memory T-lymphocytes ( TM ) to MV infection is an inherent property of these cells , and comparable between humans and macaques , we sorted naive and memory CD4+ and CD8+ T-lymphocyte populations from human or macaque PBMC on basis of CD45A expression . In vitro co-culture of these populations with MV-infected autologous cells showed that both human and macaque CD45RA− TM were preferentially infected by MV ( Figure 2A ) . In an alternative approach , unsorted human or macaque PBMC were co-cultured with autologous MV-infected cells , and infection percentages in Tn , TCM and TEM were determined by flow cytometry , resulting in similar differences in the susceptibility of T-lymphocyte subpopulations ( Figure 2B ) . This not only corroborated the in vivo results from the experimentally infected macaques , but also demonstrated a virtually identical trend for human and macaque subpopulations . The MV receptor CD150 is expressed at high levels by activated human CD45ROhigh memory T-lymphocytes [31] . We determined the expression levels of CD150 on the different human and macaque T-lymphocyte subsets . This confirmed the higher level of CD150 expression by human CD45RA− TCM and TEM when compared to CD45RA+ Tn , and showed that the expression levels of CD150 on human and macaque T-lymphocyte subsets are comparable , which likely explains the increased susceptibility of CD45RA− TM to MV infection ( Figure 2C–F ) . Although we were unsuccessful in further subtyping B-lymphocyte subsets on basis of expression patterns of different surface markers , the observed efficient infection of both Bn and BM lymphocytes seems in accordance with recent studies demonstrating CD150 expression on virtually all human B-lymphocyte subpopulations [32] . To address the impact of MV infection in situ , immunohistochemical analyses of serial sections of lymphoid tissues collected at different d . p . i . were performed . This demonstrated that MV mainly replicated in B-cell follicles ( Figure 3; 7 and 9 d . p . i . ) , as previously described [27] , [33] . Multiple syncytia were observed 9 d . p . i . , and dual immunofluorescence showed these were of B-lymphocyte origin ( Figure S2 ) . Strikingly similar to classic observations in humans [34] , lymphoid exhaustion of the centers of the B-cell follicles was observed during and shortly after the peak of viremia ( Figure 3; 9 and 11 d . p . i . ) . In vitro and in vivo recall T-lymphocyte responses to tuberculin were measured in Bacille Calmette-Guérin ( BCG ) -vaccinated macaques , prior to and 11 or 13 d . p . i . IFN-γ production of PBMC in response to purified-protein derivative ( PPD ) stimulation was reduced after MV infection . Notably , an MV-specific IFN-γ response was detected in PBMC collected from macaques sacrificed 13 d . p . i . ( Figure 4A ) . The in vivo recall response was determined by Mantoux testing . Before MV infection a characteristic delayed-type hypersensitivity response developed on the site of intra-dermal tuberculin injection , characterized by a well-delineated soft swelling of the cutus , corresponding with an influx of CD3+ T-lymphocytes ( Figure 4B ) . In line with classical observations [4] , [5] , Mantoux responses were suppressed after MV infection , with a much smaller and harder swelling in the cutus , potentially corresponding with epidermal repair , skin-infiltrating CD3+ T-lymphocytes were absent ( Figure 4B ) . Interestingly , we also observed macroscopically detectable EGFP expression in the skin at the sites where the animals had been vaccinated intra-cutaneously with BCG three months prior to MV infection ( Figure S3 ) , suggesting infection of tissue-resident memory lymphocytes in the skin . Analysis of macaque white blood cell ( WBC ) counts during the acute infection demonstrated a profound but transient leukopenia ( Figure 5B , circles ) , which coincided with the peak of viremia . There was a relative decrease in size of the CD45RA− CD4+ and CD8+ TCM and TEM populations between 0 and 9 d . p . i . ( Figure 5A ) , suggesting that leukopenia was related to depletion of MV-infected cells . WBC counts rapidly returned to pre-infection levels between 9 and 15 d . p . i . ( Figure 5B ) , paralleled by a restoration of the relative CD45RA− TCM and TEM population sizes ( Figure 5A ) . Lymph nodes were enlarged in all animals euthanized between 11 and 15 d . p . i . , which was paralleled by the clearance of EGFP+ lymphocytes from lymphoid tissues . During this period large numbers of Ki67+ proliferating cells repopulated the B-cell follicles in the germinal centers of lymphoid tissues ( Figure 3; 13 and 15 d . p . i . ) . Numbers of apoptotic lymphocytes , as detected by staining for cleaved caspase 3 ( CC3 ) , remained low at all time-points ( Figure S4 ) . Infiltration of CD3+ T-lymphocytes into the B-cell follicles suggests that MV-infected cells were cleared by cytotoxic T-lymphocyte-mediated killing , rather than undergoing apoptosis [35] . Based on the observations described above we propose a model for the events leading to measles immune suppression . Infection and subsequent immune-mediated clearance of CD150+ lymphocytes results in specific depletion of memory T-lymphocytes and follicular B-lymphocytes , whilst the naive T-lymphocyte population remains relatively unaffected ( Figure 5B , populations shown in red and blue , respectively ) . However , this leaves the question how such a short duration of measles lymphopenia can be reconciled with the long-lasting immune suppression . We hypothesize that the lymphocyte depletion is masked by the massive expansion of MV-specific and bystander lymphocytes ( Figure 5B , population shown in green ) , which has also been observed in humans by a massive expansion of CD8+ T-lymphocytes [36] and by demonstrating skewing of the T-cell receptor repertoire after measles [37] . As a consequence , the qualitative composition of the lymphocyte population immediately after recovery from lymphopenia is dramatically different from that before MV infection , as pre-existing memory lymphocytes have been depleted and replaced by MV-specific and bystander lymphocytes . The net result is a temporary immunological amnesia , and restoration of immunological memory may take several weeks . Thus this model explains why measles-associated immune suppression extends well beyond the transient lymphopenia .
Based on our observations , we conclude that measles immune suppression can , at least in part , be explained by massive infection and subsequent immune-mediated clearance of CD150+ memory T-lymphocytes and follicular B-lymphocytes . Depletion of T- and B-lymphocytes has also been described in the BALT of measles patients [38] . We show that MV preferentially infects CD45RA− TCM and TEM , which during secondary immune responses are the primary source of T-lymphocyte expansion or generation of effector T-lymphocytes , respectively [39] . Infection and subsequent immune-mediated depletion of memory T-lymphocyte subsets fits with the first description of measles-induced immune suppression , namely the disappearance of Mantoux responses in measles patients [4] . Several hypotheses for the underlying mechanism of measles immune suppression have been described previously . However , none of these adequately explain the “measles paradox”: the disease is associated with immune suppression but also with induction of strong MV-specific immune responses . Our observations not only explain the measles paradox , but also shed light on many of the in vivo and in vitro correlates of measles immune suppression described in literature [40]–[42] . Unresponsiveness of PBMC to mitogens and altered cytokine profiles during and after the acute phase of measles have been demonstrated in vitro in several studies [10] , [11] . These observations are not disputed , but we consider it difficult to extrapolate unresponsiveness of PBMC to mitogens in vitro to immune suppression in vivo . In parallel with the clearance of MV-infected cells we observed high numbers of proliferating Ki67+ cells in B-cell follicles . This fits well with classical observations of immune activation following measles [16] , suggesting that lymphoproliferation is not impaired in vivo during the convalescent phase . The observed in vitro suppression of mitogen-induced lymphoproliferation could also be explained by the altered qualitative composition of lymphocyte populations in convalescent measles patients , compared to healthy controls . Our model and previously published observations [16] show that a large proportion of the lymphocytes that circulate during the first weeks after measles infection have been recently activated in vivo , potentially making these cells less susceptible to re-stimulation in vitro . We do not exclude that a functional impairment of lymphocytes or DC contributes to immune suppression and thus may augment the extent of immune suppression . Actually , it is likely that there are many factors that contribute to immune suppression in vivo . To date , direct evidence of DC infection by MV in humans has not been obtained . However , it has been demonstrated in vitro that MV is capable of infecting mature DC and Langerhans cells ( LC ) [13] , [14] , [43]–[47] . In vivo in experimentally MV-infected macaques there was strong evidence for infection of DC in the skin and secondary lymphoid tissues [27] . The role of DC in immune suppression has not been extensively studied in vivo , but it is possible that they play a role either by directly being targeted and depleted by MV or indirectly by interaction with and silencing of T-lymphocytes . Furthermore , the capacity to function as antigen presenting cells might be affected [40] , [48] . Our study covers a time period of two weeks after MV infection , and as such does not provide experimental proof of what happens during resolution of measles immune suppression . The strength of our model using recombinant EGFP-expressing MV mainly lies in the sensitive detection of MV-infected cells , which is limited to the first two weeks after MV infection . Clearly , the memory lymphocyte populations specific for previously encountered pathogens are not completely depleted , and are largely restored during the weeks to months after measles . For instance , previously positive Mantoux responses disappear after onset of rash [4] , [5] , but eventually reappear . We have indicated this in our model ( Figure 5B ) by showing a gradual increase of the memory lymphocyte population after clearance of MV . Although we cannot fully explain the drivers of the resolution of immune suppression , it is possible that expansion of non-depleted TCM upon renewed antigen encounter may play an important role . However , homeostatic restoration by the immune system itself could be an alternative explanation . Some of the individual observations described here have been reported earlier in relationship to animal morbillivirus-related immune suppression [49]–[51] . The novelty of our model lies in the immune-mediated lymphodepletion being masked by the massive expansion of MV-specific and bystander lymphocytes . Although effective MV-specific CD8+ T-lymphocyte responses as well as immune activation and lymph node enlargement have been described earlier [16] , [36] , [37] , [51] , they have not been associated with immune suppression in this way . The most important consequence of our model is that the qualitative composition of lymphocyte populations changes dramatically upon MV infection . Although lymphocyte numbers in peripheral blood and lymphoid organs appear normal , depletion of pre-existing specific T- and B-lymphocytes subpopulations provides a direct explanation for the suppression of recall responses to other pathogens during and after measles . This allows such pathogens to cause severe disease and in the developing world leads to the high level of MV-associated mortality . In addition , MV efficiently replicates in B-lymphocytes , resulting in follicular exhaustion and disorganization of the germinal centers , which are essential in actively ongoing humoral immune responses . We observed comparable levels of MV-infected cells in TCM , TEM and Bn , and in parallel observed lymphocyte depletion and disorganization in B-cell follicles during the acute phase of MV infection ( Figure 3 , 9–11 d . p . i . ) . However , we have also shown that proliferating cells can be detected in lymphoid tissues as early as 11 d . p . i . , after which time the follicle structure is being restored . This matches the kinetics of antibody responses in the macaque model , in that MV-specific IgM and IgG responses are first detected around 11 d . p . i . and peak at 17 ( IgM ) and 24 ( IgG ) d . p . i . [27] , [52] . These kinetics fit well with our conclusions from the data and map well onto the immune suppression model . It has been described that MV infection can result in transient remissions of certain autoimmune diseases [53]–[55] . Our observations suggest that this can be explained by direct MV infection of CD150+ lymphocytes , followed by immune-mediated depletion . Similarly , this mechanism could also explain reductions in HIV-1 loads during acute measles [56] , [57]: MV infection of memory CD4+ T-lymphocytes could result in depletion of HIV-1-infected cells . In certain auto-immune diseases , and in animal studies in which lymphocyte populations were experimentally depleted , commensal or opportunistic infectious agents that would normally be controlled by the immune system have been shown to cause severe disease [58] . The high incidence of respiratory and gastro-intestinal complications following measles [1]–[3] may therefore be directly related to the observed high percentages of MV-infection and subsequent lymphocyte depletion in the adenoids , tonsils and gut-associated lymphoid tissue , which form a first line of defense against inhaled or ingested pathogens . We conclude that MV infection wipes immunological memory , resulting in increased susceptibility to commensal or opportunistic infections .
Animal experiments were conducted in compliance with European guidelines ( EU directive on animal testing 86/609/EEC ) and Dutch legislation ( Experiments on Animals Act , 1997 ) . The protocols were approved by the independent animal experimentation ethical review committee DCC in Driebergen , The Netherlands . Animal welfare was observed on daily basis , animal handling was performed under light anesthesia using ketamine and medetomidine . After handling atipamezole was administered to antagonize the effect of medetomidine . For experiments involving PBMC from human donors , written informed consent for research use was obtained by the Sanquin blood bank . PBMC and tissues were collected from cynomolgus ( Macaca fascicularis ) ( n = 35 ) or rhesus ( Macaca mulatta ) ( n = 5 ) macaques included in previously published studies [18] , [27] , [30] ( n = 26 ) or from unpublished infection experiments with rMVIC323EGFP or rMVKSEGFP ( Table S1 and Dataset S1 ) ( n = 14 ) . Macaques were infected by intra-tracheal inoculation or aerosol inhalation and euthanized at 2 ( n = 3 ) , 3 ( n = 3 ) , 4 ( n = 3 ) , 5 ( n = 4 ) , 7 ( n = 9 ) , 9 ( n = 8 ) , 11 ( n = 6 ) , 13 ( n = 2 ) or 15 ( n = 2 ) d . p . i . Although some of the experiments had been designed to address different research questions , the accumulated samples effectively covered all stages of MV infection in macaques . Four macaques received intra-dermal vaccinations with 4×0 . 1 ml of live BCG ( NVI , Bilthoven , Netherlands ) . The animals received an intra-dermal Mantoux test with old tuberculin ( 0 . 1 ml , 25 , 000 IU/ml , Statens Serum Institut ) [59] 3 months post-vaccination at 7 days pre-MV-infection , or 3 days prior to necropsy . Skin reactivity was assessed for three consecutive days . Skin samples from both Mantoux tests ( pre- and post-MV infection ) were collected into formalin . PBMC obtained 7 days pre-infection and during necropsy were thawed and plated into 96-wells round-bottom plates at 2×105 cells per well . Cells were stimulated with either PPD ( 10 µg/ml ) , UV-inactivated MV ( 10 µg/ml ) or live rMVrEdtEGFP [30] ( 5×104 CCID50 in the presence of 10 µg/ml infection-enhancing lipopeptide PHCSK4 [60] ) for 48 hours in triplicate . IFN-γ concentrations were measured in supernatants by ELISA ( U-CyTech Biosciences ) . After each blood collection , total WBC counts were obtained using an automated counter ( Sysmex pocH-100iV ) . To address measles-induced leukopenia in these animals , mean WBC counts were determined on 0 ( n = 31 ) , 1 ( n = 6 ) , 2 ( n = 20 ) , 3 ( n = 17 ) , 4 ( n = 18 ) , 5 ( n = 11 ) , 6 ( n = 20 ) , 7 ( n = 10 ) , 8 ( n = 6 ) , 9 ( n = 12 ) , 11 ( n = 8 ) and 13 ( n = 2 ) d . p . i . Different animals were included at each time-point . Animals were euthanized by exsanguination under ketamine anesthesia . For the purpose of detecting EGFP fluorescence , a lamp was custom-made containing six 5-volt LEDs ( Luxeon Lumileds , lambertian , cyan , peak emission 490–495 nm ) mounted with D480/40 bandpass filters ( Chroma ) in a frame that allowed decontamination with 70% ( v/v ) alcohol or fumigation with formaldehyde . Emitted fluorescence was visualized through the amber cover of a UV transilluminator normally used for screening DNA gels . Photographs were made using a Nikon D80 digital SLR camera . Lymphoid tissues were collected in buffered formalin for immunohistochemistry or PBS for preparation of single cell suspensions , which were used directly for flow cytometry . T-lymphocytes were subdivided into Tn , TCM and TEM populations ( Figure S1 ) by staining with CD3PerCP ( BD Biosciences , clone SP34-2 ) , CD4V450 ( BD Biosciences , clone L200 ) , CD8AmCyan ( BD Biosciences , clone SK1 ) , CD45RAPE-Cy7 ( BD Biosciences , clone L48 ) and CCR7APC ( R&D Systems , clone 150503 ) . The APC signal was enhanced using an APC-FASER Kit ( Miltenyi Biotec ) . B-lymphocytes were subdivided into Bn and BM populations ( Figure S1 ) by staining with CD20PE-Cy7 ( BD Biosciences , clone L27 ) , HLA-DRPacific Blue ( Biolegend , clone L243 ) , CD27APC ( eBioscience , clone O323 ) and a combination of IgDBiotin ( Southern Biotech , goat polyclonal ) and streptavidinPerCP ( BD Biosciences ) . CD150 expression was determined by staining with CD150FITC ( AbD Serotec , clone A12 ) . The infection percentages within the populations were determined by detection of EGFP . All flow cytometry was performed on a FACS Canto II ( BD Biosciences ) . H&E staining was performed to evaluate histological changes . Immunohistochemical staining was performed using a fully automated BondMax immunostainer with a polymer-based peroxidase detection system . MV-infected cells were detected using a polyclonal rabbit antibody to EGFP ( Invitrogen ) . Similar stainings were performed with the following monoclonal antibodies: T-lymphocyte marker CD3 ( DAKO , clone F7 . 2 . 38 ) , B-lymphocyte marker CD20 ( DAKO , clone L26 ) , proliferation marker Ki67 ( DAKO , clone MIB1 ) and apoptosis marker cleaved caspase 3 ( Cell Signaling , clone 5A1E ) . Glass slides were scanned with a 40X/0 . 75 Olympus UPlan FLN objective on an Aperio Scanscope CS-O SS5200 equipped with Spectrum Plus . An Aperio Positive Pixel Count Algorithm was applies to quantify and therefore standardize the intensity of stains present to produce optimal discrimination between immunoperoxidase diaminobenzidine ( DAB ) tetrahydrochloride reactions and hematoxylin stained nuclei . Scanners are kept at ambient temperature in a temperature-controlled area to eliminate loss of performance due to overheating . PBMC from healthy human or macaque donors were sorted into pure CD4+ or CD8+ naive ( CD45RA+ ) and memory ( CD45RA− ) T-lymphocyte populations on basis of CD45RA expression . Unsorted or sorted PBMC from humans or macaques were infected by co-culture with low numbers ( 1∶1000 ) of autologous MV-infected B-LCL or BAL cells ( infected with cell-free rMVKSEGFP at an MOI of 1 for 48 hours ) , respectively . After two days of co-culture infection percentages in the different subsets were determined by flow cytometry . Differences between percentages infected cells or CD150 expression were tested by the non-parametric Wilcoxon rank test using SPSS software . | Measles is associated with a transient immune suppression , resulting in increased susceptibility to opportunistic infections . Indeed , the main causes of measles mortality are secondary infections in the respiratory and digestive tract . Although measles is associated with lymphopenia , depletion of lymphocytes has often been dismissed as a cause of immune suppression . Lymphocyte counts rapidly return to normal after clearance of the virus , while immune suppression lasts several weeks to months . Many studies have focused on suppression of lymphocyte proliferation as an in vitro correlate of immune suppression . However , experimental infections of non-human primates show that in vivo lymphocyte proliferation is not impaired after measles . Instead , we hypothesize that massive expansion of MV-specific and bystander lymphocytes masks the fact that pre-existing memory lymphocytes have been depleted . We conclude that measles virus infection wipes out immunological memory , leaving individuals susceptible to opportunistic infectious agents that would normally be controlled by the immune system . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"medicine",
"immune",
"cells",
"immunity",
"to",
"infections",
"immunology",
"immune",
"suppression",
"microbiology",
"animal",
"models",
"histology",
"adaptive",
"immunity",
"model",
"organisms",
"immune",
"defense",
"animal",
"models",
"of",
"infection",
"infectious",
"diseases",
"biology",
"infectious",
"disease",
"modeling",
"macaque",
"immunity",
"virology",
"measles",
"co-infections",
"viral",
"diseases"
] | 2012 | Measles Immune Suppression: Lessons from the Macaque Model |
Lengthy co-evolution of Homo sapiens and Mycobacterium tuberculosis , the main causative agent of tuberculosis , resulted in a dramatically successful pathogen species that presents considerable challenge for modern medicine . The continuous and ever increasing appearance of multi-drug resistant mycobacteria necessitates the identification of novel drug targets and drugs with new mechanisms of action . However , further insights are needed to establish automated protocols for target selection based on the available complete genome sequences . In the present study , we perform complete proteome level comparisons between M . tuberculosis , mycobacteria , other prokaryotes and available eukaryotes based on protein domains , local sequence similarities and protein disorder . We show that the enrichment of certain domains in the genome can indicate an important function specific to M . tuberculosis . We identified two families , termed pkn and PE/PPE that stand out in this respect . The common property of these two protein families is a complex domain organization that combines species-specific regions , commonly occurring domains and disordered segments . Besides highlighting promising novel drug target candidates in M . tuberculosis , the presented analysis can also be viewed as a general protocol to identify proteins involved in species-specific functions in a given organism . We conclude that target selection protocols should be extended to include proteins with complex domain architectures instead of focusing on sequentially unique and essential proteins only .
Tuberculosis ( TB ) remains a major world-wide health hazard , causing to roughly 2 million deaths per year . Approximately , one third of the world's population is currently infected with Mycobacterium tuberculosis ( MTB ) , the causative agent of TB [1] , [2] . MTB is an intracellular parasite , an organism notoriously hard to fight . One of the major reasons for its persistence is the intricate network of host-pathogen interactions which is exploited by the bacterium and which creates a fine-tuned niche for its survival in macrophages [3] . This has been developed during lengthy periods of “co-habitation” and , consequently , co-evolution . The MTB genome has been molded to accommodate the circumstances of life within macrophages . In fact , the bacterium has been so successful in this process that it is notably hard to cultivate outside its physiological host . During the co-evolution process with humans ( cf . archeological data presenting experimental evidence for the co-habitation of MTB and humans back to 9000 years [4] ) , the genome changes within the bacterium have been facilitated by its error-prone DNA polymerases [5] . As a result , the present MTB organism is very close to being an obligatory intracellular parasite . Mycobacteria are intrinsically resistant to most commonly used antibiotics and chemotherapeutic agents . Due to its specific structure and composition , the mycobacterial cell wall is an effective permeability barrier , generally considered to be a major factor in promoting the natural resistance of mycobacteria . Only a few drugs are active against mycobacterial pathogens , and current treatment strategies for TB consists of 3 or 4 drugs used in combination . However , the increasing emergence of multi-drug resistant tuberculosis ( MDR-TB ) and extensively drug-resistant tuberculosis ( XDR-TB ) necessitates the development of novel drugs [6] . Furthermore , novel drugs compatible with antiretroviral therapy are needed to treat co-infected AIDS patients [7] and new drugs are also required that can specifically be employed for children . Clearly , there is an urgent need for drug development projects that actually possess novel targets and novel mechanisms of action [8] . A significant step towards understanding the biology of MTB was provided by full genome sequencing of various strains of this microorganism , including the best characterized laboratory strain , H37Rv , that contains 3 , 984 genes [9] . The complete genome sequences of several other mycobacteria have also become available , showing various levels of divergence [10] , [11] . While the genome size of M . bovis is largely similar to that of MTB , the genome of M . leprae is reduced to only 40% of that of MTB [12] . These genomes can also be compared to those of many other pathogenic and non-pathogenic bacteria , as the number of fully sequenced bacterial genomes is over 600 and is rapidly increasing . The genomes of several eukaryotic organisms have also been sequenced and are now largely annotated , including the human genome . Additionally , the Human Microbiome Project ( HMP ) has published the sequenced genomes of 178 microbes that exist within or on the surface of the human body [13] , [14] . The plethora of genomic sequences offers a novel platform for comparative analyses and large-scale studies . This new source of data can help to identify proteins in the MTB proteome that perform essential functions ensuring the survival and virulence of the bacterium . These proteins present potential targets for drug design . Target selection is the crucial starting point of any drug development process . Traditionally , this procedure relied on established knowledge of individual proteins and their functions . The availability of complete genome sequences opened a new era and lead to the development of various bioinformatics methods which can prioritize targets in an automated cost-effective way . These approaches can take various criteria into account with the aim to minimize the interactions with the host environment yet specifically attack the pathogen's growth and survival . Several such studies focused on metabolic enzymes . In their work , Anishetty and co-workers collected enzymes from the biochemical pathways of MTB using the KEGG metabolic pathway database [15] . As a result , 186 proteins were suggested as potential drug targets based on the lack of similarity to proteins from the host H . sapiens . Hasan and co-workers proposed a ranking system by targeting metabolic checkpoints based on the uniqueness of their role in the pathogen's metabolome [16] . Additionally , targets were penalized for having high sequence similarity to proteins of the host and of the host flora . The targetTB database was created based on similar principles [17] . Using flux balance analysis and network analysis , proteins critical for the survival of MTB were first identified , and then subjected to comparative genomics analysis with the host . Finally , a novel structural analysis of potential binding sites was carried out to assess the validity of a protein as a target . The selection also incorporated data about the essentiality of proteins using the results of experiments carried out under nutrition rich conditions . A recent analysis constructed a proteome-wide drug target network by linking the structural proteome of MTB with structurally characterized approved drugs [18] . In most drug target selection protocols , the existence of a protein structure or a structural homologue is treated as an advantage for rational drug design . Breaking with this tradition , Anurag and Dash suggested a list of intrinsically disordered proteins in the MTB genome as potential drug targets [19] . This is in accordance with the recent finding that these proteins can also serve as promising drug targets [20] , exemplified by the successful blocking of the p53-MDM2 interaction by a small molecule [21] . Fueled by this observation , a list of proteins with disordered protein segments were compiled and filtered for essentiality , uniqueness and involvement in protein-protein interactions . This resulted in 13 proposed drug targets . These proteins have a probable role in signaling , regulation and translation , instead of metabolisms [19] . The success of the target selection procedure critically depends on identifying distinctive features of the pathogen that are essential for its survival . The protein repertoire encoded by the genome provides the initial starting set from which potential targets can be selected based on various hypotheses . However , the optimal target selection criteria are still a matter of considerable debate [22] , [23] . The prime criteria of current target selection protocols are essentiality , lack of sequence homologues at least in the host , and the presence of functional characterization . These criteria , however , can lead to the overlook of several important candidates . In the case of MTB , there are several proteins that do not meet the aforementioned criteria but should not be disregarded as potential targets due to their eminent biological importance . For example , the genome sequence of MTB revealed that about 10% of the coding of the genome is devoted to two largely unrelated families of acidic , glycine-rich proteins , the PE and PPE families [9] . These proteins are largely sequence specific to mycobacteria and have been implicated in host-pathogen interactions and antigenic variations [24] . However , most of these proteins are not essential and their function is largely uncharacterized . An additional new class of promising targets in MTB corresponds to signaling elements , in particular to the pkn family of Ser/Thr protein kinases [25] , [26] . These MTB proteins play essential roles in both bacterial physiology and virulence [26] , but are evolutionary related to eukaryotic protein kinases . These protein families are cases that challenge current target selection protocols and indicate that different approaches for target selection are needed . In this work , we propose a novel computational strategy based on phylogenetic profiling and comparative proteomic analysis that can highlight proteins involved in specifies-specific functions . This approach takes into account the complex evolutionary scenarios that can lead to the emergence of novel species-specific functions . Novel function can arise from de-novo protein creation but also from more ancient proteins by the combination of divergence , duplication and recombination events [27] . In order to gain insights into the contribution of the various processes , we carried out a comparative proteomic analysis . By focusing on the causative agent of tuberculosis , we analyzed the protein domain and disorder content of its proteome and carried out large-scale local sequence similarity searches to identify basic evolutionary patterns in MTB . We show that the enrichment of certain protein families in the genome can automatically indicate an important function specific to this pathogen . The implications of these findings for target selection are also discussed .
In this work we carried out a comparative genomic study based on the content of domains and disordered regions and the result of large-scale sequence similarity searches . This approach is completely general and could be applied to any kind of organism with an annotated genome . Here , we focused on MTB , the causative agent of tuberculosis . Our analyses revealed two protein families in the proteome of MTB that stand out in several aspects . These proteins were also shown to have a functional importance essential for the survival of this pathogen . Next , we will examine them as potential targets for drug design . The common properties of both the pkn and PE/PPE families include unusual domain accretions specific to this organism . This is combined with an increase in their disorder content . Both families carry out important functions in the MTB and are involved in the interactions with the host cell . Various members were shown to be essential for the organism and according to a recent analysis using guinea pig model , representatives of these families are significantly enriched in the early and chronic stages of infections [76] . Furthermore , many of them are either located in the surface of the bacteria or are exported into the host cell . The properties of these protein families underscore their biological importance and suggest that they would be ideal candidates for drug design . However , conventional drug design procedures generally overlooked such proteins as targets by largely focusing on metabolic processes . The need for novel drugs for the treatment of MTB forces researchers to explore new directions for target selection . The pkn and PE/PPE families , through their complex architectures offer several options in this regard . One of the most active areas of therapeutic research involves developing protein kinase inhibitors . With recent advances , protein kinases have now become the second most important group of drug targets after G-protein-coupled receptors [77] . Currently , three protein-kinase inhibitors are already in clinical use , and several other protein-kinase inhibitors are undergoing human clinical trials , mostly targeting cancer [77] . However , kinases may also be exploited in the development of novel antibiotics against mycobacteria . Mycobacterium tuberculosis contains 11 eukaryotic-like Ser/Thr protein kinases ( STPKs ) . All of these proteins accommodate the common catalytic domain and the 12 conserved motifs that define the signature of eukaryotic protein kinases . Despite these similarities , the sequence identity to the human SPTK genes is around 30% , low enough to allow the promising search for species-specific inhibitors [26] . Recently , a high-resolution 3D structure of the pknG kinase in complex with a potent antimycobacterial compound was published [55] . This structure revealed that despite similarities in the overall fold of the conserved kinase domain , the species-specific structural characteristics of the protein still allows highly specific interaction patterns with the promising drug-like compound . The structure of the catalytic domain of pknB in complex with different ATP analogues was also published [78] , [79] and revealed that in contrast to the similar kinase structures from eukaryotes , an important regulatory loop shows high degree of flexibility in the mycobacterial kinase preventing its localization in the crystal structure . Dimerization of signaling kinases is a frequently observed phenomenon , and within the architecture of the dimer interfaces , species-specific and conserved elements are both found [80] , [81] . Specifically , the dimerization of the mycobacterial STPK proteins seems to be universal and functionally relevant [82] , therefore the identified intermolecular interfaces may present a further target surface to perturb protein function . The additional sequence regions tethered to the common kinase domain are specific for each of the pkn proteins . As most of these domains are located outside of MTB cells , they could be more accessible for potential drugs . Although there is detailed structural information available for pknB , pknD and pknG proteins , they offer very few clues about potential “druggable” sites . According to their disorder profiles ( Figure 5 ) , pknE , pknK and to some extent pknJ are likely to contain additional ordered domains . The future structural characterization of these regions could offer further drug target sites . Concluding on the question whether mycobacterial kinases may present useful targets against MTB , we wish to emphasize that due the fact that many critical cellular processes in human are regulated by kinases , cross-reactivity could have dramatic consequences [83] , [84] . In order to alleviate such effects , we propose to focus on a combined approach which allows targeting different elements of a mycobacterial kinase ( eg dimerization domain ) . In such an approach , targeting the MTB-specific active site components of a kinase ( eg . pknG ) , may be coupled to designing potentially inhibitory small molecular compounds that bind to MTB-specific sequence elements outside the active site . The PE/PPE proteins are similar to pkn proteins in the sense that they also contain a well-conserved N-terminal domain that defines these families and C-terminal parts that show large variations in terms of their length , domain composition and disorder content . Despite the obvious biological importance of these proteins , very little structural and functional information is available for them . In a large-scale structural genomics approach it was found that individual members of both PE and PPE families did not express well or expressed in insoluble or unfolded forms [73] . To explain the failure of structure determination efforts it was suggested that these proteins need partners to fold . Indeed , via a genomic analysis one specific pair of PE and PPE proteins was predicted to interact , and their structure was successfully determined [73] . The extensive interface formed between the PE and PPE domains , that also contain the conserved N-terminal motifs , can also be considered from the viewpoint of drug design . This is supported by the growing number of examples of small molecules successfully blocking protein-protein interactions [85] . PE/PPE proteins also contain several additional ordered domains ( Figure 6 ) , however , they lack detailed structural information . In most cases , functional and structural information cannot be inferred from sequence homologs , as the family itself is highly specific to Corynebacterineae with the vast majority of PE/PPE members being present in mycobacteria only . There are , however , some exceptions , like the α/β hydrolase 3 domain , that is a catalytic domain found in a very wide range of enzymes [86] , or the phosphoglycerate mutase ( PGAM ) domain that catalyses reactions involving the transfer of phospho groups between the three carbon atoms of phosphoglycerate [87] . Another common property of several members of both families is the presence of long disordered segments . Until recently , the feasibility of targeting proteins without a well-defined structure was unclear for the purpose of drug development . There is now , however , a newly sparked interest in intrinsically disordered proteins as potential drug targets [88] , [89] . This originates partly from recognizing the biological importance of disordered proteins , especially in signaling and regulatory processes , but also from realizing the specific mode of their binding [32] , [90] . Disordered proteins usually interact via a coupled folding and binding process that involves a transition from a largely flexible state to a more ordered state [34] . This transition is associated with a large entropy cost that can make the overall binding quite weak while maintaining specificity . The low binding free energy of these interactions indicates that they would be relatively easy to block by small molecules [20] . In one example , a promising small , drug-like molecule was found to bind into the groove of MDM2 . This molecule inhibits the association of the ordered protein MDM2 with the disordered segment of p53 by mimicking the short alpha helical structure of p53 adopted upon binding . In another example , dimerization of two disordered proteins , c-Myc and Max , was blocked by specific inhibitors [89] . Generally , the analysis of known examples of the druggable regions of disordered proteins indicated that these segments overlapped with the binding regions predicted by ANCHOR [91] . Therefore , ANCHOR [41] and other disordered binding region prediction algorithms that will be hopefully developed in the years to come can be extremely useful to highlight potential druggable sites directly from the amino acid sequence , especially in combination with other methods . According to our results , both pkn and PE/PPE protein families contain putative druggable sites located in disordered segments . For example , the region in the C-terminal part of pknA ( residues 419–431 ) can be of special interest , as here a high-confidence disordered binding region is predicted by ANCHOR that coincides with a high confidence predicted α helical region predicted by PSIPRED [92] ( Figure S1 ) . This is indicative of a special class of disordered binding regions that form an α helix upon binding to their partner , similarly to the binding of p53 to MDM2 . The large number of disordered binding regions predicted in the PE/PPE families indicates that there can be many other druggable sites awaiting further characterization . The two families highlighted in this study offer various options for drug design . However , most members would be omitted from traditional target selection procedures due the lack of essentiality . In the pkn families , only pknA , pknB and pknG were shown to be essential , while essentiality was showed for only 9 members of the PE/PPE family . Since MTB is an obligate parasitic pathogen , it is extremely difficult to identify genes that are required for the optimal growth of mycobacteria under in-vivo conditions . Depiction of the MTB proteome during infection is usually based on a hypothesized environment that simulates the conditions within the infected lung and defined on the basis of bacterial response to pH , starvation and hypoxia [93]–[96] . Recently , the guinea pig model was used to examine the bacterial proteome in vivo during the early and chronic stages of the disease [76] . According to this study , various members of the PE/PPE families and pknA were observed among the most dominant proteins in the infected lung samples , giving further support for the importance of these protein families . Interestingly , there were major differences between the results of these in vitro and in vivo studies , suggesting that none of the simulated in vitro model environments accurately reflects the protein profile within the lung [76] . A further limitation of studying the essentiality of individual proteins arises from the functional overlap among members of various protein families . Each member of the pkn group contains the kinase domain , and they share many of their substrates despite the differences in their sensor domains [97] . Although PE/PPE proteins are much less well-characterized , several members of them can also exhibit significant similarities with each other and this phenomenon is also reflected in their domain composition . The similarity often goes beyond the common N-terminal domain , as many members of the PE/PPE protein family share the same domain architecture ( see Figure 6 ) . There is a high likelihood of functional overlap in these cases . Protein families with overlapping functions challenge the notion of essentiality as target selection criteria . However , by targeting the common domain of protein families , several proteins of the pathogen can be attacked using the same drug molecule . Relying on such multi-target drugs can be a more efficient avenue in drug design [98] . In this respect , relatively conserved domains that occur multiple times can be of potential interest . Such examples include the PE and PPE domains , the kinase domain of pkn proteins , the NHL domain that occurs twice in the PE-PGRS groups as well as in pknD . There are several other currently uncharacterized domains occur multiple times in the genome of MTB ( Figure 6 ) . These proteins can provide interesting targets for polypharmacological drugs [18] . The increasing number of complete genome sequences has enabled comparative genomic analyses which can be used to understand the distinctive properties of various pathogens and to specifically target them based on this knowledge . In this work we analyzed the proteins encoded by the genome of MTB from this viewpoint . We identified two protein families , the pkn and PE/PPE , that showed unusual species-specific enrichment of domains . These proteins can be considered as potential targets for drug design as they are involved in vital functions that are specific to this pathogen . Members of both families have complex domain architectures that combine family-specific domains with other domains and disordered segments . The analysis presented in this study predicts that drug design against members of these two families may lead to promising hits; verification of this prediction awaits further studies . It is important to emphasize that members of these two families represent novel potentials since the compounds that are either currently used in the clinics against tuberculosis or are under clinical trials are directed against other target proteins . Although some of our findings are specific to MTB , there are several more general implications of this study . The exclusivity of certain proteins to a given pathogen is often one of prime criteria used in various target selection protocols . However , our results indicate that species-specific functions are not necessarily brought about by species-specific proteins . In contrast , many novel functions developed from already existing proteins . In the case of eukaryotes , there are several notable examples , such as the development of olfaction , reproduction , and immunity [27] , where the combination of gene duplication , divergence and recombination led to the expansion of protein families and provided jumping points in evolution . The example of MTB shows that such complex evolutionary scenarios play important roles in prokaryotes as well and can be detected by species-specific enrichment of certain protein domains or families . Protein families emerging as a result of such processes often have complex domain architectures . Consequently , these proteins can be approached from multiple directions for the purpose of drug development . Besides traditional strategies that aim to inhibit enzymatic functions , less well-established approaches should also be considered , such as developing compounds against protein-protein interaction sites or disordered binding regions . Due the potential overlap within these protein families , individual members might not comply with essentiality . In these cases , however , a single drug can be effective against more than one protein . These cases of functional overlap arising from direct similarity , unlike indirect effects [99] , may be predicted based on target similarity . Taking the various factors of our findings into account can help to improve the success rate of target selection protocols and drug development process .
For the domain assignment the protein domains contained in the Pfam database were used ( http://pfam . sanger . ac . uk/ ) [30] . Pfam is composed of two parts , Pfam-A and Pfam-B . Pfam-A domains are manually curated and have a distinct name ( e . g . Abhydrolase_3 ) . On the other hand , Pfam-B domains are automatically assigned using sequence similarity searches and usually no additional annotation is supplied concerning them . Pfam-B domains are not given a separate name and can be identified by a unique numerical ID . For the proteome-scale comparative studies , a dataset containing 1 , 904 , 578 protein sequences from 467 known complete proteomes was assembled ( 20 eukaryotic and 447 bacterial proteomes containing 392 , 401 and 1 , 512 , 177 proteins respectively ) . These proteomes were taken from the UniProt ftp server ( ftp://ftp . uniprot . org/ ) [100] . For the similarity searches between MTB proteins and the proteins in the SDCP , PSI-BLAST was used [101] . First , a PSI-BLAST profile was calculated for each of the 3 , 948 proteins in the MTB proteome using the UniRef90 database , with three iterations . Next , these profiles were used to find hits from the proteins in SDCP , the database containing protein sequences from 467 known complete proteomes . A hit was considered significant ( the MTB and the other protein was considered similar ) and was used further on , if the e-value was below 10−4 . Since the similarity between proteins is often restricted to shorter parts of the sequence , an explicit coverage threshold was not included in this work . This enabled the recognition of similar domains or other local protein regions between proteins , and was necessary for the successful clustering of related proteins , such as the PE/PPE families . The e-value cutoff ensured that these similarities are still significant . Based on the alignments , all locally similar sequences from the SDCP were collected for each protein in the MTB proteome . Next , for each MTB protein a similarity profile was built that contains the number of similar sequences for each of the 467 organism in the SDCP . In general , the locally similar segments identified by PSI-BLAST are larger than single domains . The aligned regions can also contain disordered regions that are not excluded by the low complexity filter . Furthermore , since the similarity searches were centered around the MTB proteins , this analysis could also find similarities between sequences in some cases , even if they do not share Pfam domains . For the functional categorization of MTB proteins , data were taken from the TubercuList server ( http://genolist . pasteur . fr/TubercuList/ ) [42] . According to this site , each MTB protein is unambiguously grouped into one of the following categories: virulence , detoxification , adaptation; lipid metabolism; information pathways; cell wall and cell processes; insertion sequences and phages; PE/PPE; intermediary metabolism and respiration; regulatory proteins; conserved hypotheticals . We omitted the unknown category that contained only 16 proteins and the category corresponding to RNAs . Therefore , nine functional categories were used in this study . For the prediction of protein disorder IUPred was used ( http://iupred . enzim . hu/ ) [39] , [40] . The algorithm assigns a score between 0 and 1 for every residue in the protein . This score shows the tendency of that residue being disordered . For the binary categorization of residues we consider a residue disordered if it has a score greater than 0 . 5 , and ordered if its score is less than 0 . 5 . For the prediction of disordered binding regions we used ANCHOR ( http://anchor . enzim . hu/ ) [37] , [41] . Similarly to IUPred , ANCHOR calculates the tendency of each residue being in a disordered binding region . For binary classification , a cutoff of 0 . 5 was used here as well . The input for the clustering algorithm is based on the similarity profiles generated for each MTB sequence . Each profile consisted of 467 numbers that represent the number of sequences similar to the MTB sequence in the 467 studied proteomes . In the cluster analysis Euclidean distance was used together with Ward's method . The result of clustering was largely insensitive to various parameters of the clustering , including the type of the clustering method , various types of normalizations , parameters of PSI-BLAST . The clustering was implemented in the R program package . | Mycobacterium tuberculosis ( MTB ) , the causative agent of TB , is a dramatically successful pathogen that poses a considerable challenge for modern medicine . The increase in multi-drug resistant TB necessitates the identification of novel drug targets and drugs with new mechanisms of action . In this work , we developed a novel computational strategy based on comparative proteomic analysis that can highlight proteins involved in specifies-specific functions . Our analyses of the proteins encoded by the MTB genome identified two protein families that stand out in this respect . These proteins have complex architecture combining various domains and disordered segments . They are also involved in vital functions , especially in host-pathogen interactions . Although these proteins generally do not fit into traditional drug design paradigms , there are several new strategies emerging that can be used to target these proteins during drug development . Our results challenge current target selection protocols that largely rely on the uniqueness and the essentiality of proteins . Instead , these findings emphasize the importance of complex evolutionary scenarios that can lead to the emergence of species-specific functions from more ancient building blocks of proteins . The experiences gained from this work have important implications specifically for targeting MTB , and in broader terms , to improve current target selection protocols in drug development . | [
"Abstract",
"Introduction",
"Results/Discussion",
"Methods"
] | [
"sequence",
"analysis",
"systems",
"biology",
"biochemistry",
"genomics",
"biology",
"computational",
"biology",
"comparative",
"genomics",
"proteomics",
"pharmacogenomics"
] | 2011 | Proteins with Complex Architecture as Potential Targets for Drug Design: A Case Study of Mycobacterium tuberculosis |
Candida albicans is a major life-threatening human fungal pathogen . Host defence against systemic Candida infection relies mainly on phagocytosis of fungal cells by cells of the innate immune system . In this study , we have employed video microscopy , coupled with sophisticated image analysis tools , to assess the contribution of distinct C . albicans cell wall components and yeast-hypha morphogenesis to specific stages of phagocytosis by macrophages . We show that macrophage migration towards C . albicans was dependent on the glycosylation status of the fungal cell wall , but not cell viability or morphogenic switching from yeast to hyphal forms . This was not a consequence of differences in maximal macrophage track velocity , but stems from a greater percentage of macrophages pursuing glycosylation deficient C . albicans during the first hour of the phagocytosis assay . The rate of engulfment of C . albicans attached to the macrophage surface was significantly delayed for glycosylation and yeast-locked morphogenetic mutant strains , but enhanced for non-viable cells . Hyphal cells were engulfed at a slower rate than yeast cells , especially those with hyphae in excess of 20 µm , but there was no correlation between hyphal length and the rate of engulfment below this threshold . We show that spatial orientation of the hypha and whether hyphal C . albicans attached to the macrophage via the yeast or hyphal end were also important determinants of the rate of engulfment . Breaking down the overall phagocytic process into its individual components revealed novel insights into what determines the speed and effectiveness of C . albicans phagocytosis by macrophages .
Invasive C . albicans infection can present a serious clinical complication , especially in patients with an impaired immune system . Host defence against systemic candidiasis relies mainly on the ingestion and elimination of fungal cells by cells of the innate immune system , especially neutrophils and macrophages [1]–[3] . Despite the clinical importance of phagocytosis , this process remains poorly understood at a mechanistic level . The fungal cell wall is the first point of contact with the innate immune system and plays an important role in recognition and phagocytosis by host immune cells [2] . It is a dynamic , highly organized organelle that determines both the shape of the fungus and its viability . The core structure of the C . albicans fungal cell wall is composed of a skeleton of polysaccharide fibrils composed of β- ( 1 , 3 ) -glucan that is covalently linked to β- ( 1 , 6 ) -glucan and chitin ( a β- ( 1 , 4 ) -linked polymer of N-acetylglucosamine ) , and is designed to function as a robust exoskeleton and a scaffold for the external glycoprotein layer [4] . This outer layer consists of highly glycosylated mannoproteins that are modified by N-linked and O-linked mannosylation and phosphomannosylation [5] , [6] . Another important feature of C . albicans biology thought to play a major role in host recognition is the fungus' ability to undergo reversible morphological changes between yeast , pseudohyphal , and hyphal forms in response to environmental signals [7] , [8] . Its morphological plasticity is considered to be the most important virulence attribute of C . albicans [9] and plays a major role in the fungus' capacity to successfully infect many different anatomical sites of the human host . Hyphae have invasive properties that can promote tissue penetration and escape from immune cells [10] , whereas yeasts are suited to dissemination in the bloodstream [9] . Phagocytic clearance of fungal pathogens may be considered to consist of four distinct stages; ( i ) accumulation of phagocytes at the site where fungal cells are located; ( ii ) recognition of fungal pathogen-associated molecular patterns ( PAMPs ) through pattern recognition receptors ( PRRs ) ( reviewed in [11]; ( iii ) engulfment of fungal cells bound to the phagocyte cell membrane , and ( iv ) processing of engulfed cells within phagocytes by fusion with lysosomal vesicles to form the phagolysosome [12] . There is very limited information on how alterations in C . albicans morphogenesis or cell wall composition affect phagocyte migration towards the fungus . In contrast , a significant body of literature has identified an increasing number of PRRs and downstream signalling pathways that contribute to the recognition of fungal cells by macrophages [11] , [13] . These pathways have described recognition of N-linked mannans by the mannose receptor ( MR ) , O-linked mannans by Toll-like receptor 4 ( TLR4 ) , β-glucans by dectin-1/TLR2 , and α-mannosides by galectin-3/TLR2 complexes [14] . More recently , additional PRRs have been shown to contribute to C . albicans recognition , including the scavenger receptors CD36 and SCARF1 [15] , TLR9 recognition of nucleic acids [16] , dectin-2 [17] and the C-type lectin mincle [18] . Comparatively little is known about the engulfment process once the fungus is tethered to the phagocyte cell membrane . However , a series of studies have shed some light on how the overall phagocytic uptake process is affected by alterations in C . albicans cell wall composition , morphogenesis and macrophage activation state [10] , [19] . For example , we have recently shown that the glycosylation status of the C . albicans cell wall profoundly affected the rate of macrophage phagocytosis . Distinct patterns emerged in that phosphomannan deficient strains ( mnn4Δ , pmr1Δ , and mnt3Δ mnt5Δ ) were taken up at a lower rate than the wildtype or reintegrant controls , and that O-linked and N-linked mannan deficient strains are taken up at higher rates ( mns1Δ and mnt1Δmnt2Δ ) [10] . A study by Keppler-Ross et al . conducting competitive phagocytosis experiments suggested that macrophages displayed strong preferences for phagocytosis based on genus , species and morphology . Candida glabrata and Saccharomyces cerevisiae were taken up by J774 macrophage cells more rapidly than C . albicans , and C . albicans yeast cells were favoured over hyphal cells [20] . These studies are informative but are limited in that they assess phagocytosis in its entirety and do not break down any observed differences into individual stages of the process , such as migration , recognition or engulfment , which may be affected differentially . Furthermore , such studies assess uptake at selected time points , rather than as a continuous dynamic process , with the inherent disadvantage of ignoring temporal differences in individual stages of the phagocytosis process , which are likely to play a major role in determining the overall outcome of pathogen-host interactions in vivo . Here we have conducted a comprehensive analysis of C . albicans phagocytosis by primary macrophages and macrophage cell lines , employing video microscopy , coupled with sophisticated image analysis tools . To assess the contribution of C . albicans cell wall glycosylation and the ability to switch from yeast to hyphal forms , we have taken advantage of a large collection of genetically and phenotypically characterized isogenic mutants of C . albicans , depleted in specific cell wall components or impaired in morphogenic switching . We show here for the first time a detailed minute by minute account of the specific effects of C . albicans viability , cell wall composition , morphogenesis and spatial orientation on two distinct stages ( macrophage migration and engulfment of bound C . albicans ) of the phagocytosis process . These analyses revealed that macrophage migration towards C . albicans was dependent on the glycosylation status of the fungal cell wall , but not cell viability or morphogenic switching from yeast to hyphal forms . Macrophages rapidly engulfed viable and UV-killed C . albicans , but engulfment was protracted for all glycosylation and morphogenetic mutants examined . Engulfment of hyphal C . albicans was determined by multiple components including hyphal length and spatial orientation .
C . albicans phagocytosis by macrophages is dependent on the C . albicans cell wall glycosylation status [10] , but the question remains whether differences observed in overall uptake are a consequence of changes in migration of macrophages towards C . albicans or alterations in the engulfment process itself . Live cell video microscopy enabled examination of the individual stages of the uptake process . Representative videos are available to view in Supporting Videos S1 and S2 . First we addressed the question of whether alterations in C . albicans cell wall glycosylation and morphogenesis affect migration of macrophages towards C . albicans . Primary macrophages and macrophage cell lines were challenged with glycosylation and morphogenesis defective strains of C . albicans . The strains used in this study are shown in Table 1 . Briefly , the mnt1Δmnt2Δstrain is deficient in O-glycosylation [21] and has only a single O-linked mannose sugar . The mns1Δ strain has an N-glycosylation defect due to curtailed α1 , 2-mannosidase activity in the endoplasmic reticulum [22] and the mnn4Δ strain has a complete loss of phosphomannan [23] . Morphogenesis defective strains included the hgc1Δ strain , a G1 cyclin mutant that is unable to form true hyphae , and efg1Δ that lacks a specific transcription factor that regulates yeast-hypha morphogenesis pathways [24] , [25] . Migration of macrophages was assessed by live cell video microscopy using our standard phagocytosis assay 10 , 26 , with track measurements taken at 1 min intervals over a 6 h period . Figures 1A , B and C show images derived from video microscopy depicting the track of a single macrophage migrating towards and engulfing live C . albicans ( wildtype strain ) . Initially , the macrophage's movement appeared to be random ( Figure 1A ) . However , dynamic analysis suggested that macrophages sensed C . albicans , accelerated and homed in on their target ( Figure 1B ) , leading to cell-cell contact and engulfment ( Figure 1C ) . The corresponding video is available to view in Supporting Video S3 . Visual inspection of the videos suggested enhanced macrophage migration towards C . albicans glycosylation mutants , in particular the mnt1Δmnt2Δstrain , compared to wildtype control . The suggestion that migration was enhanced in macrophages exposed to the mnt1Δmnt2Δ mutant strain was further supported by macrophage tracking diagrams ( Figures 1D and 1E ) . Tracking diagrams ( Figures 1D and E ) illustrate the distances travelled , directionality and velocity of macrophages cultured with live wildtype and mnt1Δmnt2Δ , respectively . Due to the large number of macrophages tracked per video , the data were filtered to show only macrophages with a mean track velocity greater than that of inactive macrophages not pursuing fungal cells ( 1 . 80 µm/min ) . Tracks represent the movement of individual macrophages relative to their starting position , symbols indicate the location of macrophages at 1 min intervals and arrows represent directionality . These diagrams illustrate that although macrophages can migrate rapidly and for long distances when cultured with both live wildtype and the mnt1Δmnt2Δ mutant , when incubated with live mnt1Δmnt2Δ a higher number of macrophages have a mean track velocity of greater than 1 . 80 µm/min ( Figure 1E ) . Quantitative analysis of average macrophage track velocity for the entire length of the observation period ( 6 h ) showed no significant differences between wildtype ( 1 . 8 µm/min ± 0 . 02 SE ) and yeast-locked morphogenetic mutants , but confirmed enhanced migration with UV-killed wildtype C . albicans ( 1 . 94±0 . 02 SE , p<0 . 05 ) and the glycosylation mutants mnt1Δmnt2Δ ( 2 . 1±0 . 02 SE , p<0 . 001 ) , mns1Δ ( 2 . 09±0 . 03 SE , p<0 . 001 ) and mnn4Δ ( 1 . 96±0 . 03 SE , p<0 . 01 ) ( Figure 2B ) . The macrophage average track velocity was highest for the first 30 min of the phagocytosis assay ( Figure 2B ) . The data for this period again showed increased average track velocity for the glycosylation mutant strains mnt1Δmnt2Δ ( 2 . 68±0 . 04 , p<0 . 001 ) , mns1Δ ( 2 . 47±0 . 05 , p<0 . 05 ) and mnn4Δ ( 2 . 52±0 . 07 , p<0 . 01 ) when compared with wildtype ( 2 . 19±0 . 07 ) ( Figure 2A ) . In contrast , there was no significant difference in the mean track velocity of macrophages when incubated with morphogenesis defective mutants and UV-killed wildtype . Overall track lengths were measured for the first 30 min and 6 h , and not surprisingly the data reflected the average track velocity for the wildtype and mutant strains tested ( data not shown ) . Enhanced macrophage migration in phagocytosis assays with C . albicans glycosylation mutants was not a consequence of alterations in maximal macrophage velocity ( average max velocity in µm/min: wildtype , 3 . 7±0 . 2; mnt1Δmnt2Δ , 3 . 6±0 . 1; mnn4Δ , 3 . 6±0 . 2; mns1Δ ( 3 . 1±0 . 2 ) but rather a reflection of increased macrophage activity , particularly during the first hour of the interaction assay . Experiments with primary murine peritoneal macrophages showed a similar pattern . We observed no differences between the yeast-locked mutant strain hgc1Δ and wildtype but significantly increased average track velocity for the glycosylation mutant strain mnt1Δmnt2Δ ( p<0 . 001 ) for the first 30 min and 6 h of the phagocytosis assay ( Table 2 ) . Overall the mean track velocity of peritoneal macrophages was found to be significantly lower than for J774 . 1 macrophages ( p<0 . 001 ) . However , the maximum velocity of peritoneal macrophages ( 3 . 9±0 . 2 µm/min ) was comparable to J774 . 1 macrophages ( 3 . 7±0 . 2 µm/min ) . We hypothesised that the difference in mean track velocity between peritoneal and J774 . 1 macrophages is due in part to the peritoneal macrophages being more spread out and covering a larger surface area , therefore , reducing the need to migrate to achieve close proximity with fungal cells in the phagocytosis assay . However , experiments using the same macrophage:C . albicans ratios but lower macrophage densities confirmed that this was not the case , as mean track velocities were unchanged ( Table 2 ) . Thus , changes in C . albicans cell wall composition but not hyphal morphogenesis markedly influenced macrophage migration in vitro . Effective migration of macrophages towards C . albicans is necessary to establish cell-cell contact , which is a prerequisite for initiation of the engulfment process . Next we addressed the question of whether alterations in C . albicans cell wall glycosylation and morphogenesis affected the ability and speed by which macrophages engulfed C . albicans after cell-cell contact was established . Live cell video microscopy coupled with image analysis generated a detailed minute by minute account of the engulfment process ( Figures 3A , B and C ) . Wildtype C . albicans was shown to be rapidly engulfed by macrophages once cell-cell contact was established ( Figure 3D ) . A three dimensional projection image confirming C . albicans phagocytosis is available in Supporting VideoS4 . The average time taken for engulfment of wildtype C . albicans is 6 . 7±0 . 3 min , and the vast majority ( 95% ) of fungal cells were engulfed within 15 min . UV-killed C . albicans yeast cells were engulfed even more swiftly , with all cells taken up within 15 min and engulfment taking an average of 4 . 2±0 . 1 min ( Figure 3E ) . Interestingly , the rate of engulfment of all glycosylation mutant strains ( Figures 4B–D ) was significantly slower than that of wildtype C . albicans ( mnt1Δmnt2Δ ) ( p<0 . 001 ) ; mns1Δ ( p<0 . 01; mnn4Δ ( p<0 . 001 ) ( Figure 4A ) . The delayed engulfment was most marked for the mnt1Δmnt2Δ ( 13 . 5±0 . 7 min ) and mnn4Δ ( 14 . 4±0 . 9 min ) mutant strains ( Figures 4B and D ) , as macrophages on average took twice as long to engulf these mutants than wildtype C . albicans ( Figure 4A ) . Control strains mnt1Δmnt2Δ::MNT1 and mnn4Δ::MNN4 , containing a single reintegrated copy of the corresponding deleted genes , partially restored the ability of macrophages to swiftly engulf C . albicans ( data not shown ) . Experiments using primary thioglycollate elicited murine peritoneal macrophages and human monocyte derived macrophages also showed a significant delay for the engulfment of the glycosylation deficient mutant mnt1Δmnt2Δ and this was partially restored in the corresponding reintegrant control mnt1Δmnt2Δ::MNT1 ( Table 3 ) . Engulfment of yeast-locked morphogenetic mutant strains was delayed and ultimately impaired , relative to wildtype controls in J774 macrophages ( p<0 . 001 ) . Firstly , the average time taken for engulfment of the hgc1Δ ( Figure 5A and B ) and efg1Δ ( Figure 5A and C ) mutant strains was significantly greater than for the wildtype control ( 16 . 2±1 . 4 min , 21 . 9±2 . 4 min and 6 . 7±0 . 3 min , respectively ) . Engulfment of approximately 1 . 5% of wildtype C . albicans took longer than 30 min , compared with approximately 11% and 17% for hgc1Δ and efg1Δ , respectively . Secondly , we observed that approximately 2% of hgc1Δ and 41% of efg1Δ mutants that established contact with macrophages were not internalised , even after prolonged cell-cell contact . In contrast , all wildtype C . albicans yeasts were successfully engulfed following recognition . However , 66% of the efg1Δ mutant cells were eventually engulfed by neighbouring phagocytes after detachment from the macrophage they were originally in contact with . Experiments using peritoneal macrophages showed a non significant delay in the engulfment of the yeast locked mutant strain hgc1Δ ( Table 3 ) and little evidence of detachment once cell-cell contact was established . Thus , macrophages rapidly engulfed viable and UV-killed C . albicans after cell-cell contact was established , but engulfment was markedly slower for all glycosylation ( in all macrophage subsets studied ) and yeast-locked ( solely in the J774 macrophage cell line ) morphogenetic mutants examined . The data above showed that UV-killed yeast cells were engulfed more rapidly than live wildtype cells that were able to form hyphae . The accelerated engulfment of UV-killed cells raised questions about how cell morphology affects engulfment of C . albicans by macrophages . We examined the engulfment of wildtype C . albicans cells that had established cell-cell contact with macrophages in either yeast or hyphal morphology . Hyphal C . albicans cells were engulfed at a slower rate than C . albicans yeast cells ( 10 . 8±0 . 9 min and 5 . 6±0 . 3 min , respectively ) . Furthermore , the vast majority ( 98% ) of yeast cells of C . albicans were taken up within 15 min ( Figure 6A ) , whereas there was greater variability for hyphal cells of C . albicans , with 21% taking longer than 15 min to become engulfed ( Figure 6B ) . Next we examined whether hyphal length influenced the speed of engulfment , perhaps explaining the variations observed for hyphal C . albicans engulfment . Macrophages were capable of ingesting C . albicans with hyphae of more than twice the average diameter of macrophages ( the maximum observed length of ingested hyphae was 42 . 9 µm ) , but the mean hyphal length at time of engulfment was 8 . 7±0 . 7 µm . Intriguingly , and contrary to expectations , we found no correlation between hyphal length and speed of engulfment for hyphal cells of C . albicans of less than 20 µm length ( Figure 6C ) . However , when hyphal length exceeded 20 µm there was as significant impact on the macrophage's ability to engulf C . albicans ( Figure 6D ) . Although macrophages engulfed C . albicans with hyphae larger than 20 µm , uptake was markedly slower with 64% of uptake events requiring more than 15 min . It is worth noting that despite having difficulty engulfing large hyphae macrophages were nonetheless persistent in their attempt to do so . Thus , macrophages were more effective at engulfing yeast cells rather than hyphal cells of C . albicans and engulfment of hyphal cells was influenced in part by hyphal length , with a cut off of 20 µm , above which macrophage engulfment was markedly impaired . Finally , we took advantage of the large quantity of data amassed from live cell video microscopy phagocytosis assays to address previously unanswered questions relating to how spatial orientation of C . albicans may affect the efficiency of engulfment by macrophages . First , we established that hyphal cells of C . albicans could be taken up by macrophages independent of their spatial orientation ( Figure 7A ) . C . albicans germ tubes could be engulfed yeast-end on and germ tube apex-end on ( Figures 7B and C ) , side-on ( Figure 7D ) and at an angle ( Figure 7E ) . However , it is noteworthy that although cell-cell contact could be initiated in any orientation , the rate of engulfment was affected; C . albicans that made contact in an end-on orientation were taken up more rapidly than those engulfed at an angle or where cell-cell contact was initiated side-on ( 5 . 5±0 . 7 min , 8 . 8±0 . 5 min and 9 . 5±2 . 3 min , respectively ) . The large SE observed when C . albicans makes contact side-on can be explained by the fact that macrophages had particular difficulty ingesting large hyphae ( >20 µm ) in the side-on orientation . The end initially encountered by the macrophage appeared to be random . Approximately equal numbers of encounters occurred that were yeast end-on or hyphal end-on , but there was a propensity for C . albicans to be taken up more rapidly yeast-end on ( 8 . 0±0 . 6 min ) than hyphal end-on ( 9 . 8±1 . 0 min ) . Initial C . albicans orientation when establishing cell-cell contact with macrophages influenced the rate of engulfment with end-on contact of the hyphal end resulting in the most rapid engulfment . Thus , engulfment of hyphal cells of C . albicans was influenced by multiple factors including hyphal length and spatial orientation , and whether the initial encounter was by the yeast or hyphal end .
C . albicans is a major life-threatening human fungal pathogen . Host defence against systemic Candida infection relies mainly on phagocytosis of fungal cells by cells of the innate immune system . In this study , we analysed the contribution of distinct C . albicans cell wall components and yeast-hypha morphogenesis to specific stages of phagocytosis by macrophages . We show that macrophage migration towards C . albicans was dependent on the glycosylation status of the fungal cell wall , but not cell viability or morphogenic switching from yeast to hyphal forms . This finding was not a consequence of differences in maximal macrophage track velocity , but stems from a greater percentage of macrophages pursuing glycosylation deficient C . albicans cells during the first hour of the phagocytosis assay . The rate of engulfment of C . albicans by macrophages was significantly slower for glycosylation and morphogenesis deficient mutant strains , but enhanced for non-viable cells . Hyphal cells were engulfed at a slower rate than yeast cells , especially those with hyphae in excess of 20 µm , but there was no correlation between hyphal length and the rate of engulfment below this threshold . We show that spatial orientation of the hypha and whether hyphal C . albicans attached to the macrophage via the yeast or hyphal end were also important determinants of the rate of engulfment . This is the first study , to our knowledge , to show that individual stages of C . albicans phagocytosis by macrophages are differentially affected by changes in C . albicans cell wall composition . Our previous work , using assays that globally assess phagocytosis , have shown increased phagocytosis of O-linked and N-linked mannan deficient strains ( mns1Δ and mnt1Δ mnt2Δ ) [10] . Intriguingly , we show here that changes in cell wall glycosylation enhance macrophage migration towards C . albicans , but delay engulfment once cell-cell contact is established . This illustrates that standard assays do not differentiate between the individual stages of the phagocytosis process and are unable to detect significant temporal differences in migration or engulfment . For example , we show here that phosphomannan deficient cells of C . albicans were engulfed less efficiently . This effect was much less obvious in previous studies that simply evaluated phagocytosis efficiency by single time point measurements [27] . Macrophage migration was enhanced in all glycosylation mutants but most markedly in the mnt1Δmnt2Δ O-glycosylation mutant . This translated into much higher overall uptake compared to the phosphomannan deficient mmn4Δ mutant and is in keeping with our previous published results [10] . It is conceivable that enhanced macrophage migration in response to the absence of O-linked ( mnt1Δmnt2Δ ) or N-linked mannans ( mns1Δ ) is a consequence of unmasking underlying β-glucans [28] , [29] or electrostatic signals as a consequence of alterations in surface charge following loss of phosphomannan ( mnn4Δ ) [30] , [31] . Observation of individual macrophage migration patterns indicated that macrophage movement was slow and random initially , but became directional towards a specific C . albicans cell , associated with a marked increase in macrophage velocity . Macrophage acceleration towards C . albicans occurred at distances in excess of 15 µm and , therefore , suggests the presence of a chemotactic signal . Key candidates are a number of glycolipids that are known to be shed by C . albicans and are potent inducers of macrophage cytokine synthesis in vitro and in vivo [32] . We are currently conducting detailed mathematical modelling of the macrophage tracking patterns to further elucidate the hypothesis that macrophage migration towards C . albicans is affected by differences in shedding of glycolipids between wildtype and glycosylation deficient strains . We have shown that macrophage uptake of C . albicans is a multi-step process , involving recognition and subsequent engulfment of C . albicans . C . albicans cell wall mannosylation is a key determinant in the rate of engulfment; the absence of specific PAMPs in the glycosylation mutant strains delays engulfment once cell-cell contact has been established , and this may be a consequence of differential activation of macrophage PRRs . This is in keeping with experiments in which C . albicans cell wall mutants were combined with specific macrophage receptor blocking methods that have been used to define the PAMP-PRR interactions required for cytokine induction [14] , [33] . These in vitro findings are relevant to C . albicans infections in vivo , since C . albicans mutants with defects in cell wall mannosyl residues are also less virulent in experimental models of disseminated candidiasis [21] , [22] , [34]–[36] . Morphological plasticity is one of the hallmarks of the human fungal pathogen C . albicans [37] , and its ability to switch between yeast and hyphal forms is thought to contribute to pathogenesis [24] , [25] , [38] , [39] . C . albicans mutants that are unable to form filaments are less virulent [40] , although conversely , mutants that are unable to grow as yeast are also less virulent [41] . There are conflicting reports in the literature regarding the efficiency of macrophage phagocytosis for C . albicans yeast and hyphal forms [20] , [42] , [43] . Here we show definitively data supporting the notion that macrophages are more effective at engulfing C . albicans yeasts . Furthermore , the use of video microscopy coupled with thorough analysis of large numbers of individual macrophage-C . albicans interactions provides a minute-by-minute account of the engulfment process , which offers detail that has not been previously available . A prime example is our observation that yeast-locked C . albicans cells were engulfed less efficiently than wildtype C . albicans . Not only was this not obvious in previous studies that simply evaluated phagocytosis efficiency by single time point measurements [10] , but in addition , we observed here that delayed engulfment can result in detachment of the fungal cell and engulfment by a neighbouring macrophage . One may speculate that in vivo where phagocyte numbers are limited this may have a significant impact on pathogen clearance and that yeast locked mutant C . albicans cells have properties other than the induced phenotype that differ from wildtype yeast C . albicans cells . However , the observed delay in engulfment for yeast-locked mutant C . albicans in the macrophage cell line was almost completely abrogated in experiments using primary macrophages , underlining the importance of studying host-pathogen interactions in multiple phagocyte subsets . The approach taken here further enabled us to dissect the complexity of engulfment of hyphal C . albicans by macrophages . We show that macrophages are capable of engulfing hyphal C . albicans in excess of 40 µm ( approximately twice the diameter of macrophages ) - in keeping with reports that macrophages are capable of ingesting apoptotic epithelial cells in the involuting mammary gland of similar or even larger size [44] . Hyphal length , however , does play a major role in the engulfment process of C . albicans , in that engulfment of C . albicans with hyphae in excess of 20 µm took significantly more time and phagocytosis was frequently frustrated . Interestingly , we show that below a 20 µm hyphal length threshold there was no correlation between hyphal length and the rate of engulfment . These observations are most likely related to difficulties associated with macrophages attempting to engulf very large particles . In addition to hyphal length , we have identified two other factors that influenced engulfment of hyphal C . albicans . We showed that the rate of engulfment was determined by the orientation in which C . albicans was encountered , with end-on being favourable to side-on orientation , suggesting that steric hindrance affects engulfment . We also showed that yeast end-on engulfment was more efficient than hyphal end-on encounters . This in turn may reflect differences in the wall chemistry of the hyphal tip compared to the mother cell , or be due to the efficiency of the assembly of proteins of the phagocytic cup for objects of different sizes and shapes [45] . Here we have conducted the most detailed analysis of the contribution of C . albicans viability , cell wall glycosylation and morphogenesis to phagocytosis by macrophages to date , to our knowledge . Our approach of combining live cell video microscopy with image analysis tools for the migration analysis , and minute-by-minute analysis of thousands of individual macrophage-C . albicans interactions , provides unique insight into the complexity of C . albicans phagocytosis by macrophages . The novel methods employed here to study phagocytosis of C . albicans could be applied to study other pathogens and uptake of dying host cells . Such studies would significantly enhance our understanding of the mechanisms that govern effective phagocytosis and ultimately the innate immune response to infection .
All animal experiments have been conducted in strict accordance with UK Home Office guidelines . The appropriate project and personal licenses are in place PIL 60/6194 and approved by the UK Home office . C . albicans serotype A strain CAI4+CIp10 , hitherto referred to as the parental wildtype , was used as a control and its parent strain , CAI4 , was used to construct mutants using targeted gene disruption [46] . The mutants used are listed in Table 1 . C . albicans strains containing a single reintegrated copy of the corresponding deleted genes to regenerate the heterozygous genotype acted as controls . Most of the C . albicans strains used were created in house and have been described previously [21]–[25] . C . albicans strains were obtained from glycerol stocks stored at −80°C , and plated on SC-Ura plates ( except hgc1Δand efg1Δ ) . SC-Ura plates consist of 6 . 9 g yeast nitrogen base without amino acids ( Formedium , Norfolk , UK ) , 1 ml 1 M NaOH ( BDH Chemicals , VWR International , Leicestershire , UK ) , 10 ml 1% ( w/v ) adenine hemisulphate salt ( Sigma , Dorset , UK ) , 50 ml 40% D-glucose ( Fisher Scientific , Leistershire , UK ) , 50 ml 4% SC-Ura dropout ( Formedium , Norfolk , UK ) and 2% ( w/v ) technical agar ( Oxoid , Cambridge , UK ) made up to 1000 ml in distilled H2O . The C . albicans morphogenetic mutants hgc1Δ and efg1Δ were grown on YPD plates consisting of 1% yeast extract ( Duchefa Biochemie , Haarlem , Holland ) , 2% mycopeptone ( Oxoid , Cambridge , UK ) , 2% D-glucose and 2% technical agar in distilled H2O . All plates were incubated at 30°C until colonies formed , and were then stored at 5°C . Intraperitoneal injections of 1 ml Brewer's thioglycollate broth ( BD , New Jersey , USA ) were administered to 8 week old female BALB/c mice . After 4 days , the peritoneal cavity of sacrificed mice was lavaged with 5 mM EDTA in 1× PBS , to harvest thioglycollate-induced macrophages . These Thio-macrophages were washed 3 times with RPMI medium 1640 ( Sigma , Dorset , UK ) supplemented with 10% ( v/v ) foetal calf serum ( FCS ) ( Biosera , Ringmer , UK ) , 200 U/ml penicillin/streptomycin antibiotics ( Invitrogen Ltd , Paisley , UK ) , 10 mM HEPES ( Invitrogen Ltd , Paisley , UK ) and 2 mM L-glutamine ( Invitrogen , Paisley , UK ) . For phagocytosis assays , 1×106 thio-macrophages in 2 ml supplemented RPMI medium were seeded onto glass bottomed Iwaki dishes ( VWR , Leistershire , UK ) and cultured overnight at 37°C with 5% CO2 . Immediately prior to experiments , RPMI medium was replaced with 2 ml pre-warmed supplemented CO2-independent medium ( Gibco , Invitrogen , Paisley , UK ) containing 1 µM LysoTracker Red DND-99 ( Invitrogen , Paisley , UK ) . LysoTracker Red DND-99 is a red fluorescent dye that stains macrophage acidic organelles , enabling macrophage paths to be tracked using Volocity 5 . 0 software ( Improvision , PerkinElmer , Coventry , UK ) . J774 . 1 macrophages ( ECACC , HPA , Salisbury , UK ) were maintained in tissue culture flasks in DMEM medium ( Lonza , Slough , UK ) , supplemented with 10% ( v/v ) FCS ( Biosera , Ringmer , UK ) , 200 U/ml penicillin/streptomycin antibiotics ( Invitrogen , Paisley , UK ) and 2 mM L-glutamine ( Invitrogen , Paisley , UK ) at 37°C with 5% CO2 . Human monocyte derived macrophages were prepared as previously described in detail [26] . For phagocytosis assays , 1×106 J774 . 1 macrophages in 2 ml supplemented DMEM medium were seeded onto glass based Iwaki dishes ( VWR , Leistershire , UK ) and cultured overnight at 37°C with 5% CO2 . Immediately prior to experiments , DMEM medium was replaced with 2 ml pre-warmed supplemented CO2-independent medium ( Gibco , Invitrogen , Paisley , UK ) containing 1 µM LysoTracker Red DND-99 ( Invitrogen , Paisley , UK ) . Single C . albicans colonies from plates stored at 5°C were cultured in 5 ml SC-Ura/YPD medium ( recipes as above , excluding technical agar ) and incubated overnight at 30°C , 200 rpm . In order to determine the impact of C . albicans viability on macrophage migration and engulfment , wildtype C . albicans were killed by UV-irradiation; 100×106 fungal cells in 1 ml 1× PBS were exposed to 20 doses of UV irradiation at 100 mJ/cm2 in 6 well plates . To aid visualisation of C . albicans during phagocytosis assays , 100×106 live or UV-killed C . albicans were stained using 1 mg/ml FITC ( Sigma , Dorset , UK ) in 0 . 05 M carbonate-bicarbonate buffer ( pH 9 . 6 ) ( BDH Chemicals , VWR International , Leicestershire , UK ) for 10 min at room temperature in the dark . Fungal cells were washed 3 times in PBS to remove unbound FITC and resuspended in 1× PBS . Our standard phagocytosis assays were performed as previously described [10] . In brief , 3×106 FITC-stained C . albicans were added to 1×106 macrophages in glass based Iwaki dishes ( VWR , Leistershire , UK ) immediately prior to imaging . Video microscopy experiments were performed using a DeltaVision Core microscope ( Applied Precision , Washington , USA ) with an environmental control chamber set at 37°C . Images were captured at 1 min intervals for 6 h using an EMCCD camera . At least two independent experiments were conducted for each C . albicans strain , and at least 3 movies were analysed from each experiment . One hundred macrophages were randomly selected from each movie and their phagocytic activity determined ( as below ) . Volocity 5 . 0 imaging analysis software was used to track macrophage migration at 1 min intervals throughout the 6 h phagocytosis assay . The software enabled high throughput analysis of macrophage migration , providing detailed information on the distances travelled , directionality and velocity of thousands of individual macrophages . Data were subsequently displayed in tracking diagrams and used to calculate the mean track velocity and track length of macrophages cultured with C . albicans . These analyses enabled assessment of the affects of C . albicans viability , glycosylation status and morphology on migration . One hundred macrophages from each movie were analysed individually at 1 min intervals throughout the 6 h phagocytosis assay . Measurements taken include the time points at which initial cell-cell contact occurred and at which C . albicans was fully enclosed , the number of C . albicans taken up and their morphology , the orientation of hyphal C . albicans relative to the macrophage and hyphal length . The rate of engulfment of live and UV-killed wildtype C . albicans , and glycosylation and yeast-locked morphogenetic mutant C . albicans was calculated by subtracting the time point at which initial cell-cell contact occurred from the time point at which the fungus was fully phagocytosed . This enabled accurate assessment of the affects of C . albicans viability , glycosylation status and morphology on the speed of engulfment . C . albicans spatial orientation , morphology , hyphal length and the end of hyphal C . albicans recognised were determined to assess whether these factors impact on the rate of engulfment . This strategy enabled in depth analysis of individual C . albicans-macrophage interactions in real time . Mean values and standard errors were calculated . One-way analysis of variance ( ANOVA ) and Tukey-Kramer Multiple Analysis Comparison Tests were used to determine statistical significance . | Host defence against systemic candidiasis relies mainly on the ingestion and elimination of fungal cells by cells of the innate immune system , especially neutrophils and macrophages . Here we have used live cell video microscopy coupled with sophisticated image analysis to generate a temporal and spatial analysis in unprecedented detail of the specific effects of C . albicans viability , cell wall composition , morphogenesis and spatial orientation on two distinct stages ( macrophage migration and engulfment of bound C . albicans ) of the phagocytosis process . The novel methods employed here to study phagocytosis of C . albicans could be applied to study other pathogens and uptake of dying host cells . Thus , our studies have direct implications for a much broader community and provide a blueprint for future studies with other phagocytes/microorganisms that would significantly enhance our understanding of the mechanisms that govern effective phagocytosis and ultimately the innate immune response to infection . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"mycology",
"immune",
"cells",
"immunology",
"biology",
"microbiology"
] | 2012 | Stage Specific Assessment of Candida albicans Phagocytosis by Macrophages Identifies Cell Wall Composition and Morphogenesis as Key Determinants |
Cell heterogeneity and the inherent complexity due to the interplay of multiple molecular processes within the cell pose difficult challenges for current single-cell biology . We introduce an approach that identifies a disease phenotype from multiparameter single-cell measurements , which is based on the concept of “supercell statistics” , a single-cell-based averaging procedure followed by a machine learning classification scheme . We are able to assess the optimal tradeoff between the number of single cells averaged and the number of measurements needed to capture phenotypic differences between healthy and diseased patients , as well as between different diseases that are difficult to diagnose otherwise . We apply our approach to two kinds of single-cell datasets , addressing the diagnosis of a premature aging disorder using images of cell nuclei , as well as the phenotypes of two non-infectious uveitides ( the ocular manifestations of Behçet's disease and sarcoidosis ) based on multicolor flow cytometry . In the former case , one nuclear shape measurement taken over a group of 30 cells is sufficient to classify samples as healthy or diseased , in agreement with usual laboratory practice . In the latter , our method is able to identify a minimal set of 5 markers that accurately predict Behçet's disease and sarcoidosis . This is the first time that a quantitative phenotypic distinction between these two diseases has been achieved . To obtain this clear phenotypic signature , about one hundred CD8+ T cells need to be measured . Although the molecular markers identified have been reported to be important players in autoimmune disorders , this is the first report pointing out that CD8+ T cells can be used to distinguish two systemic inflammatory diseases . Beyond these specific cases , the approach proposed here is applicable to datasets generated by other kinds of state-of-the-art and forthcoming single-cell technologies , such as multidimensional mass cytometry , single-cell gene expression , and single-cell full genome sequencing techniques .
In the life sciences , there is now a wealth of quantitative information from simultaneous measurements on many proteins and genes , from small tissue samples down to a single cell at a time [1]–[6] . Likewise , bioimaging is following a similar trend through multicolor fluorescent imaging and the emerging ability to carry out spatially resolved vibrational spectroscopy of living cells in close to real-time [7] , [8] . These groundbreaking technologies have resulted in a plethora of information for single cells , which can be represented as points in a high-dimensional space . Here we show how one can tease out the essential information from such high-dimensional data in order to diagnose human diseases and understand their molecular origins . Our approach tackles two interlinked challenges inherent to high-dimensional , single-cell information . First , single-cell measurements exhibit vast heterogeneity in the behavior of individual cells: even a simple bell-shaped distribution can contain subpopulations enriched for biologically distinct functions . For instance , subpopulations of clonally derived hematopoietic progenitor cells with low or high expression of the stem cell marker Sca-1 were observed to be in dramatically different transcriptional states and to give rise to different blood cell lineages [9] . Second , cell phenotypes are emergent products of multiple molecular actions: the phenotype of a tissue or organism often requires not only multiple cells , but also multiple attributes at the cellular level , which makes bridging scales from molecular and cellular level information to disease diagnosis a challenging , oftentimes elusive goal [10] . Here we present a new approach to analyze high-dimensional single-cell information , and apply it to two representative datasets . We address the diagnosis of progeria , a premature aging disorder [11] , where single-cell data are obtained by an automated nuclear shape analysis from hundreds of healthy and diseased cells . We also develop a multiparameter phenotype in order to distinguish two sight threatening non-infectious uveitides , the ocular manifestations of Behçet's disease and sarcoidosis , based on multicolor flow cytometry information on tens of proteins from fresh blood patient samples . Our emphasis is to assess the optimal tradeoff between the number of single cells averaged and the number of measurements needed to capture phenotypic difference . The number of available cells may be a key limiting factor when target cell subpopulations are extremely small ( e . g . hematopoietic stem cells from bone marrow or blood samples ) or when the experimental techniques are not easily scalable ( e . g . single-cell imaging and single-cell gene expression ) . In the next Section , we describe some common approaches to analyze multidimensional single-cell datasets , we show their shortcomings due to cell heterogeneity and the inherent multidimensional nature implied in a complex phenotype , and we apply our approach to the two specific cases mentioned above . In the following Section , we provide a summary and a discussion of our findings .
A commonly used method to visualize and analyze multidimensional single cell information is through sequential selection of subtypes of cells based on simple thresholds , applied to one or two parameters at a time [12] . This procedure is generally represented as a sequence of two-dimensional plots , where one attribute is plotted against another one . This method works extremely well when simple thresholds for just a few parameters lead to reliable phenotypes . However , for complex diseases such as Behçet's and sarcoidosis , even the best choice of parameters is not enough to identify a phenotype . A representative example is shown in Fig . 1A ( i ) : CD8+ T cells have very similar combinations of IL22 and CD3 levels in both Behçet's disease and sarcoidosis , even though – as we will show below – these parameters play a key role in distinguishing between the two diseases . Similarly , highly overlapping populations are observed for other cell types we investigated ( e . g . CD4+ T cells ) and other pairs of markers studied . This indicates that the distinction between Behçet's disease and sarcoidosis can only be discerned using a combination of more than two parameters , and thus is difficult to visualize and detect with established approaches . Going beyond two parameters , some mathematical tools are able to reduce the dimensionality of high-dimensional data [13] , [14] . Singular value decomposition is a simple , yet powerful technique for generating low dimensional representations [13] . However , the optimal axes selected by such a method are not designed to distinguish between health and disease , or help diagnose the disease . This is evident in Fig . 1A ( ii ) , where the two top eigenmodes from a singular-value decomposition analysis of 16-dimensional data are plotted for the same CD8+ T cell subpopulation , showing again a large overlap between the two diseases . Even in cases where a single parameter can be established as a suitable phenotype , cell-to-cell heterogeneity presents a challenge . For example , in Hutchinson-Gilford progeria syndrome ( HGPS ) , a rare genetic disease of accelerated aging , the number of “blebs” or localized protrusions visible in a cell's nucleus is an established cellular marker of HGPS [15] . However , that does not imply that a single cell showing blebs indicates HGPS . Instead , as shown by Fig . 1B ( i ) – ( iv ) , blebbed and non-blebbed nuclei are observed both within healthy and diseased cell lines . On average , nevertheless , blebbing is a reliable phenotype , as illustrated in Fig . 1B ( v ) – ( vi ) . This raises the question: can one simply measure other aspects of the nuclear shape with additional metrics to establish a disease phenotype from a single cell , or does cell heterogeneity require us to investigate the properties of cell ensembles for a reliable diagnosis ? The tradeoff between multidimensional measurements and the number of cells needed to achieve a desired confidence level of prediction certainly requires an unbiased , fully quantitative , and mathematically robust method . Here we introduce and apply an approach to develop a disease phenotype from multiparameter single-cell measurements . Our approach uses simple machine learning methods to determine what combination of parameters can serve as an indicator of disease , and how many parameters are needed to diagnose a disease . While machine learning of disease diagnostics is not new , it often fails when applied at the single-cell level due to the heterogeneity of cells . It also fails when average quantities are measured if the number of patients is not large enough for a machine learning approach . The simple additional step of averaging over a small number of cells - here tens to hundreds of cells – and varying that number allows us to optimize our ability to detect a disease phenotype . This procedure smoothes out single cell heterogeneity and , at the same time , minimizes the loss of information due to averaging . For machine learning purposes , each patient is still represented by a point cloud in parameter space , but now each point represents a group of cells , rather than an individual cell . Recently , several groups have developed computational methods for identifying cell populations in multidimensional flow cytometry data . Their goals are two-fold: on the one hand , to determine whether automated algorithms can reproduce expert manual gating; on the other hand , to determine whether analysis pipelines can identify characteristics that correlate with external variables such as clinical outcome . In the latter case , flow cytometry data is transformed into class-labeled vectors in instance space by a variety of methods such as binning of 2D and 3D measurement histograms , Gaussian mixtures , 1D and sequential gating schemes , and cell clustering using k-means and other high-dimensional clustering techniques [16]–[23] . A detailed description and comparative assessment of the performance of different approaches has been recently reported [24] . Within this context , it is important to point out that the method proposed in our work addresses the problem of phenotypic classification when single cells are highly heterogeneous and when the number of cells available may be rather small ( just a few tens or hundreds , as opposed to typical flow cytometry experiments in which the number of measured cells is one or several orders of magnitude larger ) . We will demonstrate that our method is generally applicable to different kinds of multidimensional single-cell data and one of our examples is on flow- cytometry-based phenotypes . However , the key contribution is the development of a framework that provides a quantitative assessment of the critical sample size and number of simultaneous single-cell measurements needed to identify a phenotype with strong predictive power . State-of-the-art single-cell genomics and single-cell imaging technologies are examples in which the number of measured single cells is critically small , and where flow cytometry data analysis methods that rely on high-dimensional clustering procedures , Gaussian mixture approximations , etc may be expected to fail . We will tackle the tradeoff between the number of parameters and the number of cells needed first on the example of HGPS - the mathematics are the same for any multidimensional single-cell dataset . A complete approach would entail the study of the distribution of the individual measurement vectors . Our results demonstrate , a posteriori , that simple averages suffice for carrying out the calculations successfully . We define a “supercell of size N” as the average of the individual measurement vectors of N randomly selected cells . By repeatedly taking different random subsets of N cells from the original datasets , we build “supercell samples” and we are thus able to compute “supercell statistics” . This procedure is illustrated in Fig . 2A ( i ) – ( ii ) , where we select one shape parameter ( namely , the number of invaginations of the nuclear boundary ) and compute the probability density distributions for healthy and diseased cell lines . In Fig . 2A ( i ) , the distributions for single cells are highly overlapping , reflecting the fact that , based on individual cells , one is not able to distinguish healthy cells from diseased ones ( Dataset S1 ) . After applying the cell averaging procedure ( using N = 30 randomly selected cells to generate each “supercell” ) , we obtain distributions without any significant overlap between healthy and diseased samples , as shown by Fig . 2A ( ii ) . The supercell size N = 30 has been chosen because it represents the smallest size that provides a full separation between healthy and diseased samples , regardless of the number of parameters used ( see discussion below ) . The removal of distribution overlaps is a manifestation of the central limit theorem ( CLT ) of probability theory [25]–[27] . The CLT states that , given a set of n independent random variables associated with arbitrary probability distributions with finite mean μi and variance σi2 ( for i = 1 , 2 , … , n ) , their average is a random variable whose asymptotic cumulative distribution function approaches a normal distribution with mean μ = ∑μi/n and variance σ2 = ( ∑σi2/n ) /n . As a consequence , distributions of supercells of size N are expected to become narrower by a factor of ∼1/√N . For instance , comparing Fig . 2A ( i ) with Fig . 2A ( ii ) , we observe that the width of the latter is approximately smaller by a factor of ∼1/√30≈0 . 2 . Another consequence of the CLT is that the shape of supercell distributions becomes closer to Gaussian as N is increased . It should be pointed out that the supercell framework does not rely on a priori assumptions regarding the shape of the measurement distributions . On the contrary , it incorporates all features of the original distributions , thus naturally dealing with issues such as skewed distributions with regions that could be ambiguously attributed to outliers or to poorly resolved subpopulations . However , if the measurement distributions are distinctly multimodal due to well-defined cell subpopulations , then the ability to predict reliable phenotypes might be compromised . In such a scenario , robust phenotyping might first require the identification of different cell subpopulations followed by the application of the supercell framework separately to each of them . This procedure is discussed below in the context of distinguishing healthy individuals from patients with two non-infectious uveitides by using either all cells from peripheral blood samples , or different T cell subpopulations ( see Fig . 3 ) . After cell averaging , machine learning allows us to learn what combination of parameters best distinguishes healthy from diseased cells . In order to avoid overfitting and also to obtain a straightforward interpretation of the machine-learned parameters in terms of the original measurements , we used a support vector machine with a linear kernel , which is equivalent to the machine learning method known as the perceptron [28] , [29] . Healthy and HGPS nuclear shapes were characterized by 12 parameters including eccentricity , number of invaginations , minor/major axis length , mean and standard deviation of the curvature , and perimeter . Moreover , the concentration of lamin A/C ( measured based on the fluorescence signals of lamin A/C ) was represented through 3 additional parameters for each nucleus . However , for single cells , even with these 15 parameters , the distinction between individual cells from healthy and diseased cell lines is not learnable . Fig . 2B ( i ) shows the distance from each cell to the perceptron boundary , where positive ( negative ) distances correspond to the boundary side identified with the healthy ( diseased ) class . We observe that some cells from the healthy cell lines are classified as diseased , and vice versa . Instead , machine learning applied to the supercell samples works with 100% accuracy , as displayed in Fig . 2B ( ii ) . The questions arise , then , which and how many parameters are needed to achieve a classification of desired accuracy , and how many cells need to be averaged into a “supercell” . Fig . 2C ( i ) shows the perceptron amplitudes ( i . e . , the components of the vector normal to the boundary hyperplane ) for each of the 15 parameters . A positive sign indicates that a given parameter is higher in healthy cells relative to diseased cells , while its absolute value is a measure of its overall significance ( relative to the other parameters ) in separating healthy cells from diseased ones . Therefore , we can rank-order the 15 parameters from most to least relevant according to their decreasing amplitudes ( in absolute values ) , and learn using just the top M parameters from the rank-ordered list . While this rank ordering is independent of supercell size for large supercells , it it is very different from the rank ordering for single cells ( if the single cell measurements are strongly overlapping ) . Indeed , sizable fluctuations are observed in the single-cell and small-supercell regime ( up to supercells of size ∼10 ) followed by a stable rank-order for larger supercell sizes . The fraction of cells correctly classified by the machine learning process as a function of the supercell size is shown in Fig . 2C ( ii ) . The different curves represent different numbers of parameters ( M ) . As expected , the classification accuracy increases with both M and the supercell size . While a single cell is not sufficient for classification , a single parameter ( the number of invaginations ) is sufficient for correct classification of HGPS . Indeed , this is consistent with the standard approach to assess the disease states of HGPS based on visual analysis ( i . e . the detection of “blebs” ) and indicates that the invaginations are the most distinguishing features of blebs [30]–[32] . In our second example , we apply our technique first to the simpler problem of distinguishing healthy individuals from patients with two non-infectious uveitides , and then to the formidable challenge of distinguishing Behçet's disease from sarcoidosis . Recent work has reported progress in the ability to pinpoint molecular indicators for inflammatory immune diseases , where larger-than-normal levels of a novel subset of effector memory CD4+ T lymphocytes expressing the endothelial adhesion molecule CD146 have been observed in sarcoidosis , Behçet's , and Crohn's disease [33] . However , while patients can be diagnosed with Behçet's disease or sarcoidosis based on the concurrent observation of a number of clinical indicators , molecular signatures unique to these diseases have not been found . Our analysis of a molecular phenotype uses flow cytometry experiments , in which 14 molecular markers previously reported on human CD4+ and/or CD8+ T cells were measured for each cell; additionally , forward- ( FSC ) and side-scattering ( SSC ) measures were also taken on each cell . Thus , a total of 16 simultaneous measurements were performed on each cell from patients' peripheral blood , with about one million cells measured per patient . From a cohort of 22 patients , 7 were diagnosed with sarcoidosis , 6 with Behçet's disease , 1 with retinal vasculitis , while the remaining 8 were healthy controls . We start with large supercells to assess whether molecular phenotyping is possible at all to distinguish sarcoidosis and Behçet's disease . We represent each patient sample with 100 supercells , where each supercell was obtained from averages over 500 randomly chosen cells . We carry out separate analyses for the distinction between healthy and diseased patients ( Fig . 3 ( a ) – ( c ) ) , and for the separation between the two diseases sarcoidosis and Behçet's ( Fig . 3 ( d ) – ( f ) ) . Furthermore , we perform separate analyses for all cells ( Dataset S2 ) , for CD4+ T cells ( that can be isolated using standard gating procedures based on the sequence viability−/CD3+/CD4+/CD8− ) ( Dataset S3 ) and for CD8+ T cells ( similarly identified according to viability−/CD3+/CD8+/CD4− ) ( Dataset S4 ) . Because we have a larger number of patients than we did for HGPS , we can directly assess the predictive power of our approach to correctly diagnose a new patient . We tested the predictive power of our learnt patterns using a standard data-resampling method , namely the so-called jackknife procedure: leaving out one patient at a time , one learns with the remaining data and makes a prediction on the test patient [34] . In that way , one can determine the percentage of correct and failed predictions . Since each patient is represented by a cloud of 100 supercells , it may happen that the perceptron boundary cuts across the test patient's supercell cloud . We set a threshold of 95% in order to make a prediction: e . g . if the supercell cloud is more than 95% consistent with sarcoidosis , we classify the patient as having sarcoidosis . If the supercell cloud falls on the boundary between diagnoses ( i . e . with less that 95% of the supercells on either side of the perceptron boundary ) , we leave the test patient unclassified . Naturally , setting the prediction threshold to lower values leads to less unclassified patients , but tends to increase the number of failed predictions; in contrast , increasing the threshold to higher values leads to a more conservative approach , where the number of failed predictions is smaller at the expense of a larger number of unclassified patients . By changing the prediction threshold values over the range between 80% and 100% , the observed variations of the predicted outcome were below 10% of the cohort; the method is thus largely insensitive to the choice of the threshold parameter . By learning using all available measures , we are able to rank-order the importance of the measures based on the perceptron amplitudes . The ten most important measures and corresponding amplitudes are listed in Fig . 3 ( a ) – ( f ) . The percentage of patients correctly predicted ( green ) , unclassified ( blue ) , and incorrectly predicted ( red ) are shown as a function of the number of rank-ordered measures used . The outcomes depend strongly on the type of cells used: for the “healthy vs diseased” case , no incorrect predictions are made using all cells and just the top two measures , namely viability and CD197 ( Fig . 3 ( a ) ) . The predictions are even stronger if using only CD4+ T cells , since the top marker ( CD27 ) is sufficient by itself to correctly classify all healthy patients ( with high frequency of CD4+CD27+ T cells in their peripheral blood ) and all diseased patients ( with low frequency of CD4+CD27+ T cells in their peripheral blood ) in the cohort ( Fig . 3 ( b ) ) . In contrast , failed predictions are seen for the case of CD8+ T cells , irrespective of the number of measures used ( Fig . 3 ( c ) ) . Previous reports have suggested that CD4+CD27+ T cells represent the majority of natural regulatory T cells in human peripheral blood [35] . Thus , our results indicate that patients with either Behcet's disease or Sarcoidosis have low frequency of peripheral natural regulatory T cells and , therefore , potentially compromised immunoregulatory functions during inflammatory responses . In order to separate Behçet's disease and sarcoidosis , predictions based on all cells are very poor ( Fig . 3 ( d ) ) , better for CD4+ T cells ( Fig . 3 ( e ) ) and best for CD8+ T cells ( Fig . 3 ( f ) ) , for which no failed predictions are made when five or more measures are used . This result indicates that the top measures listed in Fig . 3 ( f ) may be used as molecular phenotypes that distinguish the two diseases . This is , to the best of our knowledge , the first report pointing out that CD8+ T cells can be used to distinguish two systemic inflammatory diseases . Moreover , it is interesting to note that , in distinguishing between patients with ocular inflammation and controls without it , the CD4 marker was an important feature , while for distinguishing between the ocular manifestations of two systemic disorders , the CD8 cell marker was superior . Given our success in demonstrating the power of molecular phenotyping to distinguish the diseases , we turn now to the analysis of the balance between the number of cells we need to average , and the number of molecular markers we need to measure . For the “healthy vs diseased” case using CD4+ T cells , the percentage of correctly classified supercells is shown in Fig . 3 ( g ) as a function of the supercell size and the number of measures used . Note that for single cells , the classification performance is very poor even using many measures , but averaging over more than ten cells is sufficient for reliable classification if a large number of measures is used . In contrast , just one measured marker is sufficient provided that we average over 100+ cells . This fact is underscored in Fig . 3 ( h ) , where the intensity distribution for supercells of size 500 are shown separately for the healthy and the diseased patients , using just the top marker ( CD27 ) . The dashed line indicates the marker intensity threshold that allows a complete separation of the two classes of supercells . The “sarcoidosis vs Behçet's” classification is further studied in Fig . 3 ( i ) for CD8+ T cells , where the percentage of correctly classified supercells is shown as a function of the supercell size and the number of measures used . We find that slightly less than 100 cells are sufficient for reliable classification , as long as the top five markers are measured . Increasing the number of markers or averaging over more cells does not strongly change the reliability of the classification . Finally , the ability to classify Behçet's disease vs sarcoidosis when using the top 5 markers is visualized in a new way in Fig . 3 ( j ) . The visualization is derived from the identification of patterns in two-dimensional parameter space ( Fig . 1A ( i ) ) , which has proven to be a tremendously successful tool for the analysis of low-dimensional data in flow cytometry . The combined approach of cell averaging into supercells , followed by machine learning , allows us to find the correct linear combinations of markers needed to fully separate the two diseases ( Fig . 3 ( j ) ) . In geometrical terms , we learned that only 5 dimensions ( out of the original 16 ) are needed; moreover , we determined the preferred direction that maximizes the gradient between the “sarcoidosis class” and the “Behçet's disease class” . This optimal class separation was achieved by means of an unbiased , mathematically robust method: no additional biological information was needed to proceed from Fig . 1A ( i ) to Fig . 3 ( j ) .
We present a simple approach to quantify disease phenotypes based on single cell measurements with multiple parameters measured on each cell . For our study of autoimmune diseases , we measure 16 parameters for millions of cells with flow cytometry , and use this information to find a molecular phenotype of Behçet's disease . We also measure 15 parameters from hundreds of fluorescence images obtained via microscopy , and use this information to automate classification of HGPS . Our data span many more dimensions than the traditional two parameters used for visually-aided cell classification ( Fig . 1A ) . We use machine learning , which allows for a reproducible , objective , and automated approach to find the optimal boundary between two high- dimensional classes of data points . The question we tackle is straightforward: do we obtain more information about a disease by the analysis of more cells , or by measuring more parameters on each cell ? The key to our novel approach is to introduce variable size cell groups ( “supercells” ) , with the group size as an explicit parameter that we vary systematically . This reveals the number of cells that need to be grouped in order to obtain a robust disease classification . We also determine to what degree adding parameters reduces the number of cells needed to determine a phenotype . Our approach to separate cell groups relies on a machine learning classification method . It is tailored specifically to determine the most useful combination of parameters to distinguish among all cells or cell groups rather than finding the optimal low-dimensional representation , as in singular value decomposition or principal component analysis . This procedure is schematically illustrated in Fig . 4 , where synthetic 2D datasets were generated to represent patient samples classified in two categories: 4 samples correspond to “Class A” , while the remaining 3 samples are labeled “Class B” . At the single-cell level , the data are non-separable due to cell heterogeneity ( Fig . 4 ( a ) ) . A machine learning classifier such as support vector machines with a linear kernel ( Fig . 4 ( b ) ) can be implemented in order to find the optimal decision boundary between the two classes; however , this method requires the data to be separable . More sophisticated variants , such as soft margin classifiers and nonlinear classifiers can be designed to learn from non-separable data , even in strongly overlapping cases such as those usually encountered in single-cell datasets ( see e . g . Fig . 1A ( i ) ) . However , our analysis shows that the optimal “boundary” inferred from overlapping distributions is distinct from the boundary obtained from supercells which actually has predictive power for phenotyping: the weight of parameters is very different for single cells and supercells whose distribution is well separated . In order to avoid overlapping patient samples , one could characterize each of them by the moments of the cell multivariate distributions , the simplest example being the sample means ( Fig . 4 ( c ) ) . This approach , however , lacks robustness: the decision boundaries are very sensitive to nearby datapoints , in particular to the support vectors that determine the classification hyperplanes , thus leading to failed predictions . Supercell distributions are built by averaging over groups of single cells . By applying machine learning on supercell samples , a robust class separation is achieved ( Fig . 4 ( d ) ) . In HGPS , our approach confirms the current practice that the number of invaginations ( or mean negative curvature ) is the most valuable nuclear metrics for phenotyping the disease using nuclear images . Importantly , we find that when analyzing 30 cells or more , a robust phenotype can be obtained simply based on the invaginations of each cell , and a more in depth analysis of additional nuclear shape metrics does not significantly reduce the number of cells needed . Our findings provide a principle guideline of the minimal cell numbers used in future disease assessments and high-throughput drug screenings of age-related diseases , in which abnormal nuclear shape is considered a hallmark phenotype . This information is of extreme importance in a rare disease like HGPS with very limited availability of patient samples . In our second example , we apply our technique to distinguish healthy individuals from patients with two non-infectious uveitides , and among those patients we distinguish between Behçet's disease and sarcoidosis . In order to distinguish healthy from disease phenotypes , we found that within the CD4+ T cell subpopulation , just one marker was enough ( Fig . 3 ( b ) ) . Indeed , CD27 appears consistently overexpressed in healthy samples ( Fig . 3 ( h ) ) . The ability to predict healthy and diseased phenotypes based on CD4+ T ( super ) cells is resilient under the removal of the top markers: even by removing the top 7 markers from the list , we are still able to classify patients as healthy or diseased with no failures . In contrast , CD8+ T cells do not have a clear distinction between healthy and diseased conditions , even using all markers available from the flow cytometry experiment ( Fig . 3 ( c ) ) . However , by focusing specifically on sarcoidosis and Behçet's disease , we demonstrate a robust means of predicting a patient's diagnosis based on 5 optimally chosen markers using CD8+ T ( super ) cells ( Fig . 3 ( f ) ) . If the top marker ( IL22 ) is removed from the list , incorrect predictions are observed even using all remaining markers; therefore , phenotyping sarcoidosis vs Behçet's is inherently high-dimensional ( since it requires at least 5 markers to be accurate ) and also very specific to those markers . An important evaluation for the future will be to evaluate the efficacy of these markers in patients with these two systemic disorders who do not have ocular complications of their disease , i . e . whether these findings are specific to the ocular disorder , or a reflection of the systemic disorder itself . By using a precise linear combination of IL22 , CD3 , viability , CD8 and CD62L , we are able to separate the two diseases successfully based on molecular markers ( Fig . 3 ( j ) ) . Averages of hundreds of cells are required for this phenotyping , and increasing the number of measured parameters does not reduce the number of cells required . The molecular markers used have been reported to be important players in autoimmune disorders . Yang et al . [36] reported an increased number of Th22 cells and increased serum IL-22 levels in patients with lupus skin disease , but a decrease in patients with lupus nephritis . CD62L has been reported to be associated with CD4+CD25brightFOXP3+ cells in bullous pemphigoid patients [37] . Finally , expanded clones of CD8+ T lymphocytes are present in the lesions of multiple sclerosis [38] . Based on the observations from the analyses presented here , our evaluation of CD8+ T cells has permitted us to see CD8-subset differences in this cell type in patients diagnosed with different uveitides . Our ability to study the tradeoff between measuring more parameters or analyzing more cells , as shown in Figs . 3 ( g ) and 3 ( i ) , has far-reaching consequences for a number of emerging technologies that allow for multi-parameter single-cell measurements . For more challenging problems than those considered here , it may become necessary to study the distributions of the measurement vectors of individual cells rather than its principal surrogate of the first moment , and extend the machine learning algorithms to well-chosen non-linear kernels . High-throughput automated microscopy , where thousands of cells are imaged automatically , is quickly becoming the norm , calling for reliable approaches to classify observations and quantify phenotypes . Similarly , while simultaneous ( multicolor ) measurement of 16 parameters is the current state-of-the-art for flow cytometry , a next generation of high-throughput single-cell analysis tools is emerging that will allow the measurement of more than 50 parameters at comparable high throughput by means of mass cytometry [3] , [4] . It is now also becoming possible to analyze gene sequences or gene expression levels for individual cells , although the cost of these expensive technologies severely limits the sample size to much fewer cells than flow cytometry [5] , [6] . Optimizing the tradeoff between measuring more cells or more parameters , as we demonstrate here , should allow us to take full advantage of these powerful and promising next-generation single-cell technologies .
This investigation was conducted according to the principles expressed in the Declaration of Helsinki and was approved by institutional review boards at National Eye Institute , National Institutes of Health . The written informed consent was provided by all patients . For the study of Hutchinson-Gilford progeria syndrome , cultured fibroblasts from two patients ( HGADFN164-p15 and HGADFN167-p15 ) and two healthy individuals ( HGADFN090-p15 and HGADFN168-p15 ) were used . The cells were fed with fresh MEM medium containing 15% FBS and grown at 37°C . In order to visualize the nuclei , we performed immunofluorescence staining of the nuclear membrane with a mouse monoclonal antibody raised against lamin A/C . ( MAB3211 ) . This antibody has been well characterized in HGPS cells and has also been used in studies of other laminopathies . Fluorescence images of about 600 nuclei per cell line were taken with a Zeiss fluorescence microscope at 400× magnification , as shown in the examples from Figure 1B . Following the procedure from Driscoll et al ( 15 ) , a custom-written MATLAB program was used to extract nuclear shapes and their properties , such as the number of invaginations , the mean curvature , the standard deviation of the curvature , etc . In addition to 12 shape measurements , we obtained 3 measurements of the intensity of immunofluorescence from lamin A/C associated with each nucleus ( the full list of measurements is provided in Figure 2C ( i ) . For the study of non-infectious uveitides , peripheral blood samples were obtained from a cohort of 22 patients , out of which 7 were diagnosed with sarcoidosis , 6 with Behçet's disease , 1 with retinal vasculitis , while the remaining 8 were healthy controls . 3 different marker panels were studied on each sample , each consisting of 2 scattering measurements ( FSC and SSC ) plus 14 or 15 cell surface fluorochromes . Some common markers ( such as CD3 , CD4 , CD8 , CD27 , CD45 , and viability ) were used on all 3 panels and were checked for consistency . Separate analyses have been performed on each set of markers in order to find the best prediction accuracy . Two marker panels did not lead to accurate sarcoidosis vs Behçet's disease phenotypes; the third one , which led to an accurate phenotype and has been discussed throughout , consisted of FSC , SSC , IL23R , CD196 , CD4 , viability , CD8 , CD27 , CD45 , IL17A , CD197 , CD3 , IL22 , CD62L , CD161 , and TNFA . Pre-processed datasets are provided as Supporting Information . Multicolor flow cytometry raw datasets are available at the Dryad Repository: http://dx . doi . org/10 . 5061/dryad . v6st3 . Data analysis was performed using custom-written programs in R and Perl . | The behavior of organisms is based on the concerted action occurring on an astonishing range of scales from the molecular to the organismal level . Molecular properties control the function of a cell , while cell ensembles form tissues and organs , which work together as an organism . In order to understand and characterize the molecular nature of the emergent properties of a cell , it is essential that multiple components of the cell are measured simultaneously in the same cell . Similarly , multiple cells must be measured in order to understand health and disease in the organism . In this work , we develop an approach that is able to determine how many cells , how many measurements per cell , and which measurements are needed to reliably diagnose disease . We apply this method to two different problems: the diagnosis of a premature aging disorder using images of cell nuclei , and the distinction between two similar autoimmune eye diseases using stained cells from patients' blood samples . Our findings shed new light on the role of specific kinds of immune system cells in systemic inflammatory diseases and may lead to improved diagnosis and treatment . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] | [] | 2013 | From Cellular Characteristics to Disease Diagnosis: Uncovering Phenotypes with Supercells |
The human double-homeodomain retrogene DUX4 is expressed in the testis and epigenetically repressed in somatic tissues . Facioscapulohumeral muscular dystrophy ( FSHD ) is caused by mutations that decrease the epigenetic repression of DUX4 in somatic tissues and result in mis-expression of this transcription factor in skeletal muscle . DUX4 binds sites in the human genome that contain a double-homeobox sequence motif , including sites in unique regions of the genome as well as many sites in repetitive elements . Using ChIP-seq and RNA-seq on myoblasts transduced with DUX4 we show that DUX4 binds and activates transcription of mammalian apparent LTR-retrotransposons ( MaLRs ) , endogenous retrovirus ( ERVL and ERVK ) elements , and pericentromeric satellite HSATII sequences . Some DUX4-activated MaLR and ERV elements create novel promoters for genes , long non-coding RNAs , and antisense transcripts . Many of these novel transcripts are expressed in FSHD muscle cells but not control cells , and thus might contribute to FSHD pathology . For example , HEY1 , a repressor of myogenesis , is activated by DUX4 through a MaLR promoter . DUX4-bound motifs , including those in repetitive elements , show evolutionary conservation and some repeat-initiated transcripts are expressed in healthy testis , the normal expression site of DUX4 , but more rarely in other somatic tissues . Testis expression patterns are known to have evolved rapidly in mammals , but the mechanisms behind this rapid change have not yet been identified: our results suggest that mobilization of MaLR and ERV elements during mammalian evolution altered germline gene expression patterns through transcriptional activation by DUX4 . Our findings demonstrate a role for DUX4 and repetitive elements in mammalian germline evolution and in FSHD muscular dystrophy .
The transcription factor DUX4 is a member of a small family of double-homeodomain genes that also includes paralogs DUXA , DUXB , DUXBL , DUXC and murine Dux [1]–[3] . The primate-specific DUX4 gene likely arose from retrotransposition of the DUXC mRNA , with subsequent deletion of the DUXC gene from the primate genome [3] . DUX4 exists in the primate genome as part of a ∼3 . 3 kb repeat unit called D4Z4 , found in macrosatellite arrays in the subtelomeric regions of human chromosomes 4 and 10 . Poorly studied arrays are also found in other genomic regions and appear to contain interrupted versions of the DUX4 ORF [4] , [5] . The number of D4Z4 units in each array varies between human individuals [6] , and array size and number of genomic loci vary between different primate species [7] . DUX4 is expressed in germ-line cells of the testis and is epigenetically repressed in somatic tissues [8]–[11] , likely in part through repeat-mediated repression [12] . Deletions of the D4Z4 array to fewer than 10 repeat units or mutations in SMCHD1 , a gene necessary for repeat-mediated epigenetic repression , result in decreased epigenetic repression of DUX4 in skeletal muscle , causing a human muscle disease , facioscapulohumeral muscular dystrophy ( FSHD; OMIM #158900 , #158901 ) [6] , [9] , [13]–[15] . The decreased epigenetic repression results in occasional bursts of DUX4 expression in a subset of nuclei in FSHD muscle cells [9] , [16] , which appears to cause death of expressing muscle cells [17]–[19] . Our prior work showed , using microarrays and ChIP-seq , that mis-expression of DUX4 in human skeletal muscle activates the expression of many genes normally expressed in the germline and that DUX4 binds a double-homeobox motif at ∼60 , 000 sites in the mappable genome [10] . Although many DUX4 binding sites overlap non-repetitive regulatory elements of its activated genes , more than half of the ChIP-seq peaks overlap repeat elements of the MaLR class [10] . MaLRs , or mammalian apparent LTR-retrotransposons , are distantly related to endogenous retroviruses ( ERVs ) , and comprise ∼3 . 7% of the human genome with ∼350 , 000 copies of various subclasses of MaLR elements [20]–[23] . Like ERVs , MaLRs are selfish elements that spread in the genome by a copy-paste mechanism called retrotransposition ( transcription , reverse transcription and reintegration ) . ERVs and MaLRs have a pair of long-terminal repeats ( LTRs ) that act as strong promoters and that flank one or more open reading frames ( ORFs ) . Post-insertion deletions mediated by homology between the two LTRs of an ERV or MaLR element often leave single ( or “solo” ) LTR elements in the genome . The internal ORFs of conventional ERVs encode the gag , pro , pol , and sometimes env proteins needed for replication and re-integration , whereas the single MaLR ORF encodes a protein with an ∼90-aa stretch of homology to the ERVL gag protein , suggesting that MaLRs derived from ERVL-like retrotransposons [24] . The homology to the ERVL gag protein and the absence of the other proteins necessary for ERV replication suggest that MaLRs might rely on concomitantly expressed ERVs to provide these other proteins . Most human ERV and MaLR elements inserted into the primate genome before the divergence of Old and New World monkeys and are generally thought to be no longer capable of retrotransposition ( unlike in rodents , where ERVs and MaLRs are still active ) . However , reports of very rare polymorphic ERV and MaLR insertions in humans suggest that occasional transposition events still occur in the human population [25] , [26] . Retrotransposition provides a way to spread regulatory sequences to large numbers of new genomic locations in short evolutionary time . Some families of repetitive elements appear to have been involved in large-scale rewiring of transcriptional networks [27] , [28] and a number of cellular transcription factors have been shown to bind repetitive elements [29]–[31] . Repeats , especially LTRs , can be co-opted ( or “exapted” ) to act as promoters for genes of the host genome [32] , [33] . For example , when placentation evolved during early mammalian divergence , a large number of genes acquired expression in endometrial cells via upstream insertion of the eutherian-specific MER20 transposon [34] . In another example , binding of the OCT4 and NANOG transcription factors to various primate-specific repeats explains many of the differences observed between the transcriptional networks of human and mouse ES cells [31] . To date , no such repeat-mediated rewiring of the male germ cell transcriptional network has been reported , yet testis expression patterns are known have evolved rapidly in mammals [35] . To determine whether DUX4 binding to repetitive elements can affect transcriptional networks , we analyzed ChIP-seq and RNA-seq datasets in skeletal muscle cells that ectopically express DUX4 , as well as RNA-seq data from FSHD patient and control muscle cells . We find that repetitive elements are bound and activated by DUX4 , including primate-specific MaLR and ERV LTRs , and that some full-length repeat elements are transcribed in response to DUX4 . In addition , some DUX4-bound LTRs form novel first exons of annotated genes , long-noncoding ( lnc ) RNAs , and antisense RNAs . Together , our findings demonstrate that DUX4's transcriptional network is mediated in part through binding to repetitive elements , including some primate-specific retroelements that could contribute to lineage-specific patterns of gene expression [35] .
In previously published work , we observed that many of the ∼60 , 000 DUX4 binding sites identified by ChIP-seq following expression of DUX4 in human myoblasts overlap MaLR LTR elements [10] . Using a more recent version of the human genome assembly ( hg19 ) and including the X and Y chromosomes , we identified 63 , 795 DUX4 binding-sites ( Table S1 ) . Overall , ∼2/3 of sites are in a repetitive element , more than expected given that ∼45% of the human genome is recognizable as repeats ( Figure 1A ) [21] . Furthermore , ∼1/3 of DUX4 binding sites are in a MaLR element , ∼10-fold more than would be expected if binding sites were dispersed randomly throughout the genome ( Figure 1B ) . Because DUX4 binds an AT-rich sequence motif , its binding sites might show repeat enrichment based simply on GC content; however , when we randomly sampled genomic locations with similar AT-content to the DUX4 binding site ( see Methods ) , we find that only ∼49% overlapped repeats of any class , and only ∼3% overlapped MaLR elements . In order to determine which subclasses of MaLRs are responsible for the enrichment , and to explore which other repetitive elements might also bind DUX4 , we systematically analyzed ChIP-seq data using two complementary approaches . An analysis of repetitive element content among DUX4 ChIP-seq peaks shows that enrichment is heavily biased towards LTR elements ( Figure 1 , Table 1 , S2 and S3 ) , particularly those of the MaLR class: of 32 enriched repeat types ( using arbitrary thresholds of ≥2-fold enrichment and ≥100 DUX4-bound instances ) , 30 are in the LTR class , and of those , 23/30 are MaLRs . Many subtypes of MaLR-LTRs contribute to the MaLR family-wide enrichment , including the LTRs of THE1 elements ( A , B , C and D subfamilies ) that were active early in the primate lineage [20] and MLT1 subtypes that were active before mammalian radiation . Outside of the MaLR family , other LTR subtypes also show enrichment , including primate-specific ERVL-LTRs ( MLT2A1 and MLT2A2 ) and hominoid-specific ERVK-LTRs ( MER11B and MER11C ) ( Table 1 ) . Some non-LTR repeats are also enriched among DUX4 binding sites , including the HSAT5 and HSATII satellite repeats and some simple repeats like the ( CAAT ) n tetranucleotide repeat , although with fewer mappable DUX4 binding sites ( <100 DUX4-bound instances ) . In almost all cases , the consensus sequences for enriched repeats contain at least one DUX binding motif ( Figure S1 ) . Analysis of ENCODE ChIP-seq data for 161 other regulatory factors ( see Methods ) shows that these DUX4-bound regions are not generally bound by other transcription factors ( the factor with most overlapping peaks , Runx3 , only bound 0 . 9% of DUX4-bound repeat regions ) . It remains challenging to accurately map sequence reads to repetitive elements with high sequence identity [36] , and the peak-based ChIP-seq analysis we described above is relatively blind to recently mobilized repeats because it only uses sequence reads that map uniquely in the human genome . Therefore we also used a complementary approach , adapting a previously published method [37] that estimates enrichment regardless of whether individual ChIP-seq reads map uniquely in the genome ( Figure 1D , Methods ) . The method aggregates read counts over all genomic copies of each repeat class rather than trying to map reads uniquely to individual repeat instances . Read counts for a test sample are then compared with counts for a control sample to determine enrichment . Although this read-based method can look at recently mobilized repeats , it gives a “dampened” measure of repeat enrichment due to background reads in ChIP-seq samples and is therefore less sensitive to modest levels of enrichment ( see Methods ) . This read-based method identified a similar set of repetitive elements as the peak-based method ( Figure 1E ) , and , in addition , revealed enrichment of a small number of repeat types , such as the SVA_F subfamily , that were not detected using the peak-based method ( Figure 1E , Tables 1 , S2 and S3 ) . SVAs are composite retroelements that include segments derived from SINE , VNTR and Alu repeats , and have been very recently active in the human , chimpanzee and gorilla genomes [38] . To examine whether DUX4 binding sites are functionally important , especially those in repetitive elements , we determined whether they are evolutionarily conserved . We find that DUX4-bound motifs have tolerated fewer sequence changes than flanking sequences during the evolution of placental mammals ( Figure 2A ) , primates ( Figure 2B ) , and humans ( Figure 2C ) . These findings hold true even if we consider only motifs in repetitive sequences ( Figure 2 , gray datapoints ) , demonstrating that at least a subset of DUX4 binding sites in repetitive elements are conserved . In order to determine whether DUX4 binding to repetitive elements results in their transcriptional activation , we generated RNA-seq data ( 100 nt single-end reads ) from DUX4-transduced and control myoblasts . In a conservative analysis , we identified 738 DUX4-activated transcripts within 1 kb on either side of a DUX4 ChIP-seq peak and six DUX4-repressed transcripts ( Figure 3A , ≥2-fold change , FDR-adjusted p-value<0 . 1 ) . These 738 bound and activated regions comprise 1 . 2% of the 63 , 795 DUX4-bound sites . Peaks with multiple DUX4-binding sites are more likely to initiate transcripts than those with a single motif ( Figure 3B; 2 . 5% of peaks with more than one site were associated with lenti-DUX4-transcriptional induction , compared to only ∼0 . 8% peaks with 1 motif; p<10−15 , chi-squared test ) . Peaks with more than one motif also show greater ChIP-seq peak height , a proxy measurement for DUX4 occupancy ( Figure 3C ) . DUX4-bound repetitive elements are just as likely to initiate a transcript as DUX4 binding sites in unique sequence ( ∼1 . 2% of both classes show activation ) , with 454/738 ( ∼62% ) of DUX4-initiated transcripts arising in or near a repetitive element . The DUX4-bound repeat types most likely to be transcribed are HSATII elements ( 23% of bound HSATII repeats show activation ) and MLT2A1 ERVL-LTRs ( 5 . 4% of bound MLT2A1 LTRs are activated ) ( Tables 1 , S2 and S3 ) . Some of this effect might be explained by the number of DUX4-binding sites within each peak: 65% of MLT2A1 elements contain more than one good DUX4-motif , compared to only 9% of a class of elements that are less frequently activated , the THE1B-MaLRs ( 1 . 0% bound THE1Bs are activated ) . To explore the biological significance of DUX4's ability to bind and activate various repetitive elements , we used RNA-seq data to examine the types of transcripts that result from repeat activation . First , we asked whether full-length repetitive elements are activated . 100 , 864 regions of the human genome assembly are annotated as internal regions of ERV or MaLR elements , or fragments thereof , and many of these repetitive elements are old enough to have acquired sequence changes that allow unique mapping of short sequence reads . RNA-seq data from DUX4-transduced myoblasts shows that 184 of these regions are activated in the presence of DUX4 ( ≥2-fold , FDR-adjusted p-value≤0 . 1 ) , of which 120 are MaLRs and the rest a mix of other ERV classes . These activated internal MaLR/ERV regions tend to be flanked on one or both sides by a DUX4-bound LTR , where transcription seems to initiate . Repeats whose internal regions are activated tend to be younger than repeats that do not show obvious activation , considering only repeats flanked by at least one DUX4-bound LTR ( Figure S2 ) . It is tempting to suggest that this finding indicates that repeats have been evolving towards better DUX4 response , but might more likely reflect the fact that , because less evolutionary time has elapsed , younger repeats have retained sequence elements needed to produce stable transcripts , like TATA boxes and polyadenylation signals . Using RT-PCR , we were able to verify DUX4-mediated activation of a THE1C-MaLR element and one ERVL element , but not of a second ERVL element ( Table 2 , Figures S3 , S4 and S5 ) . Our inability to confirm activation of one of the ERVLs could be explained if mismapping of RNA-seq reads among highly related ERVL elements misled our choice of an individual element from which to design primers for this assay . Of these 184 transcriptionally activated ERV or MaLR internal elements , only one contains an ORF exceeding 300 amino acids in length . This full-length THE1D-MaLR internal region aligns to the THE1 consensus sequence without stops or frameshifts , and encodes a 464 amino acid ORF . Because the function of MaLR-encoded proteins is unknown , it is difficult to interpret this finding , but we note that this chromosome 7 THE1 is the only THE1 element in the genome that encodes an uninterrupted ORF and thus is a candidate “active” MaLR element . To examine the possibility that this ORF might be preserved as a domesticated protein , we collected its sequence from other primate genomes . We found that although the ORF is also maintained in chimpanzee and gorilla , it is interrupted by stop codons and/or frameshifts in orangutan , macaque , baboon , and marmoset . The lack of conservation among primates suggests that the ORF does not confer selective advantage . 17 other internal regions contain one or more ORFs exceeding 200 amino acids ( arbitrary threshold ) and encode fragments of various THE1 internal ORFs as well as fragments of some ERV gag and pol ORFs , although none encode full-length proteins . Repetitive elements , especially LTRs , can be co-opted as alternative promoters for mammalian genes [32] , [33] and can rewire transcriptional networks during evolution [27] , [28] , [31] . Therefore , acquisition of the DUX4 retrogene and the spread of repetitive elements around the genome had the potential to alter germ cell promoter usage and the germ cell transcriptional network during primate evolution . Analysis of spliced RNA-seq reads that join a DUX4-bound region with a gene sequence identified 238 previously unannotated DUX4-activated promoters for human genes ( Table S4 ) with 144 of those promoters in repetitive elements . Neither GO nor GREAT analyses [39] of these genes and regions revealed striking enrichment of particular functional classes , although we note that germ cell genes ( especially primate-specific genes ) are likely very poorly annotated . We selected three genes ( HEY1 , PPCS and NT5C1B ) for validation ( Figures 4 , S6 , S7 and S8 , Table 2 ) . HEY1 ( hairy/enhancer-of-split related with YRPW motif 1 ) inhibits myogenesis by repressing myogenin and Mef2C [40] , and its ∼13-fold activation by DUX4 in muscle cells could inhibit muscle differentiation in FSHD . RNA-seq data from DUX4-expressing myoblasts suggests the existence of transcripts that initiate ∼40 kb upstream of HEY1's first annotated exon in a THE1B-MaLR retrotransposon and splice via two additional exons to the second exon of HEY1 ( Figure 4A ) . These chimeric transcripts encode an ORF with an in-frame start codon in exon 2 that lacks the first 38 amino acids of the full-length HEY1 protein . We verified this THE1B-HEY1 fusion transcript by RT-PCR , 5′-RACE and Sanger sequencing and showed that its presence in myoblasts is DUX4-dependent ( Table 2 , Dataset S1 ) . This THE1B element is present at the syntenic location in Old and New World monkeys but not in more distant genomes , demonstrating that it inserted in our genome ∼40–75 million years ago [41] , long after the origin of the HEY1 gene . A DUX4-bound MLT1B-MaLR element initiates DUX4-dependent transcripts ∼6 kb upstream of the phosphopantothenoylcysteine synthetase gene ( PPCS ) that splice to exon 2 of PPCS , as suggested by RNA-seq data and verified by RT-PCR , 5′-RACE and Sanger sequencing ( Figure 4B , Table 2 , Dataset S1 ) . The predicted translation start codon of the chimeric transcript in exon 2 is also used in the shorter of the two annotated PPCS isoforms ( RefSeq NP_001070915 ) . This MLT1B element is found in the syntenic location in diverse mammalian genomes including those of primates , rodents , carnivores and bats , indicating that it inserted in our genome at least 90 million years ago . A DUX4-bound MLT1E1A-MaLR element initiates transcripts ∼6 kb upstream of the NT5C1B ( 5′-nucleotidase , cytosolic IB ) gene and uses either of two donor sites to splice to exon 2 of NT5C1B ( Figure 4C ) . This fusion transcript encodes either an ORF lacking the first 14 amino acids of NT5C1B , or a chimeric ORF with the first 10 amino acids of NT5C1B replaced by 17 amino acids encoded in the MLTE1A sequence . Our RNA-seq data shows that the gene is transcribed at low levels in control myoblasts , but is induced ∼300-fold in the presence of DUX4; RT-PCR , 5′-RACE and Sanger sequencing confirm the novel MLT1E1A-NT5C1B fusion transcript ( Table 2 , Dataset S1 ) . This MLT1E1A element is found at the syntenic location in diverse placental mammalian genomes , so must have inserted into the ancestral genome at least 98 million years ago . In addition to creating novel first exons for protein-coding genes , our RNA-seq data revealed that DUX4-bound repeats can also create promoters for long non-coding RNAs ( lncRNAs ) . Comparison of DUX4 binding and activated transcripts to a recently published dataset of lncRNAs [42] shows that 18 DUX4-bound sites initiate transcripts for lncRNAs , of which 13 are in repetitive elements ( Table S5 ) . We used RT-PCR and Sanger sequencing to verify two of these activated lncRNAs; one initiates in an MLT1C-MaLR element shared among many mammals , and one in a primate-specific THE1C-MaLR element ( Figures 5A , 5B , S9 , S10 ) . lncRNA catalogs are incomplete and more instances of DUX4-initiated lncRNAs are likely to exist . Two very recent reports [43] , [44] show that repetitive elements are enriched at the transcription start sites of lncRNAs . Only 56 of the 2045 repeat-initiated lncRNAs ( 2 . 7% ) described in one of those reports [43] start in DUX4-bound repeats . We note that the catalogs of lncRNAs used in these two recent studies include transcripts expressed in a diverse set of tissues; if suitable data are available in future , it will be interesting to determine whether DUX4-bound repeats comprise a greater proportion of lncRNA transcription start sites in germ cells than in other tissues . DUX4-bound sites also initiate transcripts antisense to annotated genes . For example , a transcript that initiates in an MLT1D-MaLR element overlaps the first exon of the DDX10 gene in the antisense orientation ( Figures 5C , S11 ) , and is confirmed by RT-PCR and Sanger sequencing . We are not aware of any genome-wide catalog of antisense transcripts and thus did not perform a systematic analysis of these RNAs . The analysis above relies on RNA-seq reads that map to fewer than 20 genomic locations and is blind to highly repeated sequences . Therefore , we also used an alternative read-based method to calculate repeat enrichment among DUX4-activated transcripts ( see above , and Methods ) . We identified many of the same repeat classes already highlighted by our analyses of uniquely mappable reads , indicating that many DUX4-bound and activated repeats have diverged enough that standard methods are effective ( Figure 6A , Tables S2 and S3 ) . However , as with ChIP-seq data , the read-based analysis uncovered additional classes of activated repeats that were not obviously enriched when we used uniquely mapped reads , including a number of LTR elements , mostly of the ERV1 and ERVK families ( e . g . MER52D , LTR12D , MER50B ) ( Table 1 , S2 and S3 ) . Most notably , however , copies of the pericentromeric satellite repeat HSATII are massively activated in the presence of DUX4 . Combining the two DUX4-transduced myoblast RNA-seq datasets , HSATIIs are activated ∼860-fold with ∼6700 reads per million in DUX4-expressing cells compared to only ∼8 reads per million in control samples ( Figure 6B ) , a baseline consistent with low HSATII expression ( 0 . 2–17 HSATII sequences per million reads; median 0 . 9 reads per million ) in a panel of sixteen normal tissues sequenced by Illumina ( the “Body Map 2 . 0” dataset , GEO accession GSE30611 ) . We aligned HSATII RNA-seq reads to the HSATII consensus sequence , finding that multiple variant sequences ( and therefore multiple repeat units ) are transcribed ( Figure S12 ) . Our ChIP-seq data also demonstrated that HSATIIs are bound by DUX4 , with ∼1 . 9-fold enrichment of HSATII sequences among individual reads and 30 DUX4 peaks in mappable HSATII regions ( ∼5-fold enrichment ) . Furthermore , each HSATII ChIP-seq peak appears to derive from multiple tandemly-arranged DUX4 binding sites . The 30 HSATII peaks span bigger genomic regions ( median peak width 1 . 2 kb ) than other ChIP-seq peaks ( median width 0 . 4 kb ) and contain multiple matches to DUX4's consensus motif - the 299 annotated HSATII regions in the human genome assembly contain a total of 820 DUX4 motifs . We show above that a large number of repeat-initiated transcripts are induced in myoblasts over-expressing DUX4 . To determine whether these transcripts are expressed in normal germ cell biology and in FSHD muscle , we used RT-PCR to assay for their presence in FSHD patient cells and various tissues from healthy individuals , using Sanger sequencing to confirm that each amplified product derives from the expected locus ( Table 2 ) . We found that most of the repeat-initiated transcripts we tested are expressed in myotube cells derived from an FSHD2 patient where disease-causing mutations result in de-repression of endogenous DUX4 , whereas we did not observe these transcripts in control myotubes that do not express DUX4 . This indicates that the low level of endogenous DUX4 present in FSHD muscle cells is sufficient to transcriptionally activate the same endogenous repetitive elements identified by our over-expression studies described above . Given the normal expression of DUX4 in testis , we assayed these repeat-initiated transcripts in human testis RNA from an individual unaffected by FSHD . We found that all tested repeat-initiated transcripts that respond to DUX4 in skeletal muscle are normally expressed in testis ( Table 2 ) , demonstrating that DUX4-repeat binding likely regulates transcription in the male germline . It is possible that factors other than DUX4 might also regulate transcription from these repetitive elements , perhaps explaining why we also detected some of these transcripts in other normal somatic tissue samples where DUX4 is not expressed ( Table 2 ) . For example , the internal regions of some ERVL and THE1C full-length repeats appear expressed in many tissues – we note that our primers recognize many copies of these repeats , and expression of only a single copy would enable us to detect transcription . Further research is needed to determine whether other transcription factors bind repetitive element promoters in those tissues . Other DUX family members might fill this role; their expression patterns and binding specificities are currently unknown . To further assess transcripts of repetitive elements in FSHD muscle cells , we performed a focused analysis of a small dataset of RNA-seq data from myotubes cultured from three control muscle and two FSHD1 muscle biopsies ( Yao et al . , manuscript in preparation ) . RNA-seq profiles show that most of the genes , lncRNAs and internal repeat regions we tested using RT-PCR are expressed in FSHD myotubes but not controls ( Figures S3 , S4 , S5 , S6 , S7 , S8 , S9 , S10 , S11 ) . In regions within 1 kb of DUX4-bound sites ( both bound repetitive elements and unique sites ) , ratios of expression in FSHD myotubes versus controls are well-correlated with the activation levels we found in lenti-DUX4 transduced myoblasts ( Figure 7A , Spearman's rho = 0 . 38 , p<10−15 ) . Expression ratios of internal MaLR/ERV regions in FSHD myotubes versus controls are also well-correlated with activation levels in lenti-DUX4-transduced myoblasts ( Spearman's rho = 0 . 49 , p<10−15 , Figure 7B ) . In addition , 13% of the 144 DUX4-bound repetitive elements that form alternative promoters for annotated genes show FSHD-specific transcripts and HSATII is expressed at ∼26-fold higher levels in FSHD myotubes ( median ∼2 . 2 reads per million ) than control myotubes ( median 0 . 08 reads per million ) . Therefore , the endogenous DUX4 that is expressed in just a subset of FSHD muscle cells is sufficient to drive expression from bound repetitive elements . We also performed a similar focused analysis using testis RNA-seq from the Illumina BodyMap data but the expression level of DUX4 was very low . Only the most abundant DUX4 targets ( according to our DUX4-transduced myoblast data ) were detected in the testis RNA-seq despite our ability to detect all tested transcripts by RT-PCR ( see Table 2 ) most likely because only a small proportion of cells in the testis express DUX4 [9] .
In this study we show that DUX4 binds many LTR repetitive elements of the MaLR and ERV families and initiates transcription at a number of those elements . Some DUX4-bound LTRs produce retrotransposon transcripts and others form previously unrecognized alternative promoters for human protein-coding genes , lncRNAs , and antisense transcripts . DUX4 also binds and activates transcription of the pericentromeric satellite HSATII . We initially identified these DUX4-activated transcripts in myoblasts transduced with lentivirally-expressed DUX4 , but show that many of the same loci are transcribed in FSHD but not control muscle cells , indicating that endogenously expressed DUX4 can activate LTR-driven transcription in FSHD muscle . We also show that all loci we tested using RT-PCR are expressed in the testis of an unaffected individual , suggesting that DUX4 drives transcription from at least some repetitive elements during normal development . Transposable elements can generate evolutionary novelty by exaptation [28] , [45]: their protein-coding regions can evolve to form a host gene , for example the mammalian placentation gene syncytin [46] , or their regulatory elements can affect the expression patterns or post-transcriptional control of pre-existing host genes [27] , [28] . Barbara McClintock initially proposed that transposable elements could alter expression of neighboring genes [47] , and her hypothesis is now supported by a growing body of literature describing repetitive elements that regulate transcription of host genes [32] , [33] . In some cases , repetitive elements of a particular family are enriched upstream of genes in similar functional classes [27] , [31] , [34] and may have provided a means to rewire transcriptional networks , distributing new transcription factor binding sites around the genome in short evolutionary time . We find that DUX4-bound MaLR and ERV repeats are used as alternative promoters for host genes and , at least in some cases , modulate gene expression in human testis . Although other TFs have been shown to bind LTR elements [29]–[31] , we provide the first demonstration of a transcription factor that binds and activates the MaLR subfamily of LTR elements , and a possible explanation for the rapid evolution of testis expression patterns that has been observed in mammals [35] . In a possible parallel with our results , Peaston et al . found that MaLR elements initiate dozens of genic transcripts in mouse oocytes [48] , although they did not identify the transcription factor ( s ) involved . DUX4-induced repeat-initiated transcripts also include a number of lncRNAs . Although the human genome contains several thousand lncRNAs , functions have been determined for only a few . Even those few functions appear diverse , including recruitment of chromatin-modifying factors , involvement in enhancer function , organization of nuclear substructures , and control of translation [49] , [50] . Many lncRNAs are testis-specific [42] , raising the question of whether DUX4 might be responsible for transcription of a subset of lncRNAs in the testis . In the future when lncRNA catalogs are more complete and their functions have begun to be elucidated , it will be interesting to revisit the question of whether DUX4-bound repeats played a role in the evolution of the germline lncRNA transcriptional network . DUX4 also activates the transcription of relatively intact copies of ERV and MaLR retrotransposons that do not splice to genes or lncRNAs , bringing up the possibility that it had a role in their genomic spread . In order for a retroelement to successfully invade the mammalian genome , it must be active in germ cells . However , because retrotransposition can also be harmful to the host organism [22] , [23] , retroelements whose activity is strictly restricted to germ cells could have an evolutionary advantage . The germline transcription factors that activated MaLR and ERV retrotransposition are currently unknown: the expression of DUX4 in testis and repression in other tissues together with its ability to bind and activate MaLR and ERV elements could suggest it had a role in repeat mobilization during evolution . DUX4's activation of relatively intact ERV and MaLR copies might additionally suggest that it currently has a role in their developmental epigenetic silencing . Recent years have yielded an increasing understanding of the mechanisms eukaryotes employ to defend against repetitive elements , including piRNA pathways and mechanisms that establish repressive chromatin marks [51] , [52] . These crucial defenses against repetitive elements are particularly active in germ cells and the early embryo and require an initial transcriptional activation of the retrotransposon to feed into the “ping-pong” cycle that produces and amplifies piRNAs that then silence homologous sequences [51] . It would be interesting in future to investigate whether DUX4 is involved in the initial activation of retrotransposon transcription in germ cells . Other pathways also exist to silence repetitive elements . For example , KAP1 controls endogenous retroviral elements by recruiting chromatin-modifying factors [53] , [54]; it is targeted to murine leukemia virus LTRs in a sequence-specific fashion by the KRAB-zinc finger protein ZFP809 [55] . To explain KAP1's more general role in ERV silencing , it is assumed that other sequence-specific DNA-binding proteins exist to target it to other ERV classes . The large number of diverse zinc finger proteins present in the mouse and human genomes may fill this role [56] , but the sequence-specificity of DUX4 for LTRs might also allow it to recruit repressive factors to repetitive elements in a cell-type specific context . Similar to retroelements , the pericentromeric satellite HSATII is bound and activated by DUX4 . HSATII is a pericentromeric satellite sequence , repeated in large tandem arrays close to a subset of human centromeres . Its consensus sequence is 170 bp long [24] and comprises ∼6 imperfect tandem copies of a smaller ∼25–28 bp repeat unit ( data not shown ) . Pericentromeric regions show evidence of transcriptional activity during specific stages of male meiosis [57] , and transcription of satellite sequences at early developmental stages appears be an important prerequisite for later establishment of repressive heterochromatin [58] . The activation of both interspersed repetitive elements and HSATII by DUX4 and its expression in germ cells of the testis could suggest a role in establishing repressive heterochromatin at both dispersed transposons and in tandemly repeated sequences near centromeres . Transcription of repetitive elements and satellite sequences in other , less appropriate biological contexts would likely be harmful , perhaps giving a strong evolutionary advantage to the location of the DUX4 retrogene within a high copy-number macrosatellite that can be tightly repressed by similar epigenetic means . DUX4 is normally expressed in the testis and epigenetically repressed in somatic tissues , but its variegated de-repression in muscle cells causes facioscapulohumeral muscular dystrophy ( FSHD ) [9] , [13]–[15] . Previous work has provided some insight into why DUX4 over-expression is pathogenic [10] , [59]–[61] . The repeat-initiated transcripts we describe here could also contribute to FSHD pathogenesis . For example , the HEY1 gene ( induced by DUX4-mediated activation of an upstream THE1B element ) can inhibit myogenesis by repressing myogenin and Mef2C [40] , and its activation might contribute to the muscle deficiencies seen in FSHD . In addition , expression of satellite transcripts can cause genomic instability [62] , [63]; DUX4-mediated activation of HSATII might similarly affect FSHD muscle cells . DUX4-induced expression of ERV and MaLR-encoded proteins or protein fragments could also have functional consequences in testis or FSHD muscle cells . Notably , some ERV-encoded env proteins contain a peptide with immunosuppressive properties [64] , perhaps contributing to the suppression of innate immunity we observe upon DUX4 over-expression in myoblasts [10] . Conversely , ERV-encoded protein fragments could be antigenic , and might elicit an immune response and some of the inflammation seen in FSHD muscle [65] , [66] . Our findings may also have clinical implications for cancer biology . The DUX4 target HSATII is expressed in a number of cancers [67] , and it has been shown that mouse cells lacking the genome caretaker gene Brca1 aberrantly transcribe satellite sequences leading to genome instability [62] . A large number of other DUX4 targets are known “cancer testis antigens” ( CTAs ) : genes normally expressed only in testis but de-repressed in some cancers , eliciting an immune response [10] . Furthermore , in Hodgkin's lymphoma cells , a THE1B-MaLR element provides an alternative promoter for the CSF1R proto-oncogene and de-repression of THE1B elements is widespread [68] . Together with the previous finding that the DUX4-containing D4Z4 repeat is hypomethylated in certain tumors [69] , these observations raise the question of whether DUX4 de-repression in cancers might mediate the activation of HSATII , CTAs and/or THE1B promoters . DUX4 is a primate-specific retrogene and a member of a small gene family that has experienced substantial change during mammalian evolution [1]–[3] . Although the binding preferences and functions of primate DUX4 orthologs and of DUX paralogs are still unknown , we note that an alignment of DUX family homeodomain sequences [1] shows that at least some of the residues predicted to determine DNA-recognition preferences [70] are different between DUX4 and the parental DUXC gene . Determining whether other members of the DUX gene family also bind and regulate retrotransposons will illuminate the importance of these largely unstudied genes and retroelements in the biology and evolution of mammalian germ cells and in muscle disease .
All experiments were performed with approval of the Institutional Review Board of the Fred Hutchinson Cancer Research Center . We wrote a number of custom scripts using R [71] , PERL , and several Bioconductor [72] , [73] and Bioperl functions [74] . We use the hg19 ( GRCh37/February 2009 ) reference human genome assembly and annotations provided by UCSC Genome Bioinformatics [75] , including the RefSeq [76] , lncRNA [42] and common SNP tracks , and phyloP scores for primates and placental mammals [77] . We also used RepeatMasker [78] analysis of the human genome assembly obtained via the UCSC site ( chromOut . tar . gz , which uses “RELEASE 20090120” of RepBase [24] ) . This version of RepBase recognizes ∼1400 human repetitive element types , classified into 56 families , with families classified into 21 classes . We obtained repeat consensus sequences from RepBase [24] . We used the Bioconductor GOstats package [79] to test for GO term enrichment , and GREAT analysis [39] was performed online ( http://great . stanford . edu ) . Our ChIP-seq data were previously published [10] ( GEO accession GSE33838 ) . Briefly , these 40 bp ChIP-seq reads derive from chromatin immunoprecipitated with a mix of two DUX4 antibodies . Chromatin was extracted from myoblasts transduced with lentivirus carrying DUX4 , or from negative control myoblasts that do not express DUX4 . The human genome contains multiple near-identical copies of DUX4: in our experiments , we used the full-length splice form of the most distal DUX4 copy on chromosome 4q35 , because this is the copy that appears to be expressed and pathogenic in FSHD patient muscle [9] , [10] , [13] . This DUX4 isoform is also expressed in testis , along with other copies containing minor sequence variants whose functional consequences are currently unknown [9] . We mapped each ChIP-seq read to the human reference assembly ( hg19 ) using BWA [80] . We eliminated multiply-mapping reads ( retaining reads with mapq >15 ) and determined peak locations [10] . We identified a position weight matrix ( PWM ) describing a motif that is strongly enriched among DUX4 peaks , using only the ∼24 , 000 peaks that do not overlap a repetitive element [10] . We determined a score threshold for this PWM of 9 . 75 , above which >97% of ChIP-seq peaks contain at least one motif . For further analysis of DUX4 binding sites we refined peak locations by identifying the single 17-mer subsequence with highest score to the DUX4 PWM , rather than using the entire peak region ( ChIP-seq resolution is limited by fragment size of ∼200 bp ) , making the simplifying assumption that each peak contains a single DUX4-binding site . We then determined whether each peak's best-scoring subsequence overlaps a repetitive element . To estimate peak-level enrichment for each repeat type , we divided the proportion of all peaks that overlap each repeat type ( observed ) by the proportion of base-pairs in the sequenced genome within that repeat type ( expected ) . We previously found that DUX4 binding sites are distributed relatively uniformly across different types of genomic regions ( promoters , intergenic regions , introns , etc . ) [10] , so it is not necessary to adjust our “expected” proportions for the different prevalence of various repeats in these region types . The DUX4 binding motif includes an average of 5 . 03 G or C residues among its 17 bases . In order to create an AT-matched set of genomic locations , we randomly selected 17-bp regions from the human genome , eliminating any that overlap assembly gaps , retaining those whose sequence contains 4–6 G or C residues , and downsampling the final set to contain 63 , 795 sites to match the dataset of DUX4 binding sites . We then determined whether these randomly selected 17-mers overlapped repetitive elements , and counted the types of repeats found among overlapping elements . We performed this sampling 10 times , and use the mean fraction of sites overlapping repeats to calculate enrichment measures shown in the columns of Tables S2 and S3 labeled “peak-based ChIP-seq enrichment estimate , compared to randomly sampled AT-matched regions” . In order to determine whether DUX4-bound repeats also tend to be bound by other regulatory factors , we obtained ENCODE ChIP-seq peak locations for 161 regulatory factors via the UCSC Genome Bioinformatics “wgEncodeAwgTfbsUniform” tables . For each factor , we chose a single representative ENCODE dataset , and determined the number of peaks that overlap DUX4 ChIP-seq peaks assigned as bound repeats ( see above ) . Using the entire peak regions ( several hundred base-pairs wide ) for both DUX4 and the other TFs ( rather than binding sites defined at higher resolution ) allows us to ask the biologically relevant question of whether TFs bind in the vicinity of DUX4 , rather than asking whether binding sites are exactly overlapping . In our standard ChIP-seq analysis ( above ) , we ignored sequencing reads that map to multiple genomic locations to ensure that called peaks likely represent true binding sites . However , this method is blind to binding in very recently duplicated regions , so we used an alternative bioinformatic method very similar to that of Day et al . [37] . This method examines repeat enrichment among individual sequencing reads , comparing counts of reads matching each repeat type in a ChIP-seq sample to counts in a negative control sample . In more detail , we first filtered ChIP-seq read datasets to remove low quality sequences , and reads that match our lentivirus-DUX4 constructs , the packaging constructs used during lentivirus preparation , or Illumina adapter sequences . We constructed an alternative repeat-based “reference genome” , where each repeat type is represented by a “chromosome” comprising every genomic instance of that repeat , with an amount of flanking sequence on each side equal to half the length of a sequence read , concatenated together with a intervening stretches of Ns that are each longer than a sequencing read . We then used BWA to map filtered reads to the repeat-based reference genomes , without filtering results for uniquely mapping sequences . We used the samtools idxstats program [81] to determine the proportion of filtered reads mapping to each repeat type . We estimated enrichment by comparing the proportion of reads mapping to each repeat in the ChIP sample with the proportion in the control sample , adding 0 . 5 to each count to avoid problems that would arise from division by zero . These enrichment estimates are “dampened” because ChIP samples contain background DNA derived from unbound genomic regions ( 50–90% of reads ) ; although immunoprecipitation depletes unbound sequences it cannot completely eliminate them . Background proportions differ between experimental and control samples , and background fraction undoubtedly contains many repetitive sequences . These read-based estimates are therefore likely a very conservative measure of true enrichment in the bound DNA fraction . Our RNA-seq data are available from GEO with accessions GSE45883 and GSE51041 . Two human myoblast cell lines ( 54-1 and MB135 ) were each transduced with lentivirus carrying DUX4 . After 48 hours ( 54-1 cells ) or 24 hours ( MB135 cells ) , RNA was extracted , poly ( A ) selected , and subjected to Illumina sequencing using standard protocols to generate 100 bp single-end reads . As negative controls , we also sequenced RNA from untransduced 54-1 cells , and from MB135 cells transduced with lentivirus carrying GFP . In addition , we sequenced RNA from two FSHD1 and three control myotube samples . Primary myoblast cell lines were received from the University of Rochester biorepository ( http://www . urmc . rochester . edu/fields-center ) and were cultured in DMEM/F-10 media ( Gibco ) in the presence of 20% heat-inactivated fetal bovine serum ( Gibco ) , 1% penicillin/streptomycin ( Gibco ) . Media was supplemented with 10 ng/ml rhFGF ( Promega ) and 1 µM dexamethasone ( SIGMA ) . Myoblasts were fused at 80% confluence in DMEM/F-12 Glutamax media containing 2% KnockOut serum replacement formulation ( Gibco ) for 36 hours . RNA was extracted , poly ( A ) selected , and subjected to Illumina sequencing using standard protocols to generate 100 bp single-end reads . Our analyses are conservative , identifying only the elements with greatest activation extents , because we lack statistical power due to small numbers of samples and a minor technical issue with the 54-1 control sample ( see below ) . In addition , we note that we are only examining polyadenylated transcripts and may be ignoring others; however , full-length transcripts of many repetitive elements are polyadenylated , including those of ERVs , L1s and Alu elements [23] , [82] . Our analysis of RNA-seq reads falls into two general categories , both described in detail below . First , we performed a read-based analysis as we had done for ChIP-seq reads , combining read counts across all instances of a particular repeat type; this method does not allow us to determine which instance of a repeat type is activated , merely that one or more elements of that class shows activation , but unlike standard methods , it does allow examination of recently duplicated sequences . We used the same read-based method as we did for the ChIP-seq reads ( see above ) on our RNA-seq reads . Although this method uses the BWA read-mapping tool and will therefore fail to map spliced reads , it does not suffer from a limitation of tophat that it suppresses mappings for reads mapping to many genomic locations . Second , we considered individual genomic locations ( repeat instances , genes , lncRNAs , etc . ) using tophat and DESeq , a method that limits our ability to examine highly similar multicopy sequences . We mapped reads to the genome using tophat [83] , allowing up to 20 map locations for multiply-mapping reads ( no map location is reported for reads that map to >20 locations ) . We counted reads overlapping each region of interest ( gene , lncRNA , repeat , etc . ) using the bedtools coverageBed function [84] with the “split” and “counts” options . We filtered regions to retain only those that had at least 10 mapped reads ( summed across the four myoblast datasets ) . We then used the DESeq Bioconductor package [85] to detect differentially expressed regions . We also repeated these analyses after filtering tophat output to retain only reads that map uniquely to the genome - results were very similar to those we obtained using all map locations ( data not shown ) . We encountered a minor technical issue: the RNA-seq read dataset for one sample , the untreated 54-1 negative control , has very low levels of contaminating reads from a lentivirus-DUX4 treated sample . A small number of reads match the lentivirus backbone , the DUX4 insert , and the lentivirus vector-DUX4 junction . In addition , we find small numbers of reads for genes ( and repeats ) that are activated to very high levels in DUX4-expressing cells but are “off” in cells that do not express DUX4 , very consistently at about 1/1000 of the number of reads found in 54-1 cells over-expressing DUX4 . The most likely explanation is that a small amount of RNA from another sample contaminated the untreated 54-1 cell RNA sample before sequencing . Although this issue only affects genes expressed to very high levels , it causes a technical problem for the DESeq statistical analysis method we used , because contaminating reads for genes expressed to high levels make dispersal estimation difficult ( data not shown ) . This issue contributes to the conservative nature of our conclusions . In order to examine the age of activated internal repeat regions relative to internal repeat regions that do not show obvious activation ( Figure S2 ) , we first applied the following filters to the full dataset of 100 , 864 regions of the human genome assembly annotated as internal regions of ERV or MaLR elements ( these regions were obtained via UCSC's track of RepeatMasker data ) . We first selected repeats flanked by DUX4-bound LTRs , requiring a ChIP-seq peak within 5 kb of the internal repeat region whose best DUX4 motif is within an LTR-type repeat element . We then filtered the dataset to only retain repeat regions spanning ≥500 bp , because repeat ages estimated from shorter regions are likely unreliable . This filtered dataset contains 10 , 190 internal repeat regions , including 92 of the 184 activated regions . We use the “milliDiv” statistic reported in UCSC's RepeatMasker track ( divergence from consensus sequence , per 1000 sites examined ) as a proxy for repeat age ( lower divergence = younger ) , dividing the number by 1000 to present a more intuitive per-site divergence measure . To analyze diversity of transcribed HSATII repeat units among RNA-seq reads ( Figure S12 ) , we first extracted all reads that mapped to any HSATII copy in our alternative repeat-based reference genome ( see above , in “Estimation of repeat enrichment among ChIP-seq reads” section ) . We then re-aligned those reads to a consensus-based HSATII reference sequence , comprising a full copy of the 170 bp consensus sequence from RepBase , concatenated to a second partial copy ( bases 1–98 ) , because HSATII is found in the genome in large tandemly repeated blocks , and we wanted to ensure we captured any RNA-seq reads that begin in the end of one repeat unit and continue into the beginning of the next unit . We used blastn [86] to align reads to this consensus HSATII sequence , tolerating mismatches , and used a custom PERL script to convert blastn output to sam format so that we could use the IGV browser [87] to view the resulting large alignment . In order to identify DUX4-bound regions that are used as previously unannotated promoters for genes or lncRNAs , we first used bedtools' intersectBed function [84] on tophat's genomic mappings to filter RNA-seq datasets to retain only reads that overlap DUX4 ChIP-seq peaks ( DUX4-bound regions ) . We additionally filtered reads to retain only those that contain an intron of at least 20 bp and that overlap annotated genes ( or lncRNAs ) . After these filtering steps , we created a table of peak-gene ( or peak-lncRNA ) pairs , counting the number of reads for each peak-gene pair in each RNA-seq dataset . We eliminated any peak-gene pairs where the peak and the gene themselves overlapped , and further focused on pairs linked by at least one read in both of the DUX4-overexpressing myoblast cell lines ( or in both of the FSHD patient myotube samples ) . For each peak-gene ( or peak-lncRNA ) pair , we estimated a DUX4 activation ratio , by comparing the proportion of reads linking that peak and gene in the two DUX4-expressing myoblasts ( or two FSHD patient myotubes ) with the proportion of reads in the two control myoblast samples ( or three control myotubes ) . Again , we added 0 . 5 to each read count to avoid problems with division by zero . We then filtered the peak-gene ( or peak-lncRNA ) list to only include pairs with a DUX4 activation ratio of ≥2 . Human tissue RNAs were purchased from BioChain ( Hayward , CA ) and had been DNase-treated by the supplier . Primary human myoblasts ( 54-1 and MB135 , neither of which has an FSHD mutation , and MB200 , from an individual with FSHD2 ) were collected and cultured as previously described ( Snider et al . , 2010 ) . 54-1 primary myoblasts were transduced with a lentiviral vector expressing either DUX4 or GFP ( as in our RNA-seq experiments ) . 24 hours after transduction , RNA was harvested for RT-PCR or 5′ RACE . Non-transduced 54-1 and MB200 cells were differentiated into myotubes by growing to 100% confluency and adding differentiation media for 48 hours ( Dulbecco's Modified Eagle Medium , 1% penicillin-streptomycin , 1% horse serum , 0 . 1% insulin , 0 . 1% transferrin ) . Total RNA was isolated from cultured cells using the RNeasy Mini Kit ( Qiagen ) followed by Invitrogen's protocol for DNase I ( Amplification Grade ) treatment with the addition of RNaseOUT ( Invitrogen ) to the reaction . DNase I reaction components were removed using the RNeasy Mini Kit ( Qiagen ) . RNA was eluted using 50 µl of RNase-free water , and the volume was reduced using a SpeedVac . cDNA synthesis was performed using 1 µg of RNA , SuperScript III reverse transcriptase ( Invitrogen ) and random hexamers ( Roche ) according to the manufacturer's instructions ( 50°C 30 min and then 55°C 30 min ) . Reactions were cleaned using the QIAquick ( Qiagen ) PCR purification system and eluted with 50 µl of water . Negative control samples corresponding to each cDNA sample were prepared by omitting reverse transcriptase . PCR reactions were performed with 10% PCRx Enhancer solution ( Invitrogen ) and Platinum Taq polymerase ( Invitrogen ) using 10% of the cDNA reaction as template in a total reaction volume of 20 µl in thin-walled MicroAmp reaction tubes ( Applied Biosystems ) . Primers are listed in Table S6 . PCR cycling conditions for cell culture samples were 95°C for 5 min , followed by 35 cycles of 95°C for 30 s , 55°C for 30 s and 68°C for 2 min , followed by a final extension of 7 minutes at 68°C . Cycling conditions for human tissue samples were the same , except that 45 cycles were used . PCR products were examined on 1% UltraPure ( Invitrogen ) agarose gels in TBE , cloned and sequenced using BigDye Terminators ( Applied Biosystems ) . 5′ RACE for the THE1B-HEY1 , MLT1B-PPCS , and MLT1E1A-NT5C1B transcripts was performed on total RNA using the GeneRacer kit ( Invitrogen ) . Prior to PCR with gene-specific primers and GeneRacer 5′ primers , the RT reaction was cleaned using QIAquick spin columns ( Qiagen ) as described above . Gene-specific reverse primers are listed in Table S6 . PCR products were gel purified , cloned into TOPO 4 . 0 ( Invitrogen ) and sequenced using BigDye Terminators ( Applied Biosystems ) . | Transposable elements ( TEs ) are found in most genomes , and many TEs create extra copies of themselves in new genomic locations by a process called retrotransposition . TEs are often thought of as genomic parasites that must be suppressed , because retrotransposition can cause great harm to their host organism . However , during evolution , the functions encoded by TEs have sometimes been co-opted to the advantage of the host genome as novel genes or as gene regulatory regions . We studied a human transcription factor called DUX4 that is normally expressed in testis and repressed in muscle . Sometimes muscle repression fails , causing the disease facioscapulohumeral muscular dystrophy ( FSHD ) . We find that DUX4 binds many TE types and can activate their transcription . Some activated TEs have been co-opted as novel promoters for human genes . DUX4's activation of these genes via TEs might be important in the biology of normal testis and may contribute to the FSHD disease process . Our findings raise the possibility that DUX4 and TEs co-evolved , as TEs may have hijacked DUX4 to aid their retrotransposition while DUX4 may have utilized TEs to modify its transcriptional network in the evolving germline . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [] | 2013 | DUX4 Binding to Retroelements Creates Promoters That Are Active in FSHD Muscle and Testis |
Stimulus-specific adaptation ( SSA ) occurs when neurons decrease their responses to frequently-presented ( standard ) stimuli but not , or not as much , to other , rare ( deviant ) stimuli . SSA is present in all mammalian species in which it has been tested as well as in birds . SSA confers short-term memory to neuronal responses , and may lie upstream of the generation of mismatch negativity ( MMN ) , an important human event-related potential . Previously published models of SSA mostly rely on synaptic depression of the feedforward , thalamocortical input . Here we study SSA in a recurrent neural network model of primary auditory cortex . When the recurrent , intracortical synapses display synaptic depression , the network generates population spikes ( PSs ) . SSA occurs in this network when deviants elicit a PS but standards do not , and we demarcate the regions in parameter space that allow SSA . While SSA based on PSs does not require feedforward depression , we identify feedforward depression as a mechanism for expanding the range of parameters that support SSA . We provide predictions for experiments that could help differentiate between SSA due to synaptic depression of feedforward connections and SSA due to synaptic depression of recurrent connections . Similar to experimental data , the magnitude of SSA in the model depends on the frequency difference between deviant and standard , probability of the deviant , inter-stimulus interval and input amplitude . In contrast to models based on feedforward depression , our model shows true deviance sensitivity as found in experiments .
Stimulus-specific adaptation ( SSA ) is the decrease in responses to a repeating stimulus ( standard ) that does not generalize to another , rarely-occurring stimulus ( deviant ) . SSA is a robust and widespread finding in the auditory system . It has been demonstrated in single neurons in primary auditory cortex ( A1 ) of anesthetized cats [1] , and in single-unit , multi-unit and local field potential ( LFP ) recordings in A1 of awake [2] and anesthetized rats [3 , 4] . Although SSA is present in the inferior colliculus [5–7] and the auditory thalamus [8–11] , it is mostly confined to the non-lemniscal pathway [8 , 9] and is thus likely generated de novo in A1 , whose major thalamic input is from the lemniscal part of the medial geniculate body . SSA in A1 has true deviance sensitivity: The response to a rare stimulus presented within a sequence composed mostly of the same standard tone ( which violates the expectation for yet another repeat of the standard ) is larger than the response to the same rare stimulus when presented within a sequence composed of many different tones ( that doesn't generate strong expectations for any stimulus ) although the level of sensory adaptation in the multi-tone sequence may be lower [4 , 12 , 13] . SSA is therefore an appealing test case for linking high-level concepts such as deviance sensitivity with mechanistic models . SSA shares many similarities with mismatch negativity ( MMN ) [14] , an event-related potential evoked by deviant stimuli that has been investigated extensively in humans [15] . While it is clear that SSA is not the direct neuronal correlate of MMN [16 , 17] , it may be one of the stages leading to MMN . In fact , human midlatency potentials correspond better with SSA temporally . Like SSA , midlatency potentials show deviance sensitivity [18] . It is therefore tempting to hypothesize that cortical SSA is the direct neural correlate of deviance sensitivity in midlatency potentials . There exist a number of models for deviance sensitivity , mostly in the context of the MMN . While some models are based on sensory adaptation [19 , 20] , others postulate an explicit prediction mechanism [21] . These models , however , are not very realistic at the single-neuron and local network level . In models based on sensory adaptation , changes in membrane excitability , which are a prominent mechanism of neural adaptation [22 , 23] , cannot easily explain SSA as they would affect similarly the responses to all stimuli . Therefore , SSA models have been primarily based on synaptic adaptation , which can be stimulus-specific . SSA may arise from synaptic depression in the thalamocortical ( ThC ) input to A1 , as studied e . g . by Lee and Sherman [24] . Indeed , computational models based on purely feed-forward connectivity show SSA [4 , 25 , 26] . However , these models provide predictions that are inconsistent with some of the experimental data [4 , 12] . Here we study an alternative mechanism: SSA may arise from the heavily recurrent cortical connectivity and the depression of the local , intracortical synapses in A1 [27] . We use a modified version of a recurrent neural network model with synaptic depression [28] , which reproduced important properties of A1 responses , including frequency tuning , forward masking , and lateral inhibition . An important feature of this model is the generation of synchronized firing events called population spikes ( PSs ) . The existence of PSs in A1 has support from intracellular [29 , 30] and multi-unit recordings [31–33] , and from network [34] and single-cell calcium imaging studies [35] . We study the basic properties of SSA and its robustness using both simulations and mathematical analysis , and provide predictions for experimental testing of the role of intracortical synaptic depression in cortical SSA .
SSA is commonly demonstrated using two tones that evoke similar responses in the recording site [1] and presenting them repetitively within an oddball protocol , where one tone occurs with high probability ( standard ) and the other occurs with low probability ( deviant; cf . Fig 2A ) . Hence , we analyzed SSA in the column midway between the “standard column” and the “deviant column” ( the columns whose best frequencies were the standard and deviant frequencies , respectively ) , referred to as the “middle column” . As evident from Fig 2A , SSA occurred in the model because standard tones often could not produce PSs in the middle column , yet the deviant tones could . SSA in this model looks somewhat different when considered from the point of view of the single neuron ( averaging many single-trial responses ) and from the point of view of the single-trial network activity ( averaging across many neurons ) , a dissociation that has been studied experimentally in auditory cortex ( e . g . [32] ) . Fig 2A shows the firing-rates of selected neurons in a single block of the oddball protocol consisting of 100 stimuli . In each single trial , neurons showed one of two response types . The first was a small transient increase in firing-rate ( Fig 2B , left , showing the responses of all neurons in trial no . 22 ) . The other is the signature of a PS , a concerted sharp increase in firing-rate ( Fig 2B , right , showing the responses of all neurons to trial no . 24 ) during which each neuron fired on average about one spike . Population spikes occurred mostly ( but not exclusively ) in response to deviant tones , as will be discussed in more detail below . The responses of single neurons across many presentations of the same stimulus could show substantial variability . The responses of a neuron to the standard tones ( e . g . neuron no . 75 . Fig 2D , top , shows responses from 500 trials , including those shown in Fig 2A ) mostly consisted of small increases in firing-rate but included also a few large responses that occurred when standards did evoke a PS ( light blue traces in Fig 2C , left; average response in blue ) . The responses of the same neuron to the deviant tones were mostly large , although they also included a few small responses , corresponding to failures of a deviant to evoke a PS ( e . g . trial no . 4 in Fig 2A; Fig 2C , right , light red traces; average response in red ) . In conclusion , sensory responses may occur in a single neuron whether or not a population spike occurred . The average response of a neuron to both standards and deviants is composed of trials with PSs , with a large and consistent response across neurons , and trials with no PSs . The difference between standard and deviant responses has to do with the probability of such events: PSs are likely to occur in response to deviant tones , with a few failures . On the other hand , standard responses show a high probability of failures , but the successful PS responses to the standard are similar in their magnitude to those evoked by the deviant . This situation is also found in experimental results [12 , 43] . To illustrate the mechanisms of SSA in the model , Fig 3A–3C show the firing-rates and time-course of synaptic resources not only in the middle column but also in its neighbors . PSs left a strong depletion of resources in their wake ( Fig 3C ) . Both the deviant and standard tones initiated PSs in their respective columns . However , standard tones initiated PSs much less often in the standard column , and even those that were initiated mostly failed to propagate into the middle column . Importantly , as discussed above , even when a stimulus failed to evoke a PS , sensory responses to that stimulus still occurred at least in some of the neurons in the column . PSs initiated in the standard column were of two types . Occasionally a PS could be evoked in the standard column during a train of standard tones . This occurred due to gradual recovery of resources from one tone to the next ( Fig 3D and 3E , left panels ) . Such PSs failed to propagate outside the standard column because of their small amplitude . The other type consisted of PSs evoked in the standard column by the first standard following a deviant ( Fig 3D , right panel ) . These PSs occurred because the deviant tone did not deplete the resources in the standard column as much as another standard tone would have done ( Fig 3E , right panel ) . This was primarily due to the adaptation in the standard column , which prevented the PS evoked by the deviant from propagating into it , resulting in little depletion of resources during the presentation of the deviant . In consequence , following the presentation of a deviant , the available resources in the standard column were higher than following the presentation of a standard tone , sometimes allowing PS generation in response to the next standard tone . However , in these circumstances the propagation of PSs from the standard column into the middle column generally failed , due to depletion of the resources in the middle column by the recent deviant-evoked PS . Also , although somewhat larger than the PSs of the first type , PSs of the second type were still of relatively small amplitude and thus limited in their ability to propagate away from the column in which they were initiated ( cf . Fig 3D , right , red and blue traces , and also Fig 3B within the green frame for the different ranges of propagation of PSs evoked in the standard and deviant columns ) . SSA in the model was not related to preference of the middle column for one tone or another . This was verified by presenting the two tones with equal probabilities of occurrence , as well as switching their roles as standard and deviant ( Fig 3F , top row , displays the average responses of these three conditions ) . Each tone evoked an average response depending mostly on its probability . When the two tones were presented with a probability of 50% , both evoked similar average response . These results reproduce the phenomenon seen in electrophysiological recordings , where responses depended on tone probability ( Fig 3F , bottom row; data from Taaseh et al . [4] ) : Responses to a tone in the deviant condition were stronger than responses to the same tone in the equal condition , and those were in turn stronger than responses to the same tone when standard . True deviance sensitivity requires more than just larger responses to rare stimuli: it is necessary that responses to deviant tones depend also on the identity of the other stimuli in the sequence . The extent of deviance sensitivity in our model was assessed using control protocols similar to those used in actual experiments [4 , 12] . These protocols were adapted from human studies [e . g . 44] , and included two types of sequences ( see Fig 4A and the Materials and Methods section ) : These three controls , together with the oddball and Equal protocols , gave rise to 6 conditions in which tones were tested ( Fig 4A ) . Since our model includes some random heterogeneity in the tuning-curves within each column , we ran all the protocols on 12 networks with different randomizations of the tuning curves . The Common-contrast SSA Indexes ( CSIs , defined in Eq 9 ) of the single neurons within the middle column of each network were all positive ( Fig 4B , left ) . Their distribution was bimodal , related to the background input received by each neuron ( see the Materials and Methods section for full description of the background inputs ) . Neurons that received strong , positive background input tended to respond strongly to standard tones and accordingly had relatively low CSI . Neurons that received negative background input and no sensory input had almost no standard responses but participated in the population spikes evoked by deviants , and accordingly had a high CSI . The CSI calculated from the mean firing-rate responses , averaged over all the neurons of the middle column , was slightly different from the mean single-neuron CSI ( CSI = 0 . 632 based on the mean firing-rate vs . 0 . 679 ± 0 . 204 from single-neuron responses , marked by an arrow and a gray line in Fig 4B , right , respectively ) . Throughout the paper we calculate the CSI from the mean responses of all neurons in a column ( rather than calculate CSIs of each neuron and then average ) . A histogram of the mean firing-rate CSI is shown in Fig 4B , right . Importantly , this histogram is very narrow ( CSI = 0 . 643 ± 0 . 007 , mean ± standard deviation ) . Our 12 networks therefore showed a similar level of SSA , despite the different randomizations of tuning-curves and the wide CSI distribution in the individual neurons of the middle column of each network . Average responses of the middle column to tones in all 6 conditions are plotted in Fig 4C . An example of multi-unit activity ( MUA ) responses in rat auditory cortex to tones presented in these conditions is displayed in Fig 4D for comparison . These data were collected by Taaseh et al . [4] , and are representative of the responses of both multiunit clusters and single neurons in rat A1 , as shown by Hershenhoren et al . [12] . In the model , as in the experimental data , three conditions tended to give rise to relatively small responses: Standard , Equal and Diverse Narrow . In these conditions , the tones had high probability ( Standard and Equal ) or were packed in a narrow frequency band , causing substantial cross-frequency adaptation ( Diverse Narrow ) . Three conditions evoked large responses: the Deviant Alone condition evoked the largest responses , as expected; and most importantly , responses to the Deviant condition were larger than those evoked by the same tone presented as part of the Diverse Broad sequence ( cf . scatters in Fig 4E; paired t-test: Deviant > Diverse Broad , t = 6 . 4 , df = 11 , p = 5 . 3x10-5 for f1 and t = 4 . 9 , df = 11 , p = 5 . 1x10-4 for f2 ) . Thus , the model passes the accepted test for true deviance sensitivity . Significantly , these simulation results are in line with the findings in rat auditory cortex by Taaseh et al . [4] ( cf . Fig 4D ) and Hershenhoren et al . [12] , who showed this pattern of responses in about half of their neurons . Why were the responses in the Deviant condition larger than in the Diverse Broad condition ? This requires consideration of the responses in the deviant column in these two conditions . In the Deviant condition , the standard tone provided some weak input to the deviant column , causing some adaptation and reducing the average deviant response relative to the Deviant Alone condition . However , since most standard presentations did not evoke PSs in the deviant column , the adaptation caused in this column by standard tones remained small ( cf . the response to deviant tones in Fig 3A relative to the response to the first standard tone , which should be similar to the Deviant Alone condition for f2 ) . In contrast , in the Diverse Broad condition the probability of a tone to evoke a PS in its column was much higher , due to the wide distribution of tone frequencies in the sequence ( Fig 4F–4H ) . These PSs propagated across the network according to its recent history , and some reached the deviant column and produced PSs in it . In consequence , PSs occurred in the deviant column at a rate that was higher than during an oddball protocol . This left the deviant column generally more adapted in the Diverse Broad condition than in the oddball protocol , and less able to generate responses that could propagate into the middle column . As an example , notice the PS generated in column 14 , just before time t = 5 s in Fig 4G . This PS propagated into the deviant column and left it too adapted , such that it could not generate a PS in response to the next presentation of its own best frequency , f2 = 12 , two stimuli later . Based on our results we conclude that the regularity of standard presentations in the oddball protocol effectively resulted in less adaptation in the deviant column compared to the Diverse Broad condition , and consequently the deviant responses in the middle column were stronger than Diverse Broad responses . Our model reproduces the experimental dependences of SSA on several parameters: The probability of deviant occurrence , P; the frequency separation , Δf ( defined here as Δf = f2 –f1 ) ; inter-stimulus interval ( ISI ) ; and input amplitude A . Smaller probabilities of deviant occurrence produced larger firing-rate responses to the deviant tone ( Fig 5A ) . This resulted in increased CSI , as in Ulanovsky et al . [1] . In addition , SSA was present only within a limited range of input amplitudes ( Fig 5B and 5C , left-hand panels ) , which was about 60% wider at the lower deviant probability ( Fig 5B , left ) . As the firing-rates in thalamocortical fibers are related to sound intensity , the decreased SSA with increased input amplitude in the model may correspond to the experimental findings of Taaseh [45] , who reported a decrease in CSI for increased sound level . Their findings , based on LFP recordings in rat auditory cortex , are re-plotted here in Fig 5D ( left ) . A general decrease in CSI for increased sound level was also reported recently by Nieto-Diego and Malmierca [42] , who recorded MUA responses across different fields of rat auditory cortex . SSA was non-monotonic not only as a function of input amplitude , but also as a function of inter-stimulus interval ( Fig 5B and 5C , right-hand panels ) . Here , the range of inter-stimulus intervals allowing SSA was about 16% wider at lower deviant probabilities ( Fig 5B , right ) . This is also in line with experimental results [1 , 4 , 12]; an example for the experimental dependence of CSI on ISI is provided in Fig 5D ( right ) , which re-plots the results of intracellular recordings by Hershenhoren et al . [12] . Increasing the frequency separation between standard and deviant , Δf , usually produced an increase in CSI , extending the ranges of input amplitudes and inter-stimulus intervals that allowed SSA ( Fig 5C ) . In general , SSA was present only for rather short ISIs ( < 1 s ) , except for the case of Δf = 6 that showed SSA for ISIs up to nearly 2 s , similar to the results of Ulanovsky et al . [1] at their largest Δf , as well as Hershenhoren et al . [12] . The effect of increasing the frequency separation on CSI in columns other than the middle column is presented below , as part of the experimental predictions of the model . Our model showed hyperacuity , which is another important feature of SSA in A1 [1]: SSA was present for Δf values substantially smaller than the width of the tuning curve . We first tested for hyperacuity by choosing f1 and f2 such that Δf was 10 times smaller than the input tuning-curve width . This setting put the two tones near the peak response frequency of the middle column , so that this column was adapted by both the standard and the deviant . Consequently , it couldn’t support the initiation of population spikes and did not show any SSA ( Fig 5E , light gray trace ) . This result emphasizes the fact that SSA in our model depends on the propagation of PSs between columns , rather than depression of thalamocortical synapses in the middle columns . As an alternative method for testing hyperacuity , we increased the frequency resolution of the model . This was achieved by increasing the width of the input tuning-curves . For these simulations , the width of the input tuning curve was set to 20 times the frequency difference between two nearby columns . In consequence , the distance between the two columns used for standard and deviant , Δf = 2 , corresponded to 1/10 of the thalamocortical tuning-curve width . Such a setting corresponds to finer columnar organization , i . e . shorter-distance connections within the cortex . SSA was present in this paradigm , albeit with lower CSI values and in a narrower range of ISIs compared to the standard network ( Fig 5E , dark gray vs . black traces ) . SSA is a non-monotonic function of most parameters of the simulation ( cf . Fig 5B and 5C ) . This property is due to the patterns of PSs evoked by the two tones in the oddball protocol . We identified four different regimes of PS initiation: ( i ) No stimulus was able to evoke a PS . This “No PS” regime occurred for low A values , when no stimulus was strong enough to elicit a PS ( Fig 6A ) , as well as for short ISIs when synaptic resources were too depleted for PS initiation and propagation . It was also found at small values of U , the fraction of resources used upon synaptic activation , which effectively reduced synaptic strength , and for long recovery time-constants of the synaptic resources , which kept the synaptic weights dynamically small . This regime showed no SSA , and was named “No PS” since it was the only regime that had no PSs , except a single PS that sometimes occurred at the beginning of a protocol . ( ii ) Deviant stimuli evoked PSs in the middle column whereas standards did not . This “Selective” regime ( Fig 6B ) is the one that showed strong SSA , as illustrated in Fig 3 . ( iii ) A Periodic regime , in which the standard tone evoked PSs in its column , and these invaded the middle column only once every few stimuli ( Fig 6C ) . PS responses to the deviant were not always successful because the middle column , and sometimes the deviant column as well , were adapted by the PSs initiated in the standard column . ( iv ) Each and every stimulus evoked a PS , regardless of its identity ( Fig 6D ) . This “Reliable” regime occurred for example at high input amplitudes and long ISIs . Regimes ( ii ) and ( iii ) are similar to the phenomenon of cycle skipping in excitable systems , where periodic excitation gives rise to responses in some but not all cycles . Fig 6E–6H illustrate the association of strong SSA with the Selective regime . CSI is plotted against selected parameters , as in Fig 5 , and the different response patterns , identified from the simulation traces , are superimposed in color . In addition to the stimulus parameters , A and ISI ( Fig 6E and 6F , black curves ) , the same regimes and the same dependence of CSI on parameters occurred when varying dynamical parameters of the local intracortical connections such as U , the fraction of synapse utilization , and τrec , the recovery time-constant of synaptic resources . Very short τrec gave rise to yet another pattern of activity , in which PSs were generated spontaneously . CSI was generally high only within the Selective regime , although there was also moderate SSA on the margins of the Periodic regime . As mentioned above , one of the modifications we made to the model of [28] was to introduce heterogeneity in the tuning curves within each column , following experimental findings in mouse A1 [46 , 47] . We use the sensitivity of SSA to stimulus parameters to assess the effect of this feature of the network ( Fig 6E and 6F ) . The same parameters were tested for a heterogeneous network , as was used throughout this work ( black ) and for one with homogeneous columns , where all the input-receiving neurons in each column have the same best frequency ( gray; the regimes marked in color on the plot area are for the heterogeneous network ) . In the homogeneous network , SSA existed in a similar range of amplitudes and in a range of ISIs that included shorter values compared to the heterogeneous network . Thus , the basic mechanism of SSA was unaffected by the heterogeneity , which primarily contributed to the existence of SSA at longer ISIs . It seems therefore that SSA exists in this model only within a relatively narrow range of parameters . To verify this claim generally , we searched exhaustively through the space of possible parameters of the stimulus sequence ( amplitude and ISI ) and through the space of possible network parameters ( U and τrec ) . Fig 7A presents the CSI for different combinations of A and ISI , with the different response regimes delineated on the map . Strong SSA is mostly confined to the Selective regime . The existence region of SSA has the shape of a narrow , curved band on the A-ISI plane . It consists of two “branches” , one spanning a wide range of ISIs but confined to low input amplitudes , and the other confined to short ISIs and spanning a wide range of input amplitudes . Fig 7B shows the CSI map and the regimes obtained in a homogeneous network , where all inputs to a column share the same best frequency . Comparison with Fig 7A reveals that network heterogeneity resulted in a slightly expanded low-amplitude , long ISI branch of the existence region for SSA . We studied the existence region for other parameters as well . For example , when keeping the stimulus parameters fixed ( A = 5 Spikes/s , ISI = 350 ms ) and varying the synaptic parameters U and τrec ( Fig 7C ) , SSA was also confined to a narrow band of parameters ( Fig 7C ) . These calculations extend the results of Fig 6G and 6H . Here , the existence region for SSA was somewhat wider when both U and τrec were large . The existence region found for SSA in our simulations can be viewed as a consequence of the difference between responses of the deviant and standard columns , which are affected differentially by adaptation . SSA exists when the slightly-adapted deviant column generates PSs while the strongly-adapted standard column doesn’t . Consider the responses of the deviant and standard columns as a function of any of the parameters of the model . Such a relationship is called here the ‘response function’ . Adaptation modifies the response function , and its net effect can be heuristically modeled by a modification of the relevant model parameter . For example , adaptation as a function of sound level can be described by effectively reducing the sound level of the input . Operations on the input of the response function fall into two major classes [48] , illustrated in Fig 8A: ( i ) Subtraction , which shifts the response function along the parameter axis while preserving its slope ( left ) ; and ( ii ) division , which scales the slope of the response curve ( right ) . In Fig 8B , the responses of the middle column , which is where the SSA was evaluated , to standard and deviant stimuli ( blue and red ) are displayed as a function of log ( ISI ) ( left ) and log ( A ) ( right ) . These responses are largely inherited from the standard and deviant columns ( where the PSs are generated; light blue and light red ) , although responses to deviants were attenuated more strongly in the middle column ( red vs . light red ) compared to the responses to standards ( blue vs . light blue ) . The responses to standards showed a sigmoidal , monotonic dependence on both parameters . The response curves to the deviants were shifted to smaller values , and also somewhat distorted: for both parameters , there was a non-monotonic region corresponding to mid-range standard responses . Comparing these results of the simulation with the schematic curves in Fig 8A , we see that adaptation acts as a subtractive operation along the logarithmic ISI scale , and at least within the low-A monotonic region of the deviant responses , it is a divisive operation along the log ( A ) scale ( Fig 8B ) . While the divisive operation on stimulus amplitude doesn’t have an obvious explanation , the subtractive effect along the log ( ISI ) scale is easily explained by considering the effective ISI . On average , this was ISI/ ( 1 – P ) in the standard column , where P is the probability of occurrence of the deviant tone , but ISI/P in the deviant column . On the logarithmic scale , these become shifts of –log ( 1 – P ) and –log ( P ) with respect to a column stimulated repetitively ( P = 1 ) . Interestingly , these shifts are the expressions for the objective surprise associated with the standard and deviant tone , respectively . Thus , at least under these conditions , the model roughly implements a predictive coder [49] . Since the existence region for SSA is rather narrow , we were interested in mechanisms that may increase its extent . We show here that the depression of the thalamocortical synapse is one such mechanism , whose inclusion in the model increases the range of parameters producing large CSI ( Fig 7D ) . Changing Us , the fraction of resource utilization in the ThC synapses , produced a pattern similar to changing A or U: Low values gave rise to the No-PS regime and higher values to the Selective and Periodic regimes ( Fig 9A ) . In contrast , changing the value of τsrec , the time-constant of recovery of the ThC synapses , did not abolish the Selective PS responses ( Fig 9B ) . The only effect of increasing τsrec was a slight increase in CSI . Robustness to changes in τsrec is highlighted by the fact that tested values spanned over two orders of magnitude . The dynamics of ThC depression in our model ( Eq 5 in the Materials and Methods section ) can be treated analytically . When presenting a sequence of identical , repetitive stimuli , the mean fraction of available resources in the ThC synapses ( denoted by z ) is periodic with a period that equals the inter-stimulus interval ( Fig 9C , bottom ) . This is approximately what happens during a long sequence of standard stimuli . Importantly , treating the system as an iterated map reveals that z assumes the same value at the onset of each stimulus ( onset z value ) and another value at the offset of each stimulus ( offset z value ) . These values can be derived analytically under some approximations ( Eqs 17 and 18; resulting values shown as dashed lines in Fig 9C , bottom ) . Importantly , the analytical derivation further reveals that this is the only steady-state behavior of the map , so that the resources of the ThC synapses cannot show oscillations with longer periods . Thus , ThC depression cannot account for the more interesting dynamical phenomena in the network . For example , in Fig 9C the population spikes within the standard column occur once in every three stimulus cycles , while the dynamical parameters of the ThC depression only follow the faster periodicity of a single stimulus cycle . Similarly , ThC depression alone cannot produce the phenomenon we observe in the Periodic regime , in which PSs may occur in the middle column on some but not all stimulus presentations ( Fig 6C ) . In both cases , Periodic PS responses to trains of standard stimuli resulted from the dynamics of the synaptic resources of the intracortical connections ( denoted respectively by x and y ) . Indeed , while z follows the rate of stimulus presentation , x and y generally follow the slower rate of the population spikes . We use the onset and offset z calculated analytically ( Eqs 17 and 18 ) to examine the effects of ThC depression within the SSA Region . The SSA Region found in our simulations consists of two “branches” , in which ThC depression plays different roles ( Fig 9D ) . The first branch , at low stimulus amplitude and spanning a wide range of ISIs , has high values of both offset and onset z . The equations show that the low-level stimuli that gave rise to this branch in the simulations hardly deplete ThC resources , and the moderate ISIs allow recovery of virtually all the depleted resources . Thus , ThC depression has little effect on network responses in this branch , explaining the robustness of SSA with respect to τsrec found in our simulations ( Fig 9A ) . The second branch lies in the short-ISI range and spans a wide range of stimulus amplitudes . In this branch , the equations show that stimuli cause significant ThC depletion ( low offset z ) and little resource recovery ( low onset z ) . This combination causes a strong effective attenuation of stimulus amplitude , explaining why in the simulations the CSI values were quite uniform ( Fig 7A ) –high-amplitude stimulation was effectively attenuated to input strength that was comparable to that of low-amplitude stimulation . The analytical treatment suggests therefore that this branch is formed by “stretching” the short-ISI , medium-amplitude region into the high-amplitude range . We note that both branches of the SSA Region have a similar , rather negligible amount of resources recovered during the ISI ( Fig 9D , right-hand plot ) . To summarize , the main effect of thalamocortical depression in our model is to weaken the dependence of CSI on some model parameters . This is illustrated in Fig 9 for input amplitude–at least at short enough ISIs , ThC depression adaptively reduces stimulus amplitude and therefore brings the activity pattern into the selective regime . The same analysis is also true for the effect of the fraction of resource utilization in the thalamocortical synapses ( Us ) on SSA , which was found to resemble that of stimulus amplitude ( cf . Figs 6E and 9A ) : At the low-level , long-ISI region Us indeed strongly affected SSA . However , within the short-ISI branch , SSA should be only weakly dependent on Us , which effectively acts as a modifier of stimulus amplitude . Thus , ThC depression has an important role in shaping the SSA Region itself , expanding the parameter region formed by intracortical depression alone . This finding was confirmed by our simulations of the same network as the one used for Fig 7A , except that ThC depression was removed . We tested this network on the same range of stimulus parameters ( Fig 7D ) : The Selective regime without ThC depression was narrower than that of the standard network , and while it showed strong SSA in the low-A branch , SSA was weak and existed for a more limited set of parameters on the short-ISI branch . Our simulations of a network with multiple columns allowed us to study model responses to the oddball protocol in columns other than the middle column . For these columns , the two tones presented were usually located on the same side of the peak of the tuning curve ( Fig 10A and 10B ) . We found that in columns whose best frequency corresponded exactly to either the standard or the deviant , SSA was considerably weaker than in the middle column . This is because adaptation to the best frequency as the standard prevented PS responses to deviants from entering the column , while responses to the best frequency when it served as deviant were attenuated by cross-frequency adaptation . It should be noted that when Δf is smaller than 2 , the smallest value shown in Fig 10B , the middle column shows weak SSA as well ( cf . Fig 5E ) . In contrast , in columns with a best frequency either lower than f1 or higher than f2 , SSA could be even stronger than in the middle column . This resulted from large deviant responses in one of the protocols . For example , columns whose best frequency was lower than the low frequency of the oddball sequence ( f1 ) showed large responses to f1 as deviant but not to f2 as deviant . Responses to f2 as deviant in these columns were small because PSs evoked in the f2 column had to travel through the strongly adapted f1 column before reaching the low-best frequency columns . The CSI in some of the low-best frequency columns was even higher than in the middle column because deviant responses to f1 suffered less cross-frequency adaptation from the standard , which was farther away on the frequency scale . The opposite occurred on the other side of the network . The results described here correspond to playing an oddball protocol with f1 and f2 not centered on the best frequency of the recording site . Such a paradigm has not been systematically studied in cortical recordings . The model makes clear predictions for the effect of tone duration on SSA . Fig 10C presents a map of CSI values as a function of tone duration and the offset-to-onset interval . The latter quantity is highly relevant for the dynamical behavior of the network , as it is the time allowed for recovery of synaptic resources . High CSI values in Fig 10C are limited to shorter tone durations and offset-to-onset intervals , and SSA was always abolished at long tone durations . However , the reason that SSA disappeared at long tone durations depended on the offset-to-onset interval: At long offset-to-onset intervals , longer tone durations produced more frequent PS responses ( Periodic and Reliable regimes ) . This was due to the ongoing recovery of intracortical resources , showing that thalamocortical depression did not play an important role at these intervals ( cf . the high onset z for long ISIs in Fig 9D ) . In contrast , at short offset-to-onset intervals , SSA was abolished at long durations because no PSs were elicited . We attribute this to the accumulating depletion of thalamocortical resources from one stimulus to the next , which rendered them too low for eliciting PS responses even in the deviant column . We note that at intervals around 300 ms , used in most of the simulations above , SSA was quite robust to changes in duration but not to changes in the interval . Indeed , Hershenhoren et al . [12] recently reported that in rats , while CSI is significant for tone duration of 30 ms and offset-to-onset intervals of 270 ms , it becomes rather weak at intervals of 670 ms and 1170 ms ( cf . Fig 5D , right ) .
SSA in our model is first and foremost a network phenomenon that does not require thalamocortical depression . It depends on intrinsic cortical dynamics that produce population spikes and allow their lateral propagation across the network . The existence of population spikes in A1 has been inferred from the distribution of EPSP amplitudes [29 , 30] , multi-electrode recordings [31–33] , and network [34] and single-cell calcium transients [35] . In-vitro recordings also show population events occurring in A1 upon stimulation of the thalamus [50] . The evidence suggests that large ensembles of neurons may be active during such events: Bathelier et al . [35] , for example , used 2-photon calcium imaging with single-cell resolution and found that the population events dominating the activity in A1 included a large fraction ( over half ) of the imaged neurons . The PSs in our model are of slightly faster rise-time and shorter duration relative to evoked population bursts recorded in rat A1 [31–33] . While the underlying mechanisms have not been identified , modeling work [36 , 37] has suggested that population spikes may be a consequence of the dynamics of a network with depressing synapses . Loebel et al . [28] further demonstrated that such networks can be used to provide a unified account for many response properties of neurons in auditory cortex , including forward masking , lateral inhibition and hypersensitive locking suppression . Here we showed that such a network can produce SSA as well . The model we presented here , similar to [28] , should be taken as a simplified representation of A1 . The use of a rate-model to represent single neurons in A1 was introduced by Loebel et al . [28] , based on a previous work that showed the equivalence of such a model to a firing-rate description of a recurrent network with synaptic depression in terms of its activity [36] . The model is stripped of many details such as layers , feedforward inhibition , cell types and numbers , accurate connectivity probabilities , or realistic connectivity profiles . The simplifications made it possible to focus on a minimal number of basic features and demonstrate that they are sufficient to support cortical SSA . Nevertheless , the model keeps some crucial features of the actual rodent auditory cortex . In cortex , connection strengths and probabilities decay smoothly with spatial distance [27] . In the model , this organization was discretized into columns . In rat A1 , the tonotopic gradient spans about 2 mm and covers about 7 octaves [40]; in mouse A1 the tonotopic axis is somewhat shorter . Since the model had 21 columns , each corresponded to about 1/3-1/2 octave . Tuning curves in the model have a half-maximum width of 5 columns , or 1–2 octaves . This is similar to curves measured in cortical recordings [40] . Based on the above numbers , the usual frequency separation of our simulations ( Δf = 2 ) corresponds to about 2/3 of an octave , or slightly above the typical frequency separation used in experiments , which is 0 . 53 octave ( the 44% of [4 , 12] and 0 . 37 of [1] ) . Thus , the main conditions of the simulations reflected the typical experimental conditions used to study SSA . The connection strengths within each column were set to prevent spontaneous PS activity ( see Materials and Methods ) . While it is difficult to compare between synaptic strengths in a rate-model and electrophysiologically-measured quantities , our tests of single-neuron activation in the model showed that connections between excitatory neurons within the same column were weaker than those reported in-vitro in mouse A1 . On the other hand , connections were more probable [27] , since our network had full connectivity within each column ( see Materials and Methods ) . Synaptic depression parameters used here are on the same order as those reported in A1 and used in modeling its activity ( U = 0 . 55 , τrec = 500 ms in [51] , Us = 0 . 8 , τsrec = 1 s in [52] ) . Finally , following [28] , our model features purely feedback inhibition , where inhibitory neurons do not receive sensory input . Feedforward inhibition is known to have an important role in A1 , arriving at a typical delay of a few milliseconds following the excitatory ( presumably thalamocortical ) currents [38] , presumably due to the additional synapse on the way . Adding an explicit delay would significantly complicate our rate model ( see e . g . [53] for an implementation of such a delay ) . We chose not to implement a synaptic delay , instead emulating the effect of delayed inhibition by feedback inhibition . Indeed , in simulations in which inhibitory neurons also received sensory input , the resulting responses were similar to those of the main simulations as long as the direct sensory input to inhibitory neurons was relatively weak . With stronger inputs to the inhibitory neurons , responses were abolished due to the simultaneous arrival of excitatory and inhibitory inputs onto the neurons in the network . The absence of direct feedforward inhibition should not affect our conclusions regarding the role of synaptic depression in shaping SSA , based on the following reasons: ( i ) The first evoked spikes in the thalamo-recipient neurons are already sufficient to trigger the population spike , with feedforward inhbition having its effect only following these initial spikes [38] . Furthermore , once a population spike is triggered , its propagation involves feedforward inhibition , since the intercolumnar connections target both excitatory and inhibitory neurons . Therefore , feedforward inhibition should not prevent the initiation of population spikes . ( ii ) Synaptic depression would come into play once the first spike is fired , and is the main factor that prevents spiking over the longer time-scale , i . e . for inter-stimulus intervals longer than 100 ms [54] . ( iii ) The feedforward inhibition itself undergoes SSA , thus scaling with the excitatory responses and at most modulating the SSA , not generating it [55] . Some insight into the effects of feedforward inhibition on SSA may perhaps be gained from the model of Schiff and Reyes [52] . Their results suggest that at the low stimulation rates used in the oddball protocol , feedforward inhibition may balance the effects of thalamocortical depression and result in a relatively constant net drive to the excitatory neurons . Such an effect would emphasize the role of depression of the recurrent synapses in generating SSA ( cf . Fig 7D , which shows the extent of the SSA region without thalamocortical depression ) . The main weakness of our model is its sensitivity to stimulus and network parameters ( Figs 6 , 7 and 9A ) . The existence range of SSA in parameter space is expanded by mechanisms such as depression of the thalamocortical synapses and , to a lesser extent , heterogeneity of the tuning curves . However , even with the action of the above mechanisms , SSA in our model is not a robust phenomenon . The sensitivity of SSA to network and stimulation parameters stems from its existence in the intersection region of the requirements of PS response to the deviant tone but no PS response to the standard tone ( as illustrated by the response curves in Fig 9 ) . The extent of this region is determined by the constraints posed by the frequency-tuning of columns and by the lateral propagation: Propagation should be strong enough to deliver deviant responses into the middle column , yet weak enough to prevent most of the successful standard responses from doing so . For the model to be a viable account of cortical SSA it is necessary to assume the existence of tight regulation of the cortical network , which drives it into the SSA Region . One possible locus of such regulation may be the narrow dynamic range of firing-rates in thalamocortical fibers [9] and/or thalamocortical EPSPs in layer-4 input neurons [56] , which in our model would both correspond to a limited range of input amplitudes . Sources of regulation might include neuromodulatory control , a mechanism that could conceivably drive A1 between states favoring SSA and states favoring other computational tasks . Cholinergic input , for example , may control important parameters such as the utilization parameters for sensory and recurrent synapses [57–59] . Two other classes of models for SSA have been studied previously . The first class of models is based on depression of the feedforward synapses [4 , 12 , 25 , 26] . Thalamocortical depression was not essential for the existence of SSA in our model , so that the mechanism studied here is indeed different from the SSA that depends on feedforward depression . The second class of models consists of networks with recurrent connections . The emergence of SSA from synaptic depression in the local , cortical connections has also been shown in a recent computational work by May et al . [20] . Their model is different from ours in that they ignored the single neurons , modeling the interaction between cortical columns . Their analysis is not detailed enough to understand whether the basic mechanisms that produce SSA in their model are the same as in the model presented here . Furthermore , they provide only limited information on the dependence of SSA on stimulus and network parameters . A different network model of SSA-like responses has been developed by Wacongne et al . [21] . Their model depends on synaptic dynamics , although not on synaptic depression . However , their model has been developed in order to test a specific model of predictive coding , and some parts of it ( e . g . the short-term memory component ) are hand-crafted to have the required properties . Thus , it is different from the network model we describe here , which uses generic network mechanisms to produce SSA . In consequence , all further comparison with existing models considers only feedforward models of SSA . Our model correctly accounts for a number of properties of SSA in rat auditory cortex . SSA increases with frequency separation between the tones and with decreasing probability of the deviant [4 , 12] . SSA decreases at high sound levels [45] . Similarly , SSA weakens as ISI is increased [4 , 12] . The model suggests that SSA should also decrease at low amplitudes and very short ISIs . The low-amplitude and short-ISI ranges are presumably overlooked in experimental studies because of the difficulty to elicit responses with such parameters . All of these properties are also correctly predicted by other models of SSA [25 , 26] . On the other hand , in the model presented here , single-trial responses to standard tones show occasional successes , consistent with a recent intracellular study of SSA [12] , where the largest responses to the standard tone were as large as those in the Deviant and Deviant Alone conditions . The existence of occasional strong responses to the standard tone may also find support in the spike-count distributions in [43] , where successful standard responses were similar in strength to the deviant responses . Feedforward models cannot easily account for such findings . Importantly , the model has true deviance sensitivity , in which responses to deviants within a regular background were stronger than to deviants within an irregular background ( “many standards” , Jacobsen and Schroger [44] ) . This phenomenon could not be explained by a feedforward model of adaptation in narrow frequency channels [4] , at least not when frequencies were close enough to produce cross-frequency adaptation . Mill et al . [25] also showed a small preference for true deviants over the “deviant-among-many-standards” , but only at large frequency separations ( equivalent to separations larger than our Δf = 2 ) . At these frequency separations , the deviant and standard were far enough apart that the standard in the oddball sequence did not induce strong cross-frequency adaptation of the deviant responses , while the many-standards condition did [25] . In contrast , in our model the many-standards ( our Diverse Broad ) condition generically gives rise to cross-frequency adaptation that is stronger than in the oddball sequence . The many-standards sequence , with its extensive propagation of population spikes across the network , caused more activity in the deviant column than the oddball sequence , and accordingly also more adaptation . Consequently , the deviant column was able to produce stronger responses to its best-frequency tones during the oddball sequence than during the many-standards sequence . Neuronal responses in rat auditory cortex indeed often show true deviance sensitivity [12] . In addition to modeling existing data , the model makes predictions that have not yet been tested in cortex . In our model , the CSI depends on the positions of the two tones relative to the column’s best frequency , with a minimum CSI when one of the tones is equal to the best frequency . To our best knowledge , this prediction has not been tested systematically in cortical recordings . Interestingly , such conditions have been tested in the inferior colliculus by Duque et al . [60] . Their results are qualitatively different from our prediction for cortical SSA in that they do not show a decreased CSI for f1 or f2 equal to the best frequency; rather , they are consistent with purely feedforward depression and the characteristic asymmetry of tuning curves in the auditory system . Bäuerle et al . [11] have also looked at SSA with frequency pairs not centered on the site’s best frequency , in gerbil ventral MGB . Most of these results favor the feedforward depression mechanism . Thus , in subcortical stations , feedforward depression may be the dominant mechanism shaping SSA . Our model predicts that SSA should be found within a limited range of stimulus durations ( cf . Fig 10C , the longer-duration “branch” ) . Significantly , our prediction contradicts that of a model based on feedforward depression [26] , which suggests that SSA should become stronger when stimulus duration is extended . We know of no study that checked systematically the dependence of SSA on stimulus duration . However , there is a tendency to use relatively short stimuli in SSA experiments in rats ( 30 ms , [3 , 4 , 12]; 75 ms [5 , 9 , 60] ) , suggesting that indeed longer stimuli result in less SSA . SSA has been demonstrated in somatosensory and in visual cortices . Mechanisms similar to those studied here may be at play in the generation of SSA in barrel cortex . In rat barrel cortex , SSA has been reported for whisker identity as well as for velocity and direction of whisker deflection [61] . The responses to non-principal whiskers are probably mediated by lateral connections [62] . The similar amplitudes of deviant and ‘many-standards' early responses in barrel cortex [61] are better explained by adaptation of intracortical synapses , as in our model , rather than adaptation in the thalamocortical synapses only [4] , since the latter mechanism predicts higher responses in the ‘many-standards’ condition ( see above for comparison of these two types of models ) . In visual cortex , studies in cats [63] and monkeys [64] have found that complex cells adapt to a specific orientation while generally retaining their responses to other orientations . Such cells , therefore , show SSA to orientation . In contrast with auditory and somatosensory cortices , in visual cortex this phenomenon can be explained by a purely feedforward mechanism , with adaptation in simple cells ( e . g . due to fatigue ) shaping the tuning curves of complex cells in a stimulus-specific way ( as suggested in [64] ) . Such a mechanism would be in line with the “cascading” adaptation characteristic of the visual system , e . g . in the way spatial adaptation in V1 appears to arise from integration of LGN responses [65] and motion adaptation in MT can be explained by broadly-tuned adaptation of their input V1 neurons [66 , 67] . Recently , true deviance sensitivity has been demonstrated in visual cortex using both extracellular recordings and calcium imaging , but the difference between deviant and many-standards responses occurs late [68] . This difference may challenge the purely feedforward explanation of adaptation in visual cortex , even if only indicating a late top-down modulation . We note that consideration of recurrent connectivity as essential to explaining V1 phenomena was emphasized in recent modeling work by Chariker et al . [69] , although not in the context of adaptation . Importantly , the mechanism generating SSA in our model generalizes to any stimulus or modality , if the tonotopic columns are replaced with populations coding for different stimuli or features . These populations need not be spatially segregated , but should have ( i ) stronger intra- vs . inter-population connection strengths and ( ii ) different sensitivities to afferent inputs . Thus , activity-dependent synaptic depression combined with heterogeneity in synaptic strengths leads to SSA for any stimuli encoded by the connectivity . In higher order areas of the auditory system , this can lead to SSA for complex stimuli .
We model primary auditory cortex ( A1 ) , using multiple cortical columns ( Fig 1A ) . Our model ( described in detail below ) closely follows Loebel et al . [28] , which was based on the rate model of Wilson and Cowan [70] except that the basic units represented single neurons rather than local populations of neurons . A previous work [36] showed the equivalence of this rate model and a network of integrate & fire units in terms of the spontaneous and evoked activities , and importantly in the generation of population spikes . Our main modifications to the model of Loebel et al . [28] were the introduction of ( i ) synaptic depression in thalamocortical synapses and ( ii ) heterogeneity of tuning curves within each column , as suggested by imaging studies of mouse A1 [46 , 47] . We note that although cortical SSA was mostly reported in the cat [1] and rat [2 , 4 , 12 , 16] , the physiological properties of A1 , e . g . the heterogeneity of tuning curves on which we rely here , are better described in mice . All simulations , data analysis and calculations were performed using MATLAB ( MathWorks ) . Each column in our model is a fully-connected recurrent network: all neurons within the column are connected to all others . Our simulations show that population spikes occur also in a network with connection probabilities as found in vitro [27] . In general , even with a connection probability of around 0 . 1 , as reported by Levy and Reyes [27] for connections between pyramidal neurons , the shortest path between any two neurons is on average less than 2 synapses ( see e . g . [71] ) , and recruitment of the neuronal population into the population spike is rapid . Each column represents an iso-frequency band along the tonotopic axis of A1 , covering about 1/3 of an octave , and consisting of NE excitatory and NI inhibitory neurons . Each neuron is described by two dynamic variables: firing-rate , denoted by EiQ for excitatory and IlQ for inhibitory neurons ( Q specifies the column index and i or l the index of the neuron within the column ) ; and the amount of resources available at each of the synapses made by this neuron , relative to its full resources , a fraction between 0 and 1 denoted xiQ for excitatory and ylQ for inhibitory neurons . Every neuron in the network receives external input , denoted eiE , Q for excitatory and elI , Q for inhibitory neurons . External inputs are uniformly distributed within a specific range of firing-rates , i . e . each neuron receives input that depends linearly on its index within the column . All neurons within each column receive input from the excitatory populations of the nearest- and second-nearest neighboring columns , with connection strengths that decrease according to inter-column distance . Most excitatory neurons in the network also receive sensory input through a set of thalamocortical ( ThC ) synapses , each mediating a different tone frequency f and having its own dynamic variable , ziQ , f , representing the fraction of synaptic resources available at this synapse . The network dynamics are defined by the following equations , following Loebel et al . [28] with the modifications described above: τEdEiQdt=−EiQ+ ( 1−τrefEEiQ ) ⋅g ( ∑R=−22JEE|R|NE∑j=1NEUxjQ+REiQ+R+JEINI∑j=1NIUyjQIjQ+eiE , Q+∑f=1MUsziQ , fsfTiQ , f ) ( 1 ) τIdIlQdt=−IlQ+ ( 1−τrefIIlQ ) ⋅g ( ∑R=−22JIE|R|NE∑j=1NEEjQ+R+JIINI∑j=1NIIjQ+elI , Q ) ( 2 ) dxiQdt=1−xiQτrecE−UxiQEiQ ( 3 ) dylQdt=1−ylQτrecI−UylQIlQ ( 4 ) dziQ , fdt=1−ziQ , fτrecs−UsziQ , fsfTiQ , f ( 5 ) The units of our rate equations ( Eqs 1 and 2 ) and rate variables ( E , I , e , s ) are 1/s . In the text and figures we specify the rate as Spikes/s for clarity . J represents synaptic efficacy and has values depending on the type of connection and the distance over which it is made: EE denotes an excitatory-to-excitatory connection , IE an excitatory-to-inhibitory connection , II an inhibitory-to-inhibitory connection and EI an inhibitory-to-excitatory connection; the distance of the connection , in columns , is specified by a superscript R ( R = 0 represents intra-columnar connections ) . U represents the fraction of utilization , i . e . how much of the available synaptic resources is utilized upon an action potential reaching the presynaptic terminal . Us is the fraction of utilization of the ThC synapses . τE and τI are the membrane time-constants of excitatory and inhibitory neurons , respectively . τref is the refractory period of excitatory or inhibitory neurons , as specified by the superscript E or I . τrec represents the time-constant for recovery of available synaptic resources in excitatory , inhibitory or ThC synapses , as specified by the superscript ( E , I or s , respectively ) . Eqs ( 1 ) and ( 2 ) describe the dynamical behavior of the excitatory and inhibitory neurons , respectively . These are rate equations–they describe the output rates of the neurons as a non-linear gain function of their summed inputs . The excitatory neurons receive intracortical excitation and inhibition ( first two terms of the right-hand side of Eq 1 ) as well as direct excitatory thalamic input ( the last term of the right-hand side of Eq 1 ) . For simplicity , inhibitory neurons receive only intracortical excitation and inhibition , with no direct thalamocortical input . The lack of direct thalamo-cortical input to the inhibitory neurons in the model is discussed in length and justified in the main paper . Most importantly for the special behavior of this model , the synapses impinging on excitatory neurons show depression . The model of synaptic depression implemented here consists of resource depletion followed by exponential recovery ( Eqs 3–5 ) . Each input class on the excitatory neurons has its own separate dynamical process of resource depletion and recovery . The gain function used in the firing-rate equations is defined as: g ( w ) ={0w<0w0≤w<EmaxEmaxw≥Emax ( 6 ) Where w is the sum of inputs to a neuron , as detailed in Eqs 1 & 2 . In all of our simulations , Εmax was set to 300 spikes/s . The implementation of a maximum firing-rate follows [28] . This value for Εmax is close to the effective maximum firing-rate that is set by the refractory period ( τref ) when its value is set to 3 ms ( as we do here for both excitatory and inhibitory neurons; see list of values below ) . Our model neurons sample the frequency axis in channels ( f ) that are ordered from 1 to M . In almost all types of protocols , M was equal to the number of columns and so f corresponds to Q ( i . e . there is one channel per cortical column ) . The exceptions to this were the Diverse Narrow protocol , which required denser frequency channels , and the tests for hyperacuity ( see below and in the Results section ) . The input mediated by each ThC fiber is represented by a firing-rate variable sf . Its magnitude at each time-point depends on the maximum amplitude of the stimulus ( A , measured in spikes/s ) and the temporal envelope of stimuli presented in that channel ( ξf ( t ) , a fraction between 0 and 1 ) : sf=ξf ( t ) ⋅A ( 7 ) In all the stimuli presented in this work , ξf ( t ) had the form of a trapezoid pulse: a 5-ms ascending linear ramp from 0 to 1 , a period of constant amplitude , and a 5-ms linear ramp descending back to 0 , all adding up to the nominal stimulus duration . TiQ , f specifies the relative amplitude at which an excitatory neuron receives input from frequency channel f compared to its best frequency ( BF ) . The values of TiQ , f over all channels make up the tuning curve of the neuron’s thalamic input . Each neuron was assigned a triangular-shaped input tuning curve . This shape was chosen in order to avoid having input to all columns upon each sensory stimulation , as was the case with the tuning curves used by Loebel et al . [28] that decayed exponentially from their peak at the best frequency . The width of the tuning curve is determined by a localization parameter λ , which scales the tuning curve according to the frequency separation between the columns . The equation for the input tuning curve of a typical neuron of column Q , i . e . with best frequency fBF = Q , is: TtypicalQ , f={0f<Q−λ1− ( Q−f ) /λQ−λ≤f<Q1− ( f−Q ) /λQ≤f<Q+λ0f≥Q+λ ( 8 ) The value of λ = 5 used in our simulations confers upon the neurons in our model tuning widths of about 2 octaves ( cf . Fig 1D ) , similar to the tuning of suprathreshold activity in rat A1 [40] . Recently , it has been shown that neurons within a local population of mouse A1 have heterogeneous best frequencies [46 , 47] . In line with these results , each column in our model contained neurons with several different best frequencies: Approximately 1/16 , 1/8 , 1/8 and 1/16 of the neurons in each column had their tuning-curves shifted by -2 , -1 , +1 and +2 , respectively , relative to the typical fBF = Q . Neurons with shifted input tuning-curves were selected randomly ( Fig 1A illustrates the resulting tonotopy; the distribution is adjusted for illustration purposes ) . The ThC input to all neurons that had a firing-rate of 0 spikes/s when there was no sound input ( “non-active” neurons ) was zero , as proposed by Loebel et al . [28] . Calcium imaging studies in mice [47 , 72] suggest that the heterogeneity of tuning-curves implemented in our model is intermediate between those of layer 4 and layers 2/3 in mouse A1 . We note that the randomization of tuning-curves was not done in order to generate independent samples , but in order to better model the physiology . We do not compare between networks with different randomized tuning-curves . Rather , almost all of our analyses were made on column averages . Our simulations of networks with homogeneous vs . heterogeneous columns showed that heterogeneity of the tuning curves slightly extended the parameter range that allowed stimulus-specific adaptation , but its effect was very small ( cf . Fig 7B ) . All simulations of this model used networks with 21 columns . The values we used for the different parameters were: NE , NI=100;λ=5;U=0 . 5;Us=0 . 7τE , τI=1×10−3s;τrefE , τrefI=3×10−3s;τrecE , τrecI=0 . 8s;τrecs=0 . 3sJEE0=6;JIE0=0 . 5;JEI=−4;JII=−0 . 5;JEE1=4 . 5×10−2;JIE1=3 . 5×10−3;JEE2=1 . 5×10−2 , JIE2=1 . 5×10−3 Values of e for both populations were distributed uniformly between -10 and 10 spikes/s . The recovery time constants for both excitatory and inhibitory neurons are referred to as τrec , since their values were identical in all of our simulations . The values above are the same as in Loebel et al . [28] . Some were based on experimental studies , while others were tuned to obtain approximate balance of excitation and inhibition , a spontaneous firing-rate of a few spikes/s , and no occurrence of spontaneous PSs . All simulations were run using a time step dt = 1 x 10−4 s . Stimuli are called “tones” here because they were set to simulate pure-tone stimulation . Each was presented in one frequency channel f . Since the behavior of the model in response to sequences of stimuli was of major interest here , tones were presented in sequences consisting of 100 stimuli . The number of different tones was determined by the protocol type [4 , 12] . In the oddball protocol there were 2 tones , the lower at frequency f1 and the higher at frequency f2 , with a frequency separation Δf = f2 –f1 . These tones were centered on the best frequency of the middle column of the network ( column 11 ) , using Δf = 2 unless specified otherwise . The deviant tone had a probability of occurrence P ( P = 10% unless specified otherwise ) . The other tone , called the standard , occurred at a probability of 1 – P . Tone sequences were generated by a random permutation of a block of standard tones followed by a block of deviant tones , in a way that allowed deviants to occur close apart , or even one after another . Responses to the two conditions were calculated as averages over all occurrences , including deviant responses that were weaker due to recent deviants ( cf . Fig 6B ) . This randomization is in line with the practice in experimental studies by our group . We do not treat the short-term effects of deviants on one another . For experimental results of this effect , see [43 , 73] . The oddball protocol was always run twice , first with f2 as deviant and f1 as standard and then with their roles reversed ( Deviant f2 and Deviant f1; Fig 3F ) . The Equal protocol was similar to the oddball but had P = 50% . In the Deviant Alone protocol , all presentations of the standard tone were replaced with silent trials . We also used two sequences of “deviant-among-many-standards” , in which we presented 10 tones spaced evenly along the frequency scale ( including f1 and f2 ) , each occurring at a probability of 10% . The Diverse Narrow protocol had two tones lower than f1 , four between f1 and f2 , and two higher than f2 . This required a denser sampling of the frequency axis , as the spacing between tones had to be Δf/5 ( cf . Fig 4A ) . In the Diverse Broad protocol we used 10 tones with a spacing of Δf = 2 , which is the same as in the oddball protocol . Four tones were below f1 and four were above f2 ( cf . Fig 4A ) . Individual stimuli had a duration of 50 ms , including onset and offset linear ramps of 5 ms . Inter-stimulus interval ( ISI ) was defined as the time between the onsets of consecutive stimuli . We used ISI = 350 ms , A = 5 spikes/s unless specified otherwise . Responses to stimuli were generally evaluated based on the mean firing-rate in the middle column , E , unless explicitly stated otherwise . In Figs 1C , 3B , 3D , 4G , 8B , 9C , 10A and 10B we present and analyze responses of other columns as well . For each stimulus , a spike-count was calculated by correcting E to the baseline ( defined as the average of E during the 5 ms before stimulus onset ) and then integrating it from stimulus onset to 45 ms after offset . Since this is a firing-rate model that has no sources of noise , a period as short as 5 ms is valid as a baseline . Its advantage is that it allowed us to calculate of the response over 45 ms even for the shortest ISIs used . The response to each tone was defined as the average spike-count for all presentations of that tone . In experimental studies of SSA the baseline is usually not subtracted ( but see [12] ) . However , it should have little effect over the response sizes and strength of the SSA , since in general our simulations showed low spontaneous rates and very large responses . We quantified the strength of SSA in a way that is becoming standard [1 , 2 , 4 , 8 , 9 , 12 , 16 , 60 , 61] . The responses in both conditions of the oddball protocol ( Deviant f2 , Deviant f1 ) were calculated and used to compute the Common-Contrast SSA Index ( CSI ) : CSI=d ( f1 ) +d ( f2 ) −s ( f1 ) −s ( f2 ) d ( f1 ) +d ( f2 ) +s ( f1 ) +s ( f2 ) ( 9 ) Where d ( f1 ) and s ( f1 ) ( d ( f2 ) and s ( f2 ) ) are the responses to f1 ( f2 ) when it was used as deviant and standard , respectively . The dynamics of resources in each thalamocortical synapse ( Eq 5 ) is a one-dimensional system , independent of the other variables in the network . The amount of resources at stimulus onsets or offsets during a train of standard stimuli can be treated as an iterated map , given that the stimulus duration and inter-stimulus intervals are constant throughout the stimulus protocol . For simplicity , we rewrite Eq 5 as: dziQ , fdt=1τrecs− ( 1τrecs+UssfTiQ , f ) ziQ , f ( 10 ) When no stimulus is presented , the solution of this equation is: ziQ , f ( t ) =1−[1−ziQ , f ( toffset ) ]exp ( −t−toffsetτrecs ) ( 11 ) Where toffset is the time when the last stimulus ended . For the period before the first stimulus in a protocol , we assume the system is in its steady-state and therefore ziQ , f = 1 for all synapses . Under stimulus presentation , the solution of Eq 10 is: ziQ , f ( t ) =11+τrecsUssfTiQ , f−[ziQ , f ( tonset ) −11+τrecsUssfTiQ , f]exp[− ( 1+τrecsUssfTiQ , f ) t−tonsetτrecs] ( 12 ) Where tonset is the onset of the current stimulus . Since we are interested only in the fractions of synaptic resources at stimulus onsets or offsets , we substitute the relevant time periods and get the following relations between the values of z for consecutive onset and offset times: zonset=1−kisi ( 1−zoffset ) ( 13 ) zoffset=z˜ss−kdur ( zonset−z˜ss ) ( 14 ) Where: zonset≡ziQ , f ( tonset ) ;zoffset≡ziQ , f ( toffset ) ;z˜ss≡11+τrecsUssfTiQ , f;kisi≡exp ( −tonset−toffsetτrecs ) ;kdur≡exp ( −toffset−tonsetτrecsz˜ss ) Substituting Eq 13 in Eq 14 , and vice versa , gives the two iterated maps: zonset ( n+1 ) =1−kisi[1−z˜ss ( 1−kdur ) ]+kdurkisizonset ( n ) ( 15 ) zoffset ( n+1 ) =z˜ss ( 1−kdur ) +kdur ( 1−kisi ) +kdurkisizoffset ( n ) ( 16 ) Which have the joint stable fixed points: z*onset=1−kisi[1−z˜ss ( 1−kdur ) ]1−kdurkisi ( 17 ) z*offset=z˜ss ( 1−kdur ) +kdur ( 1−kisi ) 1−kdurkisi ( 18 ) Eqs 17 and 18 are the expressions for the onset and offset values of z reached after enough identical stimuli were presented . These values were used in Fig 9C and 9D , as the approximate steady-state resources in synapses conveying the standard stimuli from the thalamus to the standard column ( the column whose best frequency is the standard stimulus ) . They are correct when all neurons in the column have the same best frequency , when the stimulus sequence consists of standard tones only , and when stimulation is on-off only ( no ramp in the stimulus envelope , i . e . ξf ( t ) consists only of square pulses ) . The MATLAB scripts for running our model and the data replotted from [4 , 12 , 45] are available online as supporting information . | We present a possible mechanism for the way auditory cortex emphasizes stimuli that are deviant within a regular , repetitive sequence . This enhancement is strong and widespread in auditory cortex , but not in its major thalamic input , the ventral division of the medial geniculate body . In contrast with previous models , which are based on depression of the synapses that convey the input to the cortex , here the network structure and the known dynamics of intracortical synapses play a key role . The model accounts better than previous models for available experimental data , and provides testable predictions that differentiate it from feedforward models . It is a useful starting point for studying the circuit mechanisms that underlie cortical responses to unexpected stimuli . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] | [
"auditory",
"cortex",
"medicine",
"and",
"health",
"sciences",
"action",
"potentials",
"neural",
"networks",
"nervous",
"system",
"membrane",
"potential",
"brain",
"electrophysiology",
"neuroscience",
"computational",
"neuroscience",
"mood",
"disorders",
"neuronal",
"tuning",
"sensory",
"physiology",
"computer",
"and",
"information",
"sciences",
"depression",
"animal",
"cells",
"mental",
"health",
"and",
"psychiatry",
"auditory",
"system",
"cellular",
"neuroscience",
"cell",
"biology",
"anatomy",
"synapses",
"physiology",
"neurons",
"single",
"neuron",
"function",
"biology",
"and",
"life",
"sciences",
"cellular",
"types",
"sensory",
"systems",
"computational",
"biology",
"neurophysiology"
] | 2017 | Stimulus-specific adaptation in a recurrent network model of primary auditory cortex |
Glioblastoma , the most common primary malignant brain tumor , is incurable with current therapies . Genetic and molecular analyses demonstrate that glioblastomas frequently display mutations that activate receptor tyrosine kinase ( RTK ) and Pi-3 kinase ( PI3K ) signaling pathways . In Drosophila melanogaster , activation of RTK and PI3K pathways in glial progenitor cells creates malignant neoplastic glial tumors that display many features of human glioblastoma . In both human and Drosophila , activation of the RTK and PI3K pathways stimulates Akt signaling along with other as-yet-unknown changes that drive oncogenesis . We used this Drosophila glioblastoma model to perform a kinome-wide genetic screen for new genes required for RTK- and PI3K-dependent neoplastic transformation . Human orthologs of novel kinases uncovered by these screens were functionally assessed in mammalian glioblastoma models and human tumors . Our results revealed that the atypical kinases RIOK1 and RIOK2 are overexpressed in glioblastoma cells in an Akt-dependent manner . Moreover , we found that overexpressed RIOK2 formed a complex with RIOK1 , mTor , and mTor-complex-2 components , and that overexpressed RIOK2 upregulated Akt signaling and promoted tumorigenesis in murine astrocytes . Conversely , reduced expression of RIOK1 or RIOK2 disrupted Akt signaling and caused cell cycle exit , apoptosis , and chemosensitivity in glioblastoma cells by inducing p53 activity through the RpL11-dependent ribosomal stress checkpoint . These results imply that , in glioblastoma cells , constitutive Akt signaling drives RIO kinase overexpression , which creates a feedforward loop that promotes and maintains oncogenic Akt activity through stimulation of mTor signaling . Further study of the RIO kinases as well as other kinases identified in our Drosophila screen may reveal new insights into defects underlying glioblastoma and related cancers and may reveal new therapeutic opportunities for these cancers .
Glioblastoma ( GBM ) , the most common primary malignant brain tumor , infiltrates the brain , grows rapidly , and is refractory to current therapies . Signature genetic lesions in GBM include amplification , mutation , and/or overexpression of receptor tyrosine kinases ( RTKs ) , such as EGFR and PDGFRα , as well as activating mutations in components of the PI-3 kinase ( PI3K ) pathway ( reviewed in [1] ) . More than 40% of GBMs show EGFR gene amplification , and these amplification events are often accompanied by mutations in EGFR [1] . The most prevalent mutant form of EGFR is ΔEGFR ( EGFRvIII , de2-7EGFR , EGFR* ) , an intragenic truncation mutant that displays constitutive kinase activity [2] . ΔEGFR and other constitutively active mutant forms of EGFR found in GBMs potently drive tumor cell survival , migration , and proliferation [2] , [3] . The most frequent mutation in the PI3K pathway in GBM is loss of the PTEN lipid phosphatase , which results in unopposed signaling through PI3K and robust stimulation of Akt , especially in the context of EGFR activation [1] . In mouse models , co-activation of these pathways in glia , glial progenitor cells , and/or neuro-glial stem cells induces GBM [4] , [5] , [6] , [7] . However , the full range of signaling events acting downstream of or in combination with EGFR and PI3K to drive oncogenesis remain to be determined . While several normal effectors of RTK and PI3K signaling , such as Ras , Akt , and mTor , are used by EGFR and PI3K in GBM and are required for gliomagenesis [1] , constitutive activation of RTK and PI3K pathways may evoke changes distinct from those induced by normal developmental signaling . Notably , treatments with pharmacologic inhibitors of EGFR or mTor are cytostatic at best in a subset of patients , indicating that other , unidentified factors or compensatory signals affect the survival and growth of tumor cells [8] . To uncover new factors required for EGFR- and PI3K- mediated gliomagenesis , we developed a GBM model in Drosophila melanogaster [9] . Drosophila offers several advantages for modeling cancers like GBM . Flies have orthologs for 75% of human disease genes [10] , including nearly all known gliomagenic genes; signaling pathways are highly conserved; versatile genetic tools are available for cell-type specific gene manipulation [11] , ; and Drosophila neural cell types are homologous to their mammalian counterparts [13] , [14] . While a Drosophila model cannot address all aspects of human GBM , our model recapitulates important pathologic features . Specifically , constitutive activation of EGFR-Ras and PI3K signaling in Drosophila glial progenitor cells gives rise to proliferative , invasive neoplastic glia that create transplantable malignant tumors [9] . These tumors are induced through activation of a synergistic genetic network composed of downstream pathways commonly mutated and/or activated in human GBMs , such as Akt and mTor signaling [9] . However , activating these known downstream pathways alone or in combination is not sufficient to induce glial neoplasia in Drosophila , indicating that additional , as yet unidentified , genetic pathways are involved in transformation . Thus , we undertook genetic screens using our Drosophila GBM model to discover new genes underlying EGFR and PI3K mediated neoplastic transformation , and tested whether human orthologs of the genes identified in Drosophila represent new human genes involved in GBM . Our analyses in both Drosophila and human systems uncovered that the RIOK1 and RIOK2 kinases drive the survival and proliferation of GBM cells . RIOK1 and RIOK2 are members of the RIO ( right open reading frame ) family of atypical protein kinases , named for yeast ( S . cerevisiae ) Rio1p and Rio2p , respectively [15] . The RIOK1 and RIOK2 proteins are highly conserved , and are present in all phylogenetic kingdoms , from yeast to mammals among eukaryotes . Kinases in this family are characterized by the presence of the RIO kinase domain , a kinase fold structurally homologous to eukaryotic serine-threonine protein kinase domains , but that lacks classic activation and substrate binding loops ( reviewed in [15] ) . While these kinases undergo autophosphorylation and phosphorylate nonspecific substrates in vitro , the actual in vivo substrates of RIO kinases are unknown [15] . In both yeast and human cells , RIOK1 and RIOK2 , which are not functionally redundant , are required for processing of the 18S rRNA and cytoplasmic maturation of the 40S ribosomal subunit , although neither kinase is an integral component of the ribosome [16] , [17] , [18] , [19] . Recent studies demonstrate that , in yeast , Rio2p also transiently associates with immature ribosomes to block translation initiation , although how RIOK2 is regulated in this context is unclear [20] . To date , several studies have provided suggestive evidence that the RIO kinases could be involved in RTK and PI3K signaling: RIOK2 becomes rapidly phosphorylated in response to EGFR stimulation; Rio2p binds to Tor2p , an ortholog of the mTor kinase , and RIOK1 is required for the proliferation and survival of Ras-dependent cancer cells [21] , [22] , [23] . However , to date , no specific function has been ascribed to RIOK1 or RIOK2 in the context of EGFR or PI3K signaling . In this manuscript we demonstrate that RIOK1 and RIOK2 become overexpressed in GBM tumor cells relative to normal brain cells; that RIOK1 and RIOK2 overexpression occurs in response to constitutive Akt signaling; that RIOK2 forms a complex with RIOK1 , mTor , and other signaling components to drive activation of Akt signaling and tumorigenesis; and that , in GBM cells , RIOK1 or RIOK2 loss causes a reduction in Akt signaling and provokes p53-dependent apoptosis , cell cycle exit , and chemosensitivity through the RpL11-dependent ribosomal stress checkpoint . Our data demonstrate that the RIO kinases play a key role in Akt-mediated transformation of GBM cells .
To discover new genes involved in glial pathogenesis , we performed a genetic screen using our Drosophila GBM model . Co-overexpression of constitutively active forms of Drosophila EGFR ( dEGFRλ ) and the PI3K catalytic subunit p110α ( dp110CAAX ) stimulates malignant transformation of post-embryonic larval glia , inducing lethal glial neoplasia ( Figure 1A–1C ) ( characterized in detail in [9] ) . Using this larval Drosophila GBM model , we performed an RNAi-based modifier screen for genes that suppress ( inhibit ) or enhance ( worsen ) neoplastic phenotypes caused by constitutive EGFR and PI3K signaling . In this scheme , which is an enhancer-suppressor screen , modifier kinases that block/inhibit fly glial neoplasia when their expression is reduced are referred to as ‘suppressors , ’ and modifier kinases that exacerbate neoplasia when their expression is reduced are referred to as ‘enhancers . ’ This is in keeping with standard Drosophila nomenclature , in which genes are classified by their loss-of-function phenotypes . As a side-note , in this context , the term suppressor does not refer to mammalian tumor suppressors . To enrich for new pathway components , we screened nearly all of the kinases encoded in the Drosophila genome ( Table S1 ) . Our choice to screen kinases was based on four considerations: ( 1 ) kinases regulate a broad array of biological and cellular processes , including those underlying oncogenesis; ( 2 ) kinases are highly conserved between Drosophila and humans such that every Drosophila kinase has a clear human ortholog; ( 3 ) as a group , they are well characterized , facilitating functional analysis; and ( 4 ) drug discovery efforts are focused on development of specific kinase inhibitors . We tested 553 conditional RNAi constructs targeting 223 of the 243 kinases in the fly genome ( Tables S1 , S2 ) [24] , [25] . RNAi constructs were expressed specifically within the glial lineage and were tested for their phenotypic effects on proliferation , migration , morphology , and/or viability of dEGFRλ;dp110CAAX neoplastic glia . The specificity of modifier loci was confirmed by testing multiple RNAi constructs , dominant negative constructs , and/or available mutant alleles to determine if they produced analogous phenotypes in the dEGFRλ;dp110CAAX model ( Tables S2 , S3 ) . Control assays were performed for the effects of modifier RNAi constructs when expressed specifically in other cell types , including normal glia , neuroblasts ( Drosophila neural stem cells ) , and neurons in order to distinguish those RNAi constructs that caused non-specific toxicity from RNAi constructs that caused specific changes in neoplastic glia ( Table S4 , Text S1 ) . Constructs that caused early organismal lethality in all cellular contexts tested were excluded from analysis ( Tables S2 , S4 ) . To test whether modifiers act in oncogenic signaling downstream of specific EGFR mutations found in human GBM , we created flies that overexpress human ΔEGFR . Glial-specific expression of ΔEGFR caused lethal glial neoplasia phenotypes , alone and in combination with dp110CAAX , that were similar to dEGFRλ , and these phenotypes required EGFR kinase activity and core PI3K effectors , such as dAkt ( Figure 1A–1F , Figure S1 , Table S5 ) . We tested modifier RNAi constructs for the ability to alter glial-specific ΔEGFR and ΔEGFR;dp110CAAX phenotypes ( Table S5 ) ; the results mirrored their genetic interactions with dEGFRλ;dp110CAAX ( Table S2 , S5 ) , indicating that modifiers identified in the screen are common to neoplastic phenotypes conferred by both Drosophila and human EGFR . We identified a total of 45 modifier genes ( Figure 1I , Tables S2 and S3 ) . Suppressor RNAi constructs targeting 39 genes reduced neoplasia and induced smaller brain size and lower glial cell numbers relative to dEGFRλ;dp110CAAX controls , whereas enhancer RNAi constructs targeting 6 genes worsened tumorigenesis and neoplasia and induced increased glial cell numbers , and/or aberrant glial morphologies ( Figure 1A–1D , 1I , Tables S2 and S3 ) . Modifiers that suppressed glial neoplasia include genes identified in previous studies , including dAkt [9] ( Tables S2 , S5 ) . A small subset of suppressor constructs caused strong phenotypes in the context of constitutive EGFR-PI3K signaling . Constructs targeting three modifier loci , Raf , Src42A , and Taf1 , rescued dEGFRλ;dp110CAAX animals to adult viability , allowing neoplastic glia to differentiate and function normally despite the presence of dEGFRλ and dp110CAAX ( Table S2 ) . Constructs targeting two of the strongest modifiers , CG11660 and CG11859 ( the Drosophila orthologs of the RIOK1 and RIOK2 kinases , respectively ) , caused severe reduction in brain size and glial cell number when combined with dEGFRλ;dp110CAAX , ΔEGFR;dp110CAAX , and/or ΔEGFR , as compared to wild-type control animals ( Figure 1A–1H , Figure S1 ) . dRIOK2 knockdown gave a stronger effect ( Figure 1D , 1G ) . In contrast , RNAi constructs targeting dRIOK1 and dRIOK2 did not produce a dramatic growth reduction when targeted to normal glia ( Figure 1E , Table S4 ) , indicating that dRIOK1 or dRIOK2 knockdown does not simply cause nonspecific cellular toxicity . Modifier kinases were classified by bioinformatic annotations using Flybase , Gene Ontology , KEGG pathways , the STRING database [26] , and comparisons with other Drosophila RNAi screens . These classifications ( Figure 1J , Figure S2 , Table S6 ) show that kinases with core functions in the RTK and PI3K pathways were highly represented , validating our model and screening methodology [27] , [28] , [29] . Notably , very few of our modifiers have emerged from Drosophila RNAi screens for cell viability ( Table S7 ) [30] , indicating that most of the modifiers are not generically required for cell survival . The largest group of modifiers have roles in cell proliferation ( Figure 1J , Table S6 ) , and many of these yielded reduced cell lineages upon knockdown in neuroblasts ( Table S4 ) [31] , consistent with known requirements for RTK and PI3K signaling in neural progenitor cells [9] , [32] . Several of the modifiers involved in cell proliferation , such as warts ( Table S2 ) , are also components of the hippo pathway , a pathway with a documented role in glial cell proliferation [33] . Broad comparisons with orthologs from species such as yeast ( S . cerevisiae ) , revealed modifiers kinases implicated in protein translation , such as dRIOK1 and dRIOK2 [16] , [20] , or in cell shape change and migration . Finally , comparison to human kinases shows that the majority of our modifier kinases have orthologs previously implicated in GBM ( Figure 1J , Figure S2 , Table S7 ) , leaving a set of 16 novel modifiers , including dRIOK1 and dRIOK2 . Novel modifier kinases identified in our Drosophila screens may represent human kinases directly involved in GBM pathogenesis . Kinases that block fly glial neoplasia when their expression is reduced are of interest because their human orthologs may be promising new targets for therapeutic inhibition . There are 27 human orthologs for the 16 novel Drosophila modifier kinases ( Table S8 ) . To determine if any of these human kinases are expressed or mutated in GBM , we analyzed tumor genomic databases , proteomic atlases , and GBM cell lines , which provided suggestive evidence that 12 modifier orthologs are subject to genetic alteration and/or elevated gene or protein expression in GBMs ( see Text S1 , Tables S9 and S10 , Figures S3 and S4 ) . Among these , RIOK1 and RIOK2 showed increased protein expression consistent with involvement in GBM . Given that loss of the dRIOKs strongly and specifically blocks growth and survival of EGFR and PI3K mutant glia , and that recent publications suggest that the RIO kinases may contribute to EGFR and/or mTor signaling [21] , [22] , the functional roles of RIOK1 and RIOK2 in GBM were of particular interest . A range of GBM cells and cell lines were examined to determine how RIOK1 and RIOK2 expression correlated with tumor cell genotype and phenotype . Our analyses showed that RIOK1 and RIOK2 were expressed in PTEN-null U87MG GBM cells and were upregulated in U87MG cells engineered to express ΔEGFR at levels detected in tumors [34] ( Figure 2A ) . RIOK1 and RIOK2 were also upregulated in GBM tumors with EGFR overexpression/mutation as well as activated Akt ( Figure 2B ) , although these correlations were not clear in all specimens . Primary neurosphere cultures , which are composed of neural stem cell-like human GBM cells propagated in EGF-supplemented media [35] , [36] , [37] , can maintain mutations/gene expression found in their parent tumors [37] . Neurosphere cultures showed strong RIOK1 and RIOK2 expression ( Figure 2C ) ; these included neurosphere lines with ΔEGFR , as well as neurosphere lines displaying PDGFRα overexpression , PTEN loss , and/or other mutant forms of EGFR ( Figure 2C ) . In a panel of standard GBM cell lines , RIOK1 and RIOK2 showed strong expression in cell lines known to harbor PTEN and/or EGFR mutations , and RIOK1 and RIOK2 expression was comparatively lower in a GBM cell line with intact PTEN ( Figure S5 ) [38] , [39] . In contrast , RIOK1 and RIOK2 were nearly undetectable in mixed glial cultures freshly derived from adult human cortex ( Figure 2D ) . Thus , RIOK1 and RIOK2 overexpression appeared to be correlated with RTK mutation/overexpression and/or PTEN loss in GBM tumor cells . To determine if elevated RIOK1 and/or RIOK2 expression in GBM cells depends on EGFR and/or PI3K signaling , neurosphere lines and U87MG-ΔEGFR cells were treated with relevant inhibitors ( Figure 2E and 2G , Figure S6 ) . RIOK1 and RIOK2 levels decreased upon either growth factor withdrawal or gefitinib treatment of primary neurosphere cultures ( Figure 2E ) , indicating that their up-regulation can be EGFR-dependent . Consistent with this , Pten−/−; Ink4a/arf−/− mouse astrocytes transformed by ΔEGFR showed increased RIOK1 and RIOK2 levels , which were reduced by gefitinib treatment ( Figure 2F ) . RIOK1 and RIOK2 protein levels also decreased in neurosphere cells and U87MG-ΔEGFR cells treated with inhibitors of the p110 PI3K catalytic subunit , such as BEZ-235 , and inhibitors of Akt , such as A443654 ( Figure 2G ) [40] . siRNA-mediated Akt knockdown or restoration of PTEN function in U87MG-ΔEGFR cells also reduced RIOK protein levels ( Figure 2H , Figure S7 ) . Treatments with p110 and Akt inhibitors also demonstrated that p110 and Akt signaling is required for RIOK1 and RIOK2 expression in PDGFRα-overexpressing neurospheres ( Figure 2G ) . Taken together , these data indicate that RIOK1 and RIOK2 overexpression in GBM cells is driven by Akt activity downstream of RTK mutation/overexpression and/or PTEN loss . However , the role of factors that act downstream of Akt was less clear ( mTor inhibitors did not always reduce RIOK levels , see Figure S6 ) , suggesting that RIOK1 or RIOK2 levels may be directly regulated by Akt . Given that mRNA expression levels of RIOK1 and RIOK2 did not show significant upregulation in tumor samples with PTEN and/or EGFR alterations ( Table S9 ) , and given that RIOK1 and/or RIOK2 levels decline after short-term treatments with Akt inhibitors ( Figure S7 ) , we hypothesized that Akt signaling may regulate RIOK2 and/or RIOK1 levels by modulating protein stability post-translationally . Consistent with this , addition of a proteosome inhibitor , MG132 , prevented the reduction in RIOK1 and/or RIOK2 protein levels observed upon A443654 treatment or PTEN add-back ( Figure S7 ) . RIOK2 has several mapped serine phosphorylation sites ( www . phosphosite . org ) , including putative Akt target sites ( Figure S8 ) . However , mutation of this single site did not abolish detection of RIOK2 by a phospho-Akt-substrate antibody ( Figure S9 ) . Thus , Akt-mediated regulation of RIO kinase levels does not hinge on phosphorylation at a single residue in RIOK2 , and likely involves a more complex mechanism that requires more investigation . To confirm that the RIOKs are expressed in GBM , we performed immunohistochemistry ( IHC ) for RIOK2 on a group of typed tumor specimens ( RIOK1 antibodies were unsuitable for IHC ) . Xenograft specimens of GBM39 , which is ΔEGFR-positive [41] , showed RIOK2 expression in tumor cells , with cells displaying diffuse cytoplasmic and sub-surface RIOK2 localization ( Figure 3A ) . In contrast , murine stromal cells had little or no RIOK2 immunoreactivity , although the antibody can detect mouse RIOK2 . ΔEGFR-positive and EGFR-overexpressing specimens displayed strong , but sometimes heterogeneous , cytoplasmic RIOK2 immunoreactivity , ranging from the giant cell to the small cell populations ( Figure 3B–3E , Figures S10 and S11 ) , with the strongest expression in mitotic cells and densely cellular pseudopallisades ( Figure 3C and 3D , Figure S10 ) . Heterogeneity in RIOK2 expression may possibly reflect heterogeneity of RTK expression in tumors [42] , [43] , or may reflect upregulation of RIOK2 in actively cycling cells given the increased immunoreactivity observed in mitotic cells . Cytoplasmic localization of RIOK2 in GBM is consistent with observations of RIO kinase localization in yeast and human cells [16] , [17] , [18] . In contrast , RIOK2 did not show appreciable immunoreactivity in neural cells in matched normal control brain ( n = 14 ) ( Figure 3F , 3H ) , and did not show immunoreactivity in tumor stroma ( Figure 3B ) , demonstrating that RIOK2 upregulation is tumor-specific . Akt signaling in tumors was assessed by staining for Akt phosphorylated at Serine-473 ( Akt-S473-P , example shown in Figure 3I ) , and EGFR status of tumors was primarily assessed with staining for EGFR phosphorylated on Tyrosine-1068 ( EGFR-Y1068-P , example shown in Figure 3J ) , which indicates EGFR activation . Statistical analysis demonstrated that RIOK2 expression was significantly correlated with EGFR status in tumor specimens ( Figure 3L ) , although some EGFR-negative tumors also showed RIOK2 immunoreactivity ( Figure 3G , 3L ) , while some EGFR-negative tumors did not ( Figure 3K ) . The correlation of RIOK2 expression with EGFR activity is likely secondary to Akt-mediated regulation of RIOK2: RIOK2-expressing specimens positive EGFR-Y1068-P always showed staining for Akt-S473-P ( n = 20 ) . Indeed , all specimens that showed RIOK2 immunoreactivity , whether EGFR-positive or EGFR-negative , showed staining for Akt-S473-P ( Figure 3L ) . To determine if elevated levels of RIOK2 drives oncogenic changes in mammalian cells , we tested the effects of RIOK2 overexpression in astrocytes . Pten−/−; Ink4a/arf−/− murine astrocytes , which are immortalized by tumor suppressor mutations common in GBM , express little endogenous RIOK2 and are not gliomagenic in intracranial grafts assays [44] . In two experiments , mice intracranially grafted with Pten−/−; Ink4a/arf−/− astrocytes overexpressing RIOK2 showed symptoms of hydrocephaly and neurological deficits 3 weeks following implantation , unlike mice grafted with Pten−/−; Ink4a/arf−/− control astrocytes . Histological analysis showed that control Pten−/−; Ink4a/arf−/− astrocytes yielded no intracranial tumors in 12 total animals tested , whereas RIOK2overexpresion; Pten−/−; Ink4a/arf−/− astrocytes formed invasive high-grade glial tumors composed of invasive spindle-shaped cells in 7 out of 10 total animals tested ( p< . 001 by chi-squared test ) ( Figure 4A , 4B ) . In GBM , ΔEGFR drives strong Akt activation in the context of PTEN loss , and can drive gliomagenic transformation of Pten−/−; Ink4a/arf−/− astrocytes [6] , [45] . Thus , we wondered whether RIOK2 overexpression also drives astrocyte transformation by activating Akt . Consistent with this , RIOK2overexpression; Pten−/−; Ink4a/arf−/− astrocytes displayed increased phosphorylation of Akt at Serine-473 ( Figure 4C ) , and tumor tissue from RIOK2overexpression; Pten−/−; Ink4a/arf−/− cells showed specific staining for Akt-Ser473-P ( Figure 4D ) . Phosphorylation of Serine-473 is required for Akt activity towards select substrates such as FOXO3 [46] , a direct Akt substrate that governs GBM cell tumorigenicity [47] . FOXO3 also displayed increased phosphorylation RIOK2overexpression; Pten−/−; Ink4a/arf−/− astrocytes ( Figure 4C ) . However , we did not observe increased phosphorylation of all Akt targets , including PRAS40 , with RIOK2 overexpression , suggesting that the effect of RIOK2 on phosphorylation of Akt targets was selective to Serine-473 dependent substrates [46] . Akt is phosphorylated on Serine-473 by mTor-complex-2 ( TORC2 ) , a multi-protein complex composed of the mTor kinase and several other signaling components , including Rictor [46] , a protein that becomes elevated in glioblastomas that also drives gliomagenesis when overexpressed in astrocytic cells [48] , [49] . In yeast proteomic analyses , Rio2p has been shown to bind to Rio1p and Tor2 , the yeast mTor ortholog that forms the equivalent of TORC2 [22] . From human cells overexpressing RIOK2 , RIOK2 co-immunoprecipitated with RIOK1 , mTor , and Rictor , a protein which is definitive of the TORC2 complex ( Figure 4E ) [46] . mTor also associates with another complex , mTor-complex-1 ( TORC1 ) , which phosphorylates other mTor substrates , such as EIF-4E , and is composed of signaling components including the Raptor protein [46] . Raptor did not co-immunoprecipitate in the RIOK2-RIOK1-mTor-Rictor complex ( Figure 4E ) , suggesting that RIOK2 specifically associates with TORC2 . Taken together , these data suggest that RIOK2 directly binds to TORC2 to stimulate phosphorylation of Akt at Serine-473 and activation of Akt towards select substrates , such as FOXO3 , and that this process may directly involve RIOK1 recruitment . Given that knockdown of their Drosophila cognates yields growth reduction of neoplastic glia , we tested RNAi constructs targeting human orthologs of novel suppressor kinases for their requirement in GBM cell survival and proliferation ( Text S1 , Figure S12 ) . Among these , RIOK1 or RIOK2 knockdown yielded strong effects , inhibiting U87MG-ΔEGFR and U87MG proliferation ( Figure 5A , 5B ) . ΔEGFR-positive neurosphere cultures , such as GBM301 and GBM39 , also showed a pronounced apoptotic response to RIOK2 or RIOK1 knockdown , with RIOK2 loss yielding stronger effects ( Figure 5C , 5D ) . In U87MG cells , which are dependent on Akt signaling for growth [50] , RIOK1 or RIOK2 RNAi provoked G2 cell cycle arrest and reduced proliferation ( Figure 5A–5B ) . The phenotypes caused by RIOK1 and RIOK2 knockdown were observed in other GBM cell lines that are PTEN and/or EGFR mutant , such as A172 ( Figure 6 , Figure S14 , data not shown ) . Of note , RIOK2 knockdown typically triggered a reduction in RIOK1 expression , regardless of the RIOK2 RNAi constructs used , suggesting that RIOK2 regulates RIOK1 protein levels ( Figure 5D , Figure S13 , see also Figure 6 ) . Given that the RIO kinases have been found to stimulate ribosome maturation , we initially suspected that functional reduction of RIOK1 or RIOK2 may cause generic cellular toxicity . However , in testing multiple GBM cell lines that showed strong RIOK expression , we found that a subset of GBM cells were far less affected by knockdown of RIOK1 and RIOK2 . GBM6 , a ΔEGFR-positive neurosphere line , did not undergo apoptosis upon RIOK1 or RIOK2 knockdown ( Figure 5D ) . Moreover , LNZ308 cells , which are PTEN mutant [38] , did not show cell cycle defects or strong reduction of RIOK1 expression with RIOK2 knockdown ( Figure 5B ) . Both GBM6 and LNZ308 are also mutant or null for p53 , whereas GBM cells that show cell cycle defects and apoptosis upon RIOK loss , such as U87MG , are wild-type for p53 [38] , [41] . We observed a similar lack of apoptosis upon RIOK1/RIOK2 loss in other p53 mutant/null GBM cells , such as U373 ( data not shown ) . This implies that the survival and proliferation defects induced by RIOK1 and RIOK2 loss rely on p53 . Consistent with this , concomitant knockdown of p53 with RIOK1 or RIOK2 in U87MG cells blocks the apoptosis observed upon RIOK2 or RIOK1 knockdown alone ( Figure 5E ) . In human cells , RIOK1 and RIOK2 transiently associate with immature cytoplasmic 40S ribosomal subunits to promote their maturation and stimulate rRNA processing , like their yeast counterparts [18] , [19] , [22] , [51] . Defects in ribosome biogenesis and rRNA processing can activate a p53-dependent ribosomal-stress checkpoint to suppress growth and induce cell cycle arrest and apoptosis , a process that relies on p53 upregulation and transcriptional activation mediated by release of the RpL11 ribosomal protein ( reviewed in [52] , [53] ) . In U87MG cells and other GBM cell lines , RIOK1 or RIOK2 knockdown induced up-regulation of p53 and the p21 cdk inhibitor , a p53 transcriptional target ( Figure 5F , Figure S13 , S14 ) , and coincident knockdown of RpL11 and RIOK1 or RIOK2 blocked induction of p53 and p21 ( Figure 5F and Figure S14 ) . Therefore , RIOK1 or RIOK2 loss leads to p53 activation , which requires the p53-RpL11-dependent ribosomal stress checkpoint . p53 is often downregulated in GBM tumors and tumor cells with activated Akt , PTEN loss , and/or EGFR mutation/overexpression [54] . Yet , the majority of tumors with PTEN loss or EGFR mutation/amplification have intact p53 loci ( EGFR: 85–92% , PTEN: 92–95% ) [55] , [56] . Given that RIOK loss upregulates p53 levels , we tested whether knockdown of RIO kinases could potentiate the response of GBM cells to treatments with DNA-damaging agents such as doxorubicin , which cooperates with p53 to provoke apoptosis , [57] , [58] and temozolomide , which is a DNA alkylator used to treat GBM . In GBM cells wild-type for p53 , such as GBM301 and U87MG , knockdown of RIOK1 or RIOK2 potentiated apoptotic responses to doxorubicin and/or temozolomide ( Figure 6A–6B , Figure S15 ) . In contrast , cells mutant for p53 , such as LNZ308 cells , did not show apoptosis upon RIOK1 or RIOK2 knockdown and doxorubicin-temozolomide treatments ( Figure 6A ) . Therefore , inhibition of the RIO kinases chemosensitizes EGFR- and/or PTEN mutant GBM cells . Our results suggest that elevated p53 activity can potentiate elimination of EGFR and/or PTEN mutant GBM cells . One way to increase p53 levels and activity is with nutlin-3 , a small molecule which is know to cause cell cycle arrest and sensitivity to DNA-damaging agents in U87MG cells [59] . However , nutlin-3 did not provoke the same cell cycle defects observed with RIO kinase knockdown , despite inducing high levels of p53 and p21 ( Figure S16 ) . Thus , other changes induced by RIO kinase loss must contribute to cell cycle arrest and apoptosis . We tested for signaling alterations that occur upon RIOK1 or RIOK2 knockdown that would explain reduced proliferation and survival of GBM cells . The caspase inhibitor ZVAD was used to dampen apoptosis and thus preserve signaling defects . Compared to controls , RIOK1 or RIOK2 knockdown led to reduced phosphorylation of Akt at Serine-473 and reduced phosphorylation of Akt target proteins such as FOXO3 ( Figure 6C , Figure S17 ) . This occurred in both p53 wild-type and p53 mutant GBM cells ( Figure 6C , Figure S17 ) . Serine-473 is phosphorylated by Tor complex 2 , ( TORC2 ) [46] , and in yeast and human cancer cells , TORC2 phosphorylation of Akt is stimulated by mature ribosomes , which can bind to both TORC2 and Akt to mediate their interaction , and TORC2 activity is blocked by defects in ribosome biogenesis [60] . Given that RIOK1 and RIOK2 loss causes defects in 40S ribosome maturation [17] , [18] , [19] , and that we discovered that RIOK1 and RIOK2 bind to TORC2 components , we hypothesized that RIOK1 and RIOK2 knockdown interferes with TORC2 activity . Consistent with this , other readouts of TORC2 activity , such as phosphorylation and levels of NDRG1 [60] , were reduced ( Figure 6C , Figure S17 ) , demonstrating that TORC2 activity is downregulated by RIOK1 and RIOK2 loss . This is consistent with recent findings demonstrating a requirement for TORC2 signaling in Drosophila glial neoplasia as well as human GBM cells [9] , [48] , [49] . However , the effects of RIOK loss on Akt signaling were not limited to the TORC2-dependent substrates . Phosphorylation of other Akt substrates , such as PRAS40 and TSC2 , can also be reduced upon RIO kinase knockdown ( Figure 6C ) . Thus , RIOK1 and RIOK2 are necessary for Akt signaling in GBM cells . Over-all , our results strongly suggest that functional reduction of RIOK1 and RIOK2 results in loss of Akt activity and p53 activation to drive cell cycle arrest , chemosensitivity , and apoptosis in Akt-dependent GBM cells with intact p53 ( Figure 7 ) .
From a Drosophila genetic screen , we identified genes encoding 16 novel kinases that affect EGFR- and PI3K- dependent neoplastic glial transformation . We found that a subset of human orthologs for these novel kinases , including RIOK1 and RIOK2 , are subject to alterations in GBM . RIOK1 and RIOK2 , two related and highly conserved atypical kinases , become upregulated in an Akt-dependent manner in GBM cells . Our results show that RIOK2 forms a complex with RIOK1 and TORC2 signaling components , drives activation of TORC2-dependent Akt signaling , and stimulates glial tumorigenesis . Furthermore , we found that , in GBM cells , RIOK1 or RIOK2 loss causes a reduction in Akt signaling towards TORC2-depdendent targets and provokes p53-dependent apoptosis , cell cycle exit , and chemosensitivity . Thus , our loss-of-function and gain-of-function data imply that RIOK2 creates a feedforward loop that promotes and maintains Akt activity , and disruption of this loop is sufficient to trigger chemosensitivity and apoptosis in Akt-dependent GBM cells with intact p53 ( Figure 7 ) . Our results may have broad relevance to other cancers since RIOK2 is strongly expressed in a range of other more common tumor types associated with high Akt activity , such as breast and prostate cancers ( Figure S18 ) . Further study of the RIO kinases as well as other kinases identified in our Drosophila screen may reveal new insights into the signaling defects underlying GBM and related cancers . RIOK1 and RIOK2 upregulation was associated with Akt activity in both GBM tumor specimens and cultured cells , and our results show that Akt signaling regulates RIO kinase protein stability , although the exact mechanism by which Akt regulates RIO kinase levels remains undetermined . RIOK2 has several putative and mapped phosphorylation sites , including at least one putative Akt phosphorylation site ( www . phosphosite . org , Figure S8 ) . Other studies show that RIOK2 phosphorylation can be stimulated by EGFR , and can be carried out by Polo-like kinase 1 [21] , [61] , and perhaps these events contribute to Akt-mediated regulation of RIO kinase levels . Of note , though standard GBM cells lacking PTEN showed high levels of RIO kinase expression , non-transformed astrocytes lacking PTEN did not show high levels of endogenous RIO kinase protein expression relative to astrocytes with intact PTEN . Therefore , other factors present in GBM cells must also contribute to elevated RIO kinase levels . To date , published studies show that the RIO kinases act as ribosome assembly factors that transiently associate with the 40S subunit to promote ribosome maturation and translation initiation [17] , [18] , [20] . Given that mature ribosomes are required for TORC2 activation and Akt phosphorylation at Serine-473 [60] , disruption of Akt signaling upon RIOK knockdown may be a result of defective ribosome biogenesis caused by RIO kinase loss . However , the RIO kinases may have a much more direct role in promoting and maintaining Akt activity given that RIOK2 binds to RIOK1 and to components of the TORC2 complex , which is consistent with recent studies in yeast showing that Rio2p can bind to Tor2 [22] . Given that Rio2p is released from mature ribosomes in a regulated process [20] , it is possible that the reason mature ribosomes promote TORC2 signaling is because they release free cytoplasmic RIOK2 that then stimulates TORC2 assembly or activity . The specific interplay between the RIO kinases and mTor signaling , ribosome biogenesis , protein translation , and Akt signaling will require additional investigation in the context of both normal and abnormal PI3K and RTK signaling , and may involve other as yet undetermined factors . Although RIOK1 and RIOK2 loss can cause defects in ribosome maturation [17] , [18] , in GBM cells the effects of RIO loss are not generic and instead are genotype-specific: p53 null mutant GBM cells showed no major cell cycle defects or apoptosis upon loss of these kinases . This specificity is derived from p53 upregulation and activation induced by the RpL11 ribosomal protein in response to RIOK loss . In humans , activation of the RpL11-p53-dependent ribosomal-stress checkpoint is associated with diseases caused by ribosomal protein haploinsufficiency , such as Diamond-Blackfan anemia , which are characterized by stem and progenitor cell failure [52] , [53] . Similarly , in Drosophila , haploinsufficiency of genes that encode ribosomal proteins retards developmental cell proliferation [62] . Given that cancer cells share many properties with stem and progenitor cells , induction of the RpL11-p53 ribosomal stress checkpoint may prove useful to deplete cancer cells . Indeed , recent experimental evidence indicates that the RpL11-p53-dependent ribosomal stress checkpoint suppresses tumorigenesis in mouse cancer models [63] . Moreover , several chemotherapeutic drugs induce the ribosomal stress checkpoint [64] , [65] . However , many of these drugs have deleterious effects unrelated to ribosomal stress , limiting their use . More specific induction of the ribosomal stress checkpoint , perhaps through RIO kinase inhibition , may prove therapeutically useful for GBM as well as related cancers . The importance of RIO kinases in cancer cell survival has been validated in independent studies . RIOK2 was recently identified in an RNAi-based screen for kinases that are required for survival of glioma stem-like cells , which confirms our results , although the functionality of RIOK2 in glioma was not explored [66] . In addition , RIOK1 was identified in a cell-based RNAi screen for genes required for Ras-mediated cell survival , although the functionality of RIOK1 was not explored in this study [23] . Of note , almost all other published cell culture-based RNAi screens in GBM cell lines did not pick up RIOK1 or RIOK2 because these screens were not designed to distinguish between kinases that were required for genotypic-specific survival or growth of GBM cell lines , and instead focused on kinases that showed a common requirement in all glioblastoma cell lines tested , be they mutant or wild-type for p53 , EGFR , or PTEN [67] , [68] , [69] . Our results , which are derived from independent multidisciplinary assays , are the first to establish functional connections between the RIO kinases , oncogenesis , Akt signaling , and the RpL11-p53-dependent ribosomal stress checkpoint ( Figure 7 ) . We envision that RIOK loss-of-function phenotypes in GBM cells are due to the combined effects of Akt inhibition and p53 induction , which together stimulate apoptosis and cell cycle exit of EGFR- and PTEN- mutant GBM cells , which share a common dependence on Akt signaling ( Figure 7 ) . Further studies to address the mechanisms by which the RIO kinases modulate Akt and p53 activity may lead to important new insights into the interactions between both of these pathways in both normal and cancer cells .
Flies were cultured at 25°C unless otherwise noted . Genotypes were established by standard genetics . Larval brain phenotypes were assessed and imaged as previously described [9] . Stocks were obtained from VDRC , NIG , and Bloomington stock centers ( Table S1 ) . wor-Gal4 lines were from C . Doe . To create UAS-ΔEGFR constructs , a full-length human ΔEGFR cDNA was cloned into pUAS-T , and fly stocks with stable insertions were created . The screen was based on crosses ( see Text S1 for genetic methodology ) that generated progeny containing a single RNAi construct exclusively expressed in GFP-labeled glia along with dEGFRλ and dp110CAAX . Transgenes were overexpressed using the glial-specific repo-Gal4 transcriptional driver . Screening was performed using fluorescence microscopy to visualize GFP-labeled glia in living larvae , and phenotypes were confirmed with confocal microscopy . Each positive-scoring RNAi construct was tested at least twice . Positive scoring RNAi constructs were also tested in wild-type glia , neuroblasts , and neurons ( Text S1 , Table S4 ) Established primary neurosphere cultures ( gifts of H . Kornblum ) were maintained as previously described in DMEM/F12 medium supplemented with bFGF and EGF [35] , [41] . Neurosphere cultures of GBM39 and GBM6 were created from serial xenografts of human GBMs ( gifts of C . D . James ) . Cultured normal human glia were derived from a fresh surgical specimen of normal human cortex ( gift of J . Olson ) procured under a protocol approved by the Emory University institutional review board . Cultured mouse PTEN−/−; Ink4a/arf−/− astrocytes ( gift of R . Bachoo ) were maintained in DMEM with 10% serum . The RIOK2 cDNA ( Origene ) was overexpressed in PTEN−/−; Ink4a/arf−/− astrocytes from the pBabe retroviral vector . The following drugs were used: Nutlin-3 ( Cayman ) , MG132 , Akt inhibitor IV ( Calbiochem ) , temozolomide ( Tocris ) , doxorubicin , rapamycin , PP242 ( Santa Cruz ) , PI-103 , ZVAD ( Enzo ) , A443654 ( gift of Greg Riggins ) , gefitinib ( LC Laboratories ) , BEZ-235 ( Biovision ) , LY294002 ( Cell Signaling Technology ) , MK-2206 , and GDC-0941 ( Selleck ) . Doses of LY294002 , BEZ-235 , PI-103 , GDC-0941 , and MK-2206 used on GBM cells were determined by using dose response assays to find the concentrations at which cells showed substantial reduction ( approximately <20% of normal ) in Akt-mediated phosphorylation of PRAS40 ( as detected by immunoblot ) . Lentiviral shRNA pLKO . 1 plasmids were obtained from the Broad Institute of MIT . RIOK1 shRNAs: TRCN0000196278 and TRCN0000196981 . RIOK2 shRNAs: TRCN0000197250 , TRCN0000196672 , and TRCN0000196684 . pLKO . 1-GFP and a nontargeting shRNA against lacZ ( in pLKO . 1 ) were used as controls . Lentivirus was produced and used as per standard protocols ( Sigma ) . Knockdown was evident by western blot 96 hrs post-infection . For neurosphere cultures , lentivirus was prepared in DMEM/F12 without serum , and infections were done on cells were plated adherently [36] . For siRNAs , all constructs were transfected at 50–100 µM with RNAimax ( Invitrogen ) . Unless otherwise noted , siRNA-treated cells were harvested at 72 hrs post-transfection . 2 sets of pooled siRNAs were tested each for RIOK1 and RIOK2 ( Dharmacon ) , and two different nontargeting siRNAs against GFP or luciferase were used as controls ( Dharmacon ) . For knockdown of p53 , p53 siRNAs were used and compared to matched control nontargeting siRNAs ( Cell Signaling Technologies ) . Target sequences are listed in Text S1 . For dual-knockdown experiments , U87MG cells were preferred because , with the necessary higher doses of siRNAs , U87MG-ΔEGFR cells showed nonspecific alterations in ΔEGFR expression that affected RIOK levels . Cells were lysed in RIPA buffer and cleared lysates were subjected to standard immunobloting . The following primary antibodies were used: RIOK1 ( Novus ) , RIOK2 ( Sigma ) , p53 ( Santa Cruz ) , p21 , EGFR ( BD ) , actin ( DSHB ) , RpL11 ( Invitrogen ) , NRBP2 ( Abcam ) , STK17a/DRAK1 ( Anaspec ) , PDGFRα , VRK1 , CDK9 , CDK7 , STK17B/DRAK2 , TLK1 , phospho-Akt ( S473 ) , phospho-Akt ( T308 ) , phospho-PRAS40 ( T246 ) , phospho-FOXO1 ( T24 ) /FOXO3 ( T32 ) , Akt , phospho-NDRG1 ( T346 ) , NDRG1 , PRAS40 , phospho-4E-BP1 , 4E-BP1 , PARP , cleaved caspase , FOXO3 , TSC2 , phospho-TSC2 , mTor , phospho-mTor ( T2448 ) ( Cell Signaling Technologies ) For WST-1 assays , cell lines were infected with lentiviral shRNA constructs and placed under selection for 48 hrs . Following selection , cells were plated for WST1 assays for cell proliferation/viability as per manufacturer's instructions ( Clontech ) . For flow cytometry ( FACS ) analysis of DNA content , cells were dissociated and stained with propidium iodide ( PI ) . For FACS analysis for apoptosis , cells were treated with indicated siRNAs or lentiviral vectors and stained with Annexin V-FITC and PI or 7AAD ( Invitrogen , BD Biosciences ) . Assays were performed on a FACScaliber II flow cytometer and data were collected using FACSdiva software ( BD Biosciences ) . Cell cycle profiles were generated using ModFit LT ( Verity Software House ) . In all cases , at least 5000 cells were analyzed per sample . 293T cells were transiently transfected with Myc-DDK-tagged RIOK2 constructs . Cells were lysed in 50 mM HEPES pH 7 . 5 , 150 mM NaCl , 5 mM EDTA , 1 mM DTT buffer with protease and phosphatase inhibitors [22] . RIOK2 was immunoprecipitated with M2-agarose ( Sigma ) from cleared lysates , and washed immunoprecipitates were subjected to immunobloting . Human brain tumor biopsies and tissues were obtained from the Brain Tumor Translational Resource under a protocol approved by the University of California , Los Angeles institutional review board . Paraffin embedded human brain tumor specimens and tumor tissue microarrays with matched control tissue were prepared and sectioned using the UCLA Pathology Histology and Tissue Core Facility . Immunohistochemical staining was performed as previously described [70] or as specified by manufacturer guidelines ( Sigma ) . The results were scored by neuropathologists according to standard clinical criteria , and images of RIOK2 immunoreactivity were taken on an Olympus DP72 . For orthotopic implantation of mouse astrocytes engineered in vitro , low passage cells ( no more than 8–10 passages ) were used in two separate experiments . 1×105 cells in 5 µl of PBS were injected stereotactically 2 mm lateral to the midline and 1 mm anterior to the bregma into the brains of 5–6 week old athymic nu/nu mice . Mice were monitored and all animals were sacrificed upon evidence of neurological symptoms in experimental groups such that all samples were time-matched . Brains were removed for processing and histological analysis . Sections were scored independently by two neuropathologists for the presence of tumors and injection-associated needle scars . Animals injected with RIOK2-expresing cells that developed tumors outside of the brain ( n = 1 ) were excluded from the final tally . All animal experiments were approved and conducted according to animal welfare guidelines of the IACUC at the University of California , San Diego . | Glioblastomas , the most common primary brain tumor , harbor mutations in receptor tyrosine kinases ( RTKs ) , such as EGFR , and components of the Pi-3 kinase ( PI3K ) signaling pathway . However , the genes that act downstream of RTK and PI3K signaling to drive glioblastoma remain unclear . To investigate the genetic and molecular basis of this disease , we created a glioblastoma model in the fruit fly Drosophila melanogaster . To identify new genes involved in glioblastoma development , we performed a screen for the genes required for tumor cell proliferation using our Drosophila glioblastoma model and then functionally assessed the activity of human versions of novel genes identified in this screen . Our results revealed that the RIO kinases become overexpressed in human glioblastomas but not in normal human glial or neuronal cells . We found that overexpression of the RIO kinases promotes and maintains signals that drive tumor cell proliferation and survival in RTK- and PI3K-dependent human glioblastoma , and reduction of RIO kinase expression decreased proliferation and prompted cell death and chemosensitivity in glioblastoma cells . Therefore , disruption of the RIO kinases may provide new therapeutic opportunities to target glioblastoma and other RTK- or PI3K-dependent cancers . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"cancer",
"genetics",
"signaling",
"in",
"selected",
"disciplines",
"neuroscience",
"model",
"organisms",
"oncogenic",
"signaling",
"biology",
"phospholipid",
"signaling",
"cascade",
"signal",
"transduction",
"genetic",
"screens",
"genetics",
"molecular",
"cell",
"biology",
"genetics",
"of",
"disease",
"genetics",
"and",
"genomics",
"signaling",
"cascades"
] | 2013 | A Kinome-Wide RNAi Screen in Drosophila Glia Reveals That the RIO Kinases Mediate Cell Proliferation and Survival through TORC2-Akt Signaling in Glioblastoma |
Cryptochromes are a class of flavoprotein blue-light signaling receptors found in plants , animals , and humans that control plant development and the entrainment of circadian rhythms . In plant cryptochromes , light activation is proposed to result from photoreduction of a protein-bound flavin chromophore through intramolecular electron transfer . However , although similar in structure to plant cryptochromes , the light-response mechanism of animal cryptochromes remains entirely unknown . To complicate matters further , there is currently a debate on whether mammalian cryptochromes respond to light at all or are instead activated by non–light-dependent mechanisms . To resolve these questions , we have expressed both human and Drosophila cryptochrome proteins to high levels in living Sf21 insect cells using a baculovirus-derived expression system . Intact cells are irradiated with blue light , and the resulting cryptochrome photoconversion is monitored by fluorescence and electron paramagnetic resonance spectroscopic techniques . We demonstrate that light induces a change in the redox state of flavin bound to the receptor in both human and Drosophila cryptochromes . Photoreduction from oxidized flavin and subsequent accumulation of a semiquinone intermediate signaling state occurs by a conserved mechanism that has been previously identified for plant cryptochromes . These results provide the first evidence of how animal-type cryptochromes are activated by light in living cells . Furthermore , human cryptochrome is also shown to undergo this light response . Therefore , human cryptochromes in exposed peripheral and/or visual tissues may have novel light-sensing roles that remain to be elucidated .
Cryptochromes are blue-light–absorbing photoreceptors found throughout the biological kingdom , involved in diverse and important signaling roles [1–3] . Cryptochromes were first identified in plants from a mutant of Arabidopsis thaliana ( A . thaliana ) , hy4 , which failed to show normal plant growth and developmental responses to blue light [4] . The N-terminal region of the HY4 encoding protein , renamed cryptochrome or A . thaliana cry1 ( Atcry1 ) was found to be highly homologous to a previously characterized class of enzymes , DNA photolyases , which utilizes blue light as a source of energy for the repair of UV-light–generated DNA lesions [2 , 5] . However , cryptochrome did not repair DNA but instead , participated in numerous blue-light–dependent plant growth responses , including early seedling development , leaf and stem expansion , initiation of flowering , and gene regulation [6 , 7] . The defining characteristic of a cryptochrome-type photoreceptor is therefore a light receptor molecule that is structurally highly similar to DNA photolyases , but has lost DNA repair activity and acquired a novel role in signaling . Subsequent to their discovery in plants , cryptochromes were identified in animal ( human and mouse ) systems by isolation of homologous cDNAs whose encoded proteins were likewise not functional in DNA repair [8] . Interestingly , these animal-type cryptochromes were more similar to a type of DNA photolyase that repairs 6–4 photoproducts , than to the type I cyclobutane pyrimidine dimer ( CPD ) -repairing DNA photolyases to which the plant cryptochromes are most closely related . Therefore , animal-type cryptochromes are thought to have evolved independently from different photolyase ancestors than plant cryptochromes [9] . A signaling role for animal-type cryptochromes was first identified in insects , Drosophila melanogaster ( D . melanogaster ) , through isolation of a mutation in the D . melanogaster cryptochrome ( Dmcry ) resulting in failure to properly entrain the peripheral circadian clock [10 , 11] . The role of Dmcry as a light-sensing input to the circadian clock is now well established , occurring by interaction of Dmcry with known clock proteins such as timeless or period [12 , 13] . Thus , although derived from different evolutionary photolyase ancestors , both plant Atcry1 and Dmcry act as signaling molecules that undergo light-sensitive interactions with partners to initiate signaling reactions . A schematic of the basic structural characteristics and evolutionary relation of cryptochromes and photolyases is presented in Figure 1A . Despite rapid advances in identifying the signaling pathways and molecular targets of cryptochromes in various organisms , the primary light-driven reactions that initiate the signaling process have remained elusive until very recently . Like DNA photolyases , cryptochromes bind flavin adenine dinucleotide ( FAD ) as a blue-light–absorbing cofactor [1 , 2] . However , the resting state of flavin in Atcry appears to be the fully oxidized redox form rather than the reduced form as found to be catalytically active in DNA photolyases [14] . Through a combination of in vitro studies with purified proteins and whole-cell in vivo spectroscopic techniques , it was deduced that protein-bound flavin both in Atcry1 and Atcry2 is reduced by blue light through intraprotein electron transfer resulting in accumulation of a relatively long-lived semiquinone intermediate form which is believed to represent the active signaling state [15–18] . This photoreaction is thought to be the basis for conformational changes that occur in the protein to initiate signaling [19 , 20] . When returned to darkness , plant cryptochromes slowly reoxidize back to the fully oxidized state of the flavin chromophore . In DNA photolyases , a similar light-induced photoreduction , known as photoactivation , can also occur but is relatively unimportant to biological activity since generally the FAD cofactor of DNA photolyase is fully reduced and mostly remains in this redox state independent of light conditions in vivo [21] . Therefore , plant cryptochromes have apparently transformed a minor photoreaction intrinsic to DNA photolyases to form the basis for a novel function as photoreceptor; a summary of the plant cryptochrome photocycle is presented in Figure 1B . The mechanism of light activation of animal-type cryptochromes is currently unknown . Although structurally similar to plant cryptochromes , the animal proteins apparently stem from 6–4 photolyase-type ancestors and not from CPD photolyases ( which are related to plant cryptochromes ) [9] and so could have evolved a different photocycle . Nevertheless , purified preparations of Dmcry have been recently shown to undergo a photoreduction reaction similar to plant cryptochromes in vitro , to a relatively stable radical intermediate [22] . Furthermore , studies of wavelength sensitivity show little Dmcry activity above 500 nm ( green/red ) light [23 , 24] . These characteristics are consistent with the known mechanism of activation of plant cryptochromes . On the other hand , mutation of conserved tryptophan residues that play a role in photoreduction [21] to redox inactive Phe have not been reported to affect biological activity of either mouse cry [25] or a recently characterized insect cryptochrome from monarch butterfly ( Danaus plexippus ) ( Dpcry1 ) and from other insects [26 , 27] . A further complication arose with the identification of a light-independent function for mammalian cryptochromes . Transgenic mice in which both existing cryptochrome alleles ( mcry1 and mcry2 ) had been knocked out showed a complete absence of rhythmic activity . This led to the conclusion that mammalian cryptochrome functions as a central component of the circadian oscillator [28 , 29] . Like insect cryptochromes , mcry1 and mcry2 were shown to interact with known components of the mammalian clock and thereby obtain a novel biological role . However , these interactions were entirely independent of light [30] , and moreover , the clock phenotype of mcry1 and mcry2 occurs in continuous darkness without light interruption over several days . Mammalian cryptochromes are similar to insect cryptochromes and apparently stem from the same 6–4 photolyase ancestor . Moreover , they are flavoproteins and show conservation in amino acids required for light activation in photolyases and other cryptochromes . Despite these similarities , the light-independent nature of mammalian cryptochrome response leads to the question of whether these signaling molecules retain the ability to respond to light at all . The aim of the present study is to resolve the question of whether and how light activates animal-type cryptochromes . We have employed in vivo spectroscopic techniques including a novel application of electron paramagnetic resonance ( EPR ) to detect photoconversion of flavin and accumulation of radical in living whole cells . We have examined both Dmcry and Homo sapiens cryptochrome-1 ( Hscry1 ) as representative light-sensitive and light-insensitive cryptochromes , respectively . These experiments showed that light activation of animal cryptochromes occurs by photoreduction and accumulation of a radical signaling intermediate , similar to plant cryptochromes and unlike photolyases . Furthermore , because Hscry1 undergoes the same photoreactions , mammalian cry is demonstrated to have the capacity to function as a light sensor .
In plant cryptochromes cry1 and cry2 , the dark state of the flavin is in the oxidized form in vivo ( Figure 1B ) . To determine the nature of the dark state of flavin in animal-type cryptochromes , we measured a classic action spectrum for Dmcry activity in living flies . An action spectrum is a dose-response curve for photoreceptor sensitivity in which the response of an organism is determined at multiple wavelengths of light and at multiple light intensities at each wavelength . In this way , the response will depend on how well the photoreceptor absorbs light at the given wavelength . The wavelength at which peak activity can be observed in the living organism indicates the absorption maximum ( in which the light is absorbed at highest efficiency ) of the responsible photoreceptor . If performed to sufficient resolution , such action spectra can be compared to the absorption spectrum of a purified pigment or photoreceptor and in this way identify the nature of the photoactive pigment implicated in a given biological response [6] . As a possible assay for Dmcry function , we investigated its characteristic property of degradation that followed upon photoreceptor activation . Levels of Dmcry protein decrease rapidly in flies subsequent to blue-light irradiation , likely due to conformational change in the photoreceptor leading to targeting to the proteasome [11 , 23 , 31] . This degradation can be quantitatively monitored by western blot analysis with Dmcry antibody . However , in order to be useful for action spectroscopy , the amplitude of the response ( decline in Dmcry concentration ) must be proportional to the number of photons of light energy absorbed by the photoreceptor , and not simply a delayed response with little direct relation to the light input signal . To test this property , we irradiated living flies for fixed time intervals with blue light ( 450 nm ) and observed decrease in levels of Dmcry protein over time as previously described [11 , 23 , 31] ( Figure 2A ) . Importantly , when blue-light irradiation was performed at different blue-light intensities , the time required to reach a given decline in Dmcry protein levels was proportional to the given light intensity . For example , to obtain a decrease to 50% of the original Dmcry protein concentration requires 6 . 2 min at irradiance of 200 μmol m−2 sec−1 , 9 . 3 min at 150 μmol m−2 sec−1 , and 12 . 5 min at 100 μmol m−2 sec−1 ( from Figure 2A ) , respectively . Therefore , Dmcry degradation obeys the Bunsen-Roscoe law of reciprocity , indicating that it is a response to the total number of photons , independent of irradiance time , and so represents an accurate measure of photoreceptor light responsivity [32] . For generation of this Dmcry action spectrum , living flies were first dark adapted to accumulate maximum levels of dark state cryptochrome . Flies were then subjected to continuous irradiation at a set photon fluence rate of 17 μmol m−2 s−1 at wavelengths between 380 and 502 nm . Shorter wavelengths are impractical because of the increasing absorption from cell components and therefore increasing errors . Levels of Dmcry protein were monitored by western blot analysis . Illumination time was varied to provide the different irradiance , as permitted by reciprocity . Shorter or longer wavelengths of light proved to be ineffective at eliciting significant response ( unpublished data ) . Dose-response plots of the time course of Dmcry protein degradation at different wavelengths of light showed linear decay as a function of the log irradiation time for points between 20% and 90% of dark levels of protein accumulation ( Figure S1 ) . The action spectrum is plotted from these dose-response curves using the total irradiance required to reduce Dmcry protein levels to 50% of dark controls at each wavelength ( Figure 2B ) . The curve is inverted to give a visual image whereby the peak efficiency ( the wavelength that required the shortest time to elicit 50% Dmcry degradation ) represents the absorption maximum of the responsible photoreceptor . Peak wavelength sensitivity was at 450 nm , with defined shoulders around 420 and 480 nm , matching the spectrum of protein-bound oxidized flavin [33 , 34] . Oxidized flavin is therefore the likely photoactive pigment of Dmcry in living whole flies , similar to plant cryptochromes [14] and in marked contrast to DNA photolyases in which flavin is fully reduced [2] . We next determined the nature of the chemical reaction induced by light in animal cryptochromes in living cells . We performed baculovirus-driven expression of Dmcry and Hscry1 cryptochromes in Sf21 insect cells , where photoreceptor protein accumulates to sufficiently high levels for direct application of spectroscopic and biophysical techniques in vivo [18] . To verify whether cryptochrome-bound flavin can be directly observed , expressing whole Sf21 cells were harvested and placed intact inside a fluorimeter . Fluorescence emission was measured at 525 nm ( characteristic of oxidized flavin ) over an excitation range of 400–500 nm . Despite the substantial scatter due to measurements of these living intact cells , there was clearly observable signal increase peaking for excitation at 450 nm in cells overexpressing both Dmcry and Hscry1 as compared to uninfected control cells . These results are consistent with oxidized flavin bound to the dark state of the photoreceptors ( Figure S2 ) . These data showing increased oxidized flavin in cryptochrome-expressing cells are in agreement with the resting state of Dmcry determined from action spectroscopy ( Figure 2B ) . Interestingly , mammalian cryptochrome also accumulates in the oxidized form and thereby shows functional similarity to Dmcry and not to DNA photolyases . Similar results have been previously obtained for Atcry1 [18] . To initiate the photochemical reaction , Dmcry- or Hscry1-expressing cells were irradiated with blue light and returned to the fluorimeter at intervals for measurement of excitation spectra . This assay detects change in levels of oxidized flavin in these living cells . For both Dmcry and Hscry1 , peak excitation at 450 nm showed a progressive decrease over time that matches the spectra for photoreduction of oxidized flavin ( Figure 2C and 2D ) . This decrease was not due to protein degradation since both Dmcry and Hscry1 protein levels remain stable throughout the time course of illumination in Sf21 cells ( Figure S3 ) . Furthermore , flavin reoxidation is observed when illuminated cell cultures are returned to darkness ( unpublished data ) , indicating that no cryptochrome degradation has occurred . Therefore , both tested cryptochromes had undergone a photoreaction in vivo , leading to change in redox state of protein-bound flavin ( Figure 2D ) . A similar reaction , known as photoactivation , occurs in DNA photolyases , wherein the flavin chromophore is converted to the fully reduced form by an electron transfer reaction ultimately fed by an extrinsic reductant . An intraprotein electron transfer pathway from the protein surface to the buried flavin has been derived for this light-driven reaction in Escherichia coli DNA photolyase ( EcPl ) based on crystallographic structural information and on a combination of site-directed mutagenesis and spectroscopy [35–38] . This pathway comprises a chain of three tryptophan residues ( W382–W359–W306 ) that are highly conserved throughout the photolyase/cryptochrome family . Recently , a study with purified Atcry1 has demonstrated the functional relevance of this reaction to cryptochrome photoreceptor activity [16] by substitution of redox-inactive phenylalanines for two tryptophan residues , W400 and W324 , which are found in the Atcry1 sequence and crystal structure [39] at the homologous positions to W382 and W306 of EcPl , respectively . These mutant proteins ( W400F and W324F ) lack the predicted electron donor proximal to the flavin ( W400 ) or exposed to the protein surface ( W324 ) . Both proteins were found to have impaired electron transfer activity in vitro and reduced biological activity in living plants . To determine whether flavin photoreduction may occur by a similar pathway of intermolecular electron transfer in animal-type cryptochromes , we have made point mutations in Dmcry of two conserved tryptophan residues . One mutation is distal to the flavin in this pathway ( W342F ) , which corresponds to W306 in EcPl and W328F in Dpcry1 [26] , respectively . Second , we have introduced a substitution into the middle member of the electron transfer chain of Dmcry , corresponding to W359 of EcPl . The mutant Dmcry proteins were expressed in Sf21 insect cells to high levels and subjected to in vivo fluorescence spectroscopy to follow flavin photoreduction . For determination of in vivo photoreduction , cryptochrome-expressing Sf21 cells were irradiated with blue light and returned to the fluorimeter periodically to determine remaining levels of oxidized flavins . Unexpectedly , both wild-type and mutant Dmcry proteins showed similar rates of photoreduction in these living cells at high light intensity ( 150 μmol m−2 sec−1 white light ) , as did the W400F mutant of Atcry1 ( Figure 3A ) . This result is surprising since purified preparations of Atcry1 W400F protein and of the W328F Dpcry1 ( homolog to Dmcry W342F ) showed greatly reduced photoreduction in vitro at even higher light intensities , and there was no significant radical accumulation after this time period [16 , 26 , 27] . Therefore , the efficiency of cryptochrome photoconversion in vivo is much higher than that of the purified , isolated protein in vitro , perhaps due to a more conducive redox environment and the presence of relevant electron donors/acceptors in vivo . Nevertheless , at lower photon fluence ( 10 μmol m−2 sec−1 blue light ) , a significant decline in the rate of photoreduction is observed in the phenylalanine mutants of both Dmcry and Atcry1 as compared to wild-type proteins ( Figure 3B—see also Figure 4SA and 4SB for experiments performed with further reduced photon fluence ) . This phenomenon provides a consistent explanation for the observed biological activity of amino acid–substitution mutants in Dpcry and Atcry1 . In Arabidopsis , biological activity of the mutant proteins ( W400F and W324F ) was determined at only low blue-light intensity and found impaired at this irradiance in vivo . To determine the state of the photoreceptor ( radical or fully reduced ) in the activated cryptochrome , fluorescence emission techniques in whole cells are not sufficient as they can identify only the oxidized form of the flavin chromophore . It cannot , therefore , be concluded from the above studies whether photoreduction in vivo leads to accumulation of a semiquinone intermediate as for plant cryptochromes [16–18] , or whether the fully reduced form of flavin accumulates in animal cryptochromes as for DNA photolyases [1 , 2] . To directly monitor for radical accumulation in response to light in vivo , whole-cell EPR spectra were recorded as previously described [17 , 18] . Intact Sf21 insect cells with overexpressed Dmcry or Hscry1 protein were irradiated in parallel with nonexpressing control cells at the identical intensities of blue light and rapidly frozen for EPR analysis . A paramagnetic species that does not accumulate in control cells was induced by blue-light irradiation of Dmcry- ( Figure 4A , traces B and C ) and Hscry1-expressing cells ( Figure 4A , traces E and F ) . This species appears with similar kinetics to both plant cryptochromes [17 , 18] . Interestingly , there was detectable amount of a radical present even in the dark in some samples ( see trace D ) , although not in all trials , perhaps due to concentrations below our level of detection . This result is in marked contrast to previous experiments , for which radical accumulation was never observed in unilluminated cells [17 , 18] . Finally , we have examined the Drosophila mutant proteins W397F and W342F for radical accumulation in vivo ( Figure 4A , traces H–K ) . Saturating illumination ( 40 μmol m−2 sec−1 blue light ) leads to accumulation of a radical intermediate form . To further characterize these signals , X-band–pulsed ENDOR spectroscopy was applied to illuminated whole cells expressing Dmcry ( Figure 4B ) . The observed spectrum in the expressing cells ( trace A ) is very similar to that obtained from the purified Dmcry protein ( trace B ) . Both spectra differ from those of neutral flavin radicals as seen in plant cryptochromes and corroborate the assignment to an anionic radical species as given previously for the purified protein [22] . Taken together , the in vivo spectroscopic data conclusively indicate that the photocycle for both Dmcry and Hscry1 involves light-dependent flavin reduction and accumulation of the radical state . Finally , it is necessary to establish the biological relevance of the observed in vivo photoconversion of animal cryptochromes . In plant cryptochromes , the radical state has been demonstrated to be the biologically active signaling state for both Atcry1 and Atcry2 [17 , 18] . This conclusion resulted from the observation that green light reversed the effect of blue light in the course of cryptochrome activation , due to photoconversion of the active , radical form to the fully reduced , inactive species [17 , 18 , 40] ( see also Figure 1A ) . A simple means to determine whether light-induced radical accumulation also has biological relevance for animal cryptochromes in vivo is therefore to measure whether green light ( above 525 nm ) affects both Dmcry protein accumulation and the kinetics of cryptochrome photoreduction . To test this prediction , we performed bichromatic irradiation of flies simultaneously with blue and green light ( B+G ) and compared the response to that obtained with the identical intensity of blue light by itself ( B ) ( Figure 5A ) . Green-light irradiation by itself resulted in no change in Dmcry protein levels ( unpublished data ) . In each of three independent trials , we observed more rapid decline in Dmcry protein levels in blue light ( B ) as compared to coirradiation with blue and green light ( B+G ) . This antagonistic effect can only be explained by photoconversion of the ( green-light absorbing ) radical signaling state to an inactive redox form . In the case of Dmcry-expressing cell cultures , an effect of green light on cryptochrome photoreduction was directly monitored . Cell cultures irradiated with blue and green light ( B+G ) show accelerated photoreduction of Dmcry as compared to blue light ( B ) alone ( Figure 5B ) . Although the accumulation of fully reduced flavin can not be directly monitored by this technique , these data are consistent with a shift in the overall flavin photoequilibrium subsequent to formation of the radical , and thereby consistent with the effect of green light on biological activity observed in living flies . Although the present study so far has shown that mammalian cryptochromes undergo similar photoreactions to those of insect and plant , a functional role for light in biological activation remains to be demonstrated . To address this question , we have assayed for a form of activation of Hscry1 in response to light in living flies , where endogenous cryptochrome ( Dmcry ) is known to undergo light-dependent changes resulting in proteolysis ( Figure 2 ) . Transgenic flies expressing full-length Hscry1 under the control of the UAS promoter element were obtained by established procedures ( see Materials and Methods ) . Expression of the recombinant Hscry1 was verified by western blot analysis in two independent transformed lines ( A and B ) . Expressing flies were then dark adapted to accumulate maximal cryptochrome protein and subsequently irradiated with white light . Levels of Hscry1 were assayed during the course of the irradiation . Interestingly , as is the case for Dmcry , significant decrease in Hscry1 protein levels were observed shortly after transfer to white light ( Figure 6 ) . These results indicate that Hscry1 undergoes light-dependent proteolysis as does Dmcry in living flies . Since degradation of Dmcry correlates with activation by light and biologically relevant radical formation , a similar mechanism of biological activation is also likely for Hscry1 .
In this work , we provide , for the first time , evidence for a photocycle of animal-type cryptochromes such as found in insects and mammals . Cryptochrome-bound flavin is found in an oxidized redox state in vivo , and light activation results in flavin photoreduction to a radical intermediate that represents the likely signaling state . The biological significance of this reaction is supported by the observation of antagonistic effects of green light on Dmcry function , which reduces levels of radical intermediate [17 , 18 , 40] . This mechanism contrasts with that of DNA photolyases in which flavin is fully reduced for catalytic activity . Most importantly , Hscry1 from a cryptochrome subfamily with no established light response also has the capacity to undergo this photoreaction in living cells , suggesting the possibility of novel light-sensing capabilities in humans . A number of studies have indicated that Dmcry responsivity occurs primarily at wavelengths below 500 nm [23 , 24] . The current study extends these prior observations by providing sufficient fine structure to identify oxidized flavin , with peak of activity at 450 nm and defined shoulders around 420 and 480 nm , as the likely responsible photopigment in the visible range . Further corroboration for the assignment of oxidized flavin as the ground state for animal-type cryptochromes is provided by a classic action spectrum of phase shift in pupal emergence of Drosophila [41] , a response involving phase shift of the circadian clock which is now known to be under the control of cryptochrome [11] . Consistent with the current work , peak activity was at 450 nm , and the spectrum matches that of protein-bound oxidized flavin . Interestingly , Dmcry degradation in Schneider cells has been reported to have a peak of activity in the near-UV spectral region [24] , whereas in living flies and Dmcry-expressing Sf21 cells , the peak of activity is at 450 nm ( see Figure 2B ) . Like DNA photolyases , cryptochromes are proposed to bind folate derivatives as cofactors in addition to flavin [8] . In DNA photolyases , a folate derivative absorbs light primarily in the near-UV spectral region ( 370–400 nm ) and transfers energy to the flavin chromophore [2] . Recently , a similar role for folate has been postulated in plant cryptochromes [42] whereby light energy for photoreduction is transferred to flavin through a UV antenna pigment . It is therefore likely that the reported near-UV responsivity of Dmcry in Schneider cells also results from light absorption by a folate ( or another , yet unspecified ) antenna pigment . In that case , absence of near-UV responsivity in Dmcry extracted from whole flies ( Figure 2B ) suggests that the second chromophore of animal-type cryptochromes may not be available in the majority of insect cell types . This is in line with older experiments done with DNA photolyases , where a low binding constant of the folate chromophore and a therefore heterogeneous folate concentration was concluded . In plant ( Atcry1 and Atcry2 ) cryptochromes , flavin photoreduction leading to a meta-stable neutral radical accumulation can be observed in in vitro experiments . This property of purified plant cryptochrome contrasts published DNA photolyase data , in which oxidized flavin is rapidly converted to the fully reduced redox state ( necessary for DNA repair ) . Recently , photoreduction experiments were performed with purified preparations of several insect cryptochromes in vitro resulting in similar photoreactions ( accumulation of radical and not fully reduced flavin ) , although the anionic radical , and not the neutral radical , was accumulated [22 , 26 , 27] . These results are consistent with the presently derived in vivo photocycle for animal cry activation . Results from recent studies performed with various insect cryptochromes ( fruitfly [Dmcry] , butterfly [Dpcry1] , mosquito [Agcry1] , and silk moth [Apcry1] ) [26 , 27] have called into question the assignment of oxidized flavin as the ground state for animal cryptochromes and argued against a photocycle involving flavin photoreduction . Their interpretation focused on the observation that substitution of amino acids necessary for flavin photoreduction in vitro does not abolish biological activity of these proteins in vivo . This apparent discrepancy between the absence of photoreduction in vitro yet significant biological activity in vivo is resolved by the observation that amino acid substitutions abolishing in vitro photoreduction of purified Dmcry does not , in fact , abolish photoreduction activity in vivo . Photoreduction of oxidized flavin measured by fluorescence techniques ( Figure 3A and 3B ) in these substitution mutants correlates with concomitant appearance of anionic radical as determined by EPR spectroscopic techniques ( Figure 4 ) . The same is true for the W400F mutant of Atcry1 , which shows normal rates of photoreduction in vivo under high light intensities even though flavin photoreduction in vitro is virtually zero under these conditions [16] . Cryptochrome photoreduction , therefore , occurs far more efficiently , and by additional alternate pathways , in vivo than is observed for the purified protein in vitro . A similar discrepancy between the light required to activate a purified photoreceptor protein in vitro as compared to activation in vivo has been noted for other photoreceptors , for instance , the class of phototropins [34 , 43 , 44] in which blue-light–dependent autophosphorylation requires a much higher irradiance in vitro to obtain a similar effect than is required in vivo . Results from recent studies showing reduced biological activity at lower light intensity in the W342F mutant of Dmcry [27] are consistent with our observed reduced rates of photoreduction at low photon fluence ( Figure 3B ) . Although quantitation was not formally performed , a prior study in which function of amino acid substitutions in the Trp triad of Dmcry was analyzed [25] also showed apparent reduced biological activity in W-F substitution mutants . In this study , the authors proposed that F can function as electron donor similarly to W , which , however , is not observed [16 , 26 , 27]; nevertheless , their data are entirely consistent with the present work . Finally , observations from point mutations that reduce the rate of radical formation in Apcry1 ( C402A ) abolish protein function in vivo [27] are entirely consistent with our assignment of the radical as the signaling state of the receptor [16] . A proposed mechanism whereby the anionic radical may be the resting state [27] is unlikely given that peak activity is not observed at 470 nm either in the present ( Figure 2 ) or former studies of wavelength sensitivity for this receptor [24 , 41] . The derived photocycle of animal cryptochromes is therefore similar to the reaction mechanism of plant cryptochromes ( Figure 1B ) . Both photocycles involve reduction of flavin leading to cycling between radical ( active ) and oxidized ( inactive ) redox forms . Since these different families of cryptochromes evolved independently from unrelated DNA repair enzyme ancestors , there must be a latent property of DNA photolyases that lends itself to development of photoreceptor properties . The likeliest possibility is that the flavin semiquinone form confers some conformational change on the protein , which can be recognized by yet-unidentified signaling partners and thereby be readily adapted to a role in a signaling pathway . In addition to the classic plant and animal-type cryptochromes , a third family of cryptochromes ( cryDASH ) has been identified in Synechocystis and many other organisms [45] . CryDASH cryptochromes are structurally similar to DNA photolyases , but do not efficiently repair DNA . They evolved from a 6–4 photolyase ancestor but are apparently unrelated to either the plant ( Atcry1 and Atcry2 ) or animal-type cryptochromes described in this study . Although no information as yet exists on the in vivo redox state and photocycle of cryDASH , the purified proteins are converted by light to the fully reduced flavin in vitro [45 , 46] and retain some single-strand DNA repair activity [47] , in this respect appearing more similar to DNA photolyases . It is therefore possible that an additional , entirely unrelated photocycle has evolved for cryDASH-type cryptochromes that is not similar to plant or animal-type cryptochromes . Perhaps the most intriguing finding of the present study is that mammalian cryptochromes , in particular Hscry1 , are responsive to light in vivo . Mammals are generally large , dense ( and also often nocturnal ) animals , and it makes sense that a molecule such as cryptochrome , which is an essential component of the circadian clock , should be regulated by other means than by direct absorption of light . In fact , with the exception of an isolated report of cryptochrome effect on pupil dilation in mice [48] , there has to date been no definite report of light-responsive phenotypes attributed to mammalian cryptochromes at all . Nevertheless , as the present study has shown , Hscry1 can undergo photoconversion in living organisms via a mechanism conserved with that of light-responsive cryptochromes . Hscry1 further undergoes light-dependent proteolysis in living flies , similar to the response mediated by appropriate E3 ubiquitin ligases such as COP1 in the case of Arabidopsis cry2 [49] , Therefore , the observed light sensitivity of Hscry1 in Drosophila is likely to result from surface changes leading to fortuitous recognition of the activated form by a fly E3 ligase . Since Hscry1 is widely distributed in many tissue types of humans , it could be activated by light in the retina or in locations close to the surface , providing the basis for novel biological signaling functions that remain to be determined .
As a light source , white light was produced by slide projectors placed before interference filters of 8–15-nm half-band width ( Schott Glaswerke ) . Filters used in the irradiations for the action spectrum were 380 ± 10 nm , 402 ± 12 nm , 418 ± 13 nm , 428 ± 10 nm , 438 ± 11 nm , 445 ± 10 nm , 466 ± 11 nm , 471 ± 16 nm , 492 ± 15 nm , 502 ± 15 nm , 515 ± 15 nm , and 525 ± 12 nm . Preparation of fly extracts and detection of dcry was performed essentially as described [50] . Two-week-old adult Drosophila ( after eclosion ) were adapted to dark for 3 d prior to an experiment . Between ten and 15 flies were placed under the indicated wavelengths of light during their subjective night phase ( between circadian time [CT] 20–22 ) and irradiated for the indicated times . Unirradiated control flies did not show variation in Dmcry levels during the time period that illuminations were performed on the test flies . Whole flies were harvested into liquid nitrogen , and extracted proteins prepared as described [50 , 51] . A total of 20 μg of protein was loaded per lane of an SDS polyacrylamide gel and transferred to PVDF membrane . Protein load was verified prior to load by Bio-Rad Bradford assay and subsequently by Coomassie staining of the blotted gels . Western blot analysis was performed by established methods with affinity-purified anti-Dmcry antibody [51] or Hscry1 antibody to a peptide fragment comprising amino acids 178–194 of the coding sequence ( Neosystems ) . Quantitation of resolved bands was performed digitally using Quantity One imaging software from Bio-Rad . cDNA for Dmcry and Hscry1 were cloned into a baculovirus transfer vector ( pBakPak9; Clontech ) by established protocols ( Clontech ) . A histidine tag was introduced upstream of the initial ATG in each construct to allow fast purification . For protein expression , amplified viral supernatant of the recombinant clones were mixed with cell culture and incubated as recommended ( Clontech ) . Protein expression was verified by western blot analysis with appropriate antibody of both whole-cell extract and of proteins purified by metal-affinity chromatography . Presence of flavin was verified by absorption and fluorescence spectra of the purified proteins . For construction of Dmcry mutants W397F and W342F , side-directed mutagenesis was performed by the recommended protocol using Altered Sites II in vitro mutagenesis kit ( Promega ) . The primers designed for mutagenesis are as follows: W397F , GTGCTGCAGTCCATGCTCGAAGCTCTGCCACAA; W342F , CTCGTTCGGCTTAGCGAACGGGATGCTCAGGCA . All clones were sequenced in entirety prior to protein expression . Whole-cell fluorescence experiments of Sf21 insect cells expressing recombinant cryptochrome photoreceptors were performed essentially as described [18 , 42] . Living Sf21 insect cells expressing cryptochrome protein or uninfected controls were centrifuged from culture medium , resuspended in PBS buffer ( 0 . 02 M sodium phosphate [pH 7 . 4] , 0 . 15 M sodium chloride ) , and placed directly into cuvettes at 10 °C for measurement of fluorescence spectra . Fluorescence emission at 525 nm was monitored in a Varian fluorescence spectrophotometer over a range of excitation wavelengths or at a single designated wavelength as indicated ( see Figure 3 legend ) . Excitation and/or emission spectra were always determined in parallel , both for infected ( cryptochrome-expressing ) and uninfected cell cultures at identical cell density . For light treatments , samples were removed from the spectrophotometer and placed on ice . Illumination was carried out for the indicated times , using the designated interference filters . Samples were then returned to the fluorescence spectrophotometer to monitor differences in excitation and emission spectra . All experiments were repeated for a minimum of three independent trials . Sf21 insect cells expressing both Dmcry and Hscry1 , control Sf21 cells , and purified Dmcry from Sf21 insect cells were resuspended in phosphate-buffered saline supplemented with 50% ( v/v ) glycerol in the dark . Aliquots were transferred into EPR quartz tubes ( 3 mm i . d . ) and illuminated for different times at 290 K with blue light ( Halolux 100HL; Streppel ) using a 420–470-nm band filter ( Schott ) . Samples were then frozen rapidly under illumination in liquid nitrogen and stored therein . X-band continuous-wave ( cw ) EPR spectra were recorded using a pulsed EPR spectrometer ( Bruker Elexsys E580 ) with a cw resonator ( Bruker ER 4122SHQE ) immersed in a helium-gas flow cryostat ( Oxford CF-935 ) . X-band–pulsed ENDOR spectra were recorded on the same spectrometer using an ENDOR accessory ( Bruker E560-DP ) , an RF amplifier ( Amplifier Research 250A250A ) , and employing a dielectric-ring ENDOR resonator ( Bruker EN4118X-MD-4W1 ) . The temperature was regulated to ±0 . 1 K by a temperature controller ( Oxford ITC-503S ) . The cw-EPR spectra were recorded at 120 K with a microwave power of 3 . 0 μW at 9 . 7 GHz microwave frequency with field modulation amplitude of 0 . 3 mT ( at 100 kHz modulation frequency ) . For Davies-type ENDOR spectroscopy , a microwave pulse sequence π–T–π/2–τ–π with 64- and 128-ns π/2- and π-pulses , respectively , and a RF pulse of 10-μs duration starting 1 μs after the first microwave pulse were used . The pulse separations T and τ between the microwave pulses were selected to be 13 μs and 500 ns , respectively . To avoid saturation effects due to long relaxation times , the entire pulse pattern was repeated with a low repetition frequency of 200 Hz . Spectra were taken at a magnetic field of 345 . 7 mT and a microwave frequency of 9 . 71 GHz . The entire coding sequence from the 5′ ATG onwards of HSCRY1 was introduced behind the upstream UAS promoter element of the pP ( UAST ) vector via PCR amplification and verified by sequencing; constructs were subsequently introduced into flies by standard methods [51] . Two high-expressing transformed lines with insertions on chromosome II and III , respectively , were selected for further study . Expression from the UAS upstream promoter element was induced by crossing flies to homozygous tim-GAL4 lines expressing the gal4 transcription activator driven by the TIMELESS ( tim ) promoter as described [52] . Parental lines used for crosses were: yw;tim-GAL4 [53] and Hscry1 insertion lines w;31 . 2A and w;26 . 9A . Expression was verified in F1 heterozygote progeny by western blot analysis with anti-Hscry1 antibody ( monoclonal antibody ref: BIN165979; http://Antikoerper-online . de ) . | Vision in animals is generally associated with light-sensitive rhodopsin pigments located in the eyes . However , animals ranging from flies to humans also possess ancient visual receptors known as cryptochromes in multiple cell types . In this work , we study the mechanism of light sensing in two representative animal cryptochromes: a light-sensitive Drosophila cryptochrome ( Dmcry ) and a presumed light-insensitive mammalian cryptochrome from humans ( Hscry1 ) . We expressed recombinant cryptochromes to high levels in living cells , irradiated the cells with blue light , and analyzed the proteins' response to irradiation with electron paramagnetic resonance and fluorescence spectroscopic techniques . Photoreduction of protein-bound oxidized FAD cofactor to its radical form emerged as the primary cryptochrome photoreaction in living cells , and was correlated with a light-sensitive biological response in whole organisms . These results indicate that both Dmcry and Hscry1 are capable of undergoing similar light-driven reactions and suggest the possibility of an as-yet unknown photo-perception role for human cryptochromes in tissues exposed to light . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"biochemistry",
"biophysics",
"molecular",
"biology"
] | 2008 | Human and Drosophila Cryptochromes Are Light Activated by Flavin Photoreduction in Living Cells |
In development , lineage-restricted transcription factors simultaneously promote differentiation while repressing alternative fates . Molecular dissection of this process has been challenging as transcription factor loci are regulated by many trans-acting factors functioning through dispersed cis elements . It is not understood whether these elements function collectively to confer transcriptional regulation , or individually to control specific aspects of activation or repression , such as initiation versus maintenance . Here , we have analyzed cis element regulation of the critical hematopoietic factor Gata2 , which is expressed in early precursors and repressed as GATA-1 levels rise during terminal differentiation . We engineered mice lacking a single cis element −1 . 8 kb upstream of the Gata2 transcriptional start site . Although Gata2 is normally repressed in late-stage erythroblasts , the −1 . 8 kb mutation unexpectedly resulted in reactivated Gata2 transcription , blocked differentiation , and an aberrant lineage-specific gene expression pattern . Our findings demonstrate that the −1 . 8 kb site selectively maintains repression , confers a specific histone modification pattern and expels RNA Polymerase II from the locus . These studies reveal how an individual cis element establishes a normal developmental program via regulating specific steps in the mechanism by which a critical transcription factor is repressed .
Metazoan development is characterized by complex transcriptional programs specified by gene regulatory networks [1] , [2] . Transcription factors in these networks occupy specific cis elements at target gene loci where they modulate chromatin remodeling and modification , and thereby transcription . The covalent modification of histones to yield specific histone marks promotes either the activation or repression of transcription [3] . Models of gene regulation have led to an attractive paradigm in which repression occurs in sequential stages of increasing stability [4] . While transcription factors bind and recruit chromatin-modifying and remodeling proteins , the relative contribution of individual cis elements residing within clusters of cis elements to the transcriptional control of endogenous loci is incompletely understood . GATA factor cross-regulation represents an instructive model system for investigating the contribution of individual cis elements to the initiation and maintenance of transcriptional repression . The GATA family of transcription factors plays diverse roles in multiple developmental contexts [5] . GATA factors are often expressed in an overlapping but reciprocal pattern , such that expression of one GATA factor increases as expression of another decreases . For example , GATA-1 directly represses Gata2 transcription via displacing GATA-2 from chromatin sites at its own locus , a process termed a “GATA Switch” [6] , [7] . GATA factor function has been extensively studied in the context of hematopoiesis , where GATA-1 , GATA-2 , and GATA-3 are key regulators . GATA-2 has a broad role in hematopoietic development , as demonstrated by impaired hematopoiesis in Gata2 knock-out mice resulting in lethality during midgestation [8] , [9] . GATA-1 is critical for the production of red blood cells and platelets [10] , and GATA-3 is required for specification of T cells [11] . Forced expression of GATA-2 blocks erythroid development [12] , [13] , [14] , leading to a model in which GATA-1-mediated repression of Gata2 through specific cis elements is required for differentiation . Genome-wide studies revealed GATA-1 occupancy at only a small subset of cis elements in the genome [15] . These cis elements exist as single or more complex GATA motifs , although the functionality of different permutations of GATA motifs at endogenous loci has not been investigated . The role of individual GATA-binding sites in gene regulation has been investigated extensively at the Gata2 locus , where several conserved GATA motif-containing regions span approximately 100 kb of the locus [16] . To test whether GATA switch sites function collectively or independently to regulate Gata2 expression , and to investigate the underlying mechanisms , we generated mice lacking one of these regulatory regions residing −1 . 8 kb upstream of the Gata2 promoter . We find that while this site is not essential for Gata2 expression in hematopoietic progenitors or initiation of Gata2 repression during erythropoiesis , it maintains Gata2 repression in erythroblasts . Molecular analyses demonstrate that loss of the −1 . 8 kb site reduces GATA-1 binding , allows for increased RNA Polymerase II ( Pol II ) occupancy at the locus , and results in changes in select histone marks . Further , elimination of the −1 . 8 kb site dysregulates Gata2 transcriptional control , disrupts the GATA-2-dependent genetic network , and interferes with red blood cell maturation . These results highlight the qualitatively distinct activities of individual cis elements in specific aspects of gene repression during development .
Previous studies in erythroid cell lines [17]–[22] and transgenic mouse models [23]–[25] have identified five GATA-binding regions upstream and in an intron of the Gata2 locus ( Figure 1A ) . It remains unknown whether these regions function collectively to confer Gata2 transcriptional regulation , or if individual regions function uniquely at specific developmental stages and/or in select cell types . The site at −1 . 8 kb is of considerable interest , since it possesses strong GATA-2 binding activity that is lost upon repression [17] . Thus , we reasoned that removal of this site would phenocopy GATA-2-deficiency . As definitive analysis of cis element function requires genetic ablation of endogenous loci , we generated a mouse strain lacking the palindromic GATA-binding site 1 . 8 kb upstream of the Gata2 transcriptional start site ( Δ-1 . 8 allele ) ( Figures 1B , S1 ) . Mice homozygous for the Δ-1 . 8 allele were born at expected Mendelian ratios , as assessed by PCR genotyping ( Figure 1C . ) , implying that embryonic development was largely unaffected . Morphologically , E12 . 5 wild-type and mutant embryos were similar ( Figure S2A ) , and adult mutant mice lacked gross abnormalities ( data not shown ) . We analyzed fetal liver erythropoiesis in Δ-1 . 8 mice for alterations in Gata2 expression . Using fluorescence-activated cell sorting with the erythroid markers CD71 and Ter119 [26] , we isolated cells from Stages I , II , III , and IV , corresponding to CD71loTer119− ( committed erythroid progenitors , Stage I ) , CD71hiTer119− ( proerythroblasts , Stage II ) , CD71hiTer119+ ( basophilic erythroblasts , Stage III ) , and CD71loTer119+ ( late erythroblasts , Stage IV ) ( Figure 2A ) . In wild-type mice , Gata2 was most highly expressed in Stage I progenitors , after which it was repressed in Stages II , III , and IV ( Figure 2B ) . Gata2 expression was modestly increased in Stage IV , to about one fifth of that observed in Stage I . In Δ-1 . 8 mice , Gata2 expression was normal in Stage I , and decreased normally in Stage II and III , indicating that the −1 . 8 kb site is not required for initiation of GATA-1-mediated repression . However , Gata2 expression was significantly elevated in Stage IV cells from the Δ-1 . 8 versus wild-type mice ( p≤0 . 05 ) ( Figure 2B ) . Thus , the −1 . 8 kb site is selectively required to maintain Gata2 repression in Stage IV erythroblasts . To determine if GATA-2 derepression has functional consequences in erythropoiesis , we analyzed erythroid cells in E12 . 5 fetal livers from wild-type and Δ-1 . 8 mice . Total cell numbers from wild-type and mutant fetal livers were similar ( Figure 2C ) . Cytospins of peripheral blood and fetal liver cells from wild-type and mutant E12 . 5 embryos had similar appearance upon May-Gruenwald-Giemsa staining ( Figure S2B ) . At this stage in development , most of the embryonic blood is comprised of primitive erythroid cells . However , some enucleated definitive cells were detected in both wild-type and Δ-1 . 8 embryos ( Figure S2B ) . Hematopoietic colony assays from wild-type and Δ-1 . 8 E14 . 5 fetal livers revealed that the total number of colonies and lineage distribution of colony types ( representing multipotential and lineage-restricted progenitors ) were similar ( Figure S2C ) . Examination of cells spanning different stages of erythroid development revealed no difference in the absolute number of Stage I , Stage II or Stage IV erythroid progenitors . However , the absolute number of Stage III erythroid progenitors was increased significantly ( p≤0 . 05 ) in the Δ-1 . 8 mice ( Figure 2D ) . These results demonstrate that at E12 . 5 , Stage III progenitors from Δ-1 . 8 mice expand relative to both their precursors and progeny , implying a block in the Stage III to Stage IV transition . The increased number of Stage III progenitors is in accordance with other models of ineffective erythropoiesis , in which impairment of erythroid cell maturation is accompanied by a compensatory increase in earlier red blood cell precursors [27] . The timing of this block corresponds to the stage at which Gata2 is reactivated ( Figure 2B ) , indicating that Gata2 dysregulation perturbs erythroid development . To examine this further , we utilized red blood cell enucleation as a cellular read-out of erythroid differentiation . Enucleation was measured using Draq5 to quantitate DNA content in Stage IV cells from wild-type and mutant embryos ( a representative FACS plot is shown in Figure S2D ) . Stage IV cells in Δ-1 . 8 embryos contained a significantly reduced ( >2-fold , p≤0 . 05 ) proportion of enucleated cells compared to those from wild-type , demonstrating that mutant cells fail to differentiate efficiently upon reactivation of Gata2 expression ( Figure 2E ) . We reasoned that aberrant expression of GATA-2 target genes in Δ-1 . 8 mice might underlie the block in the transition from early to late erythroblasts . Increased Gata2 expression could reactivate GATA-2 target genes expressed in early erythropoiesis , including those associated with proliferation , at a stage in which cells should exit the cell cycle . Alternatively , increased Gata2 expression could aberrantly repress late erythroid genes necessary for efficient differentiation . Finally , abnormal reactivation of Gata2 expression in cells expressing GATA-1 and other transcription factors involved in specifying alternate lineage programs could lead to the aberrant transcription of non-erythroid genes . To distinguish among these possibilities , we quantified gene expression in fetal liver erythroid cells from E12 . 5 mice . Several gene expression changes were apparent in Stage IV erythroblasts ( Figure 3A ) . Expression of Gata1 and Eraf , a globin chain stabilizing protein , were reduced by ∼40% ( p≤0 . 01 ) and ∼50% ( p≤0 . 05 ) , respectively , in the late erythroblasts of Δ-1 . 8 versus wild-type mice . In contrast to Gata1 and Eraf , most late erythroid genes examined , including the transcription factors Scl , Eklf , and the heme synthesis enzyme Alas2 , were expressed at similar levels , indicating that erythroid genes are differentially sensitive to Gata2 reactivation . Whereas expression of β-like globin genes ( Hbb-y , Hbb-bh1 , Hbb-b1 ) was normal ( Figure 3B ) , expression of α-globin ( Hba-a1 ) was reduced by 50% ( p≤0 . 05 ) and ζ-globin ( Hba-x ) was increased by 2-fold ( p≤0 . 05 ) ( Figure 3B ) . We also examined two genes expressed early in erythropoiesis . Both cMyb and the established GATA-2 target cKit were upregulated 4-fold ( p≤0 . 05 ) ( Figure 3C ) . In mast cells and megakaryocytes , GATA-2 is expressed in combination with other transcription factors including SCL and GATA-1 . As GATA-2 is aberrantly coexpressed with these factors in the Δ-1 . 8 erythroblasts , we examined select GATA-2 target genes from the mast cell and megakaryocyte lineages in our wild-type and mutant erythroblasts . Cpa3 , active in mast cells , and cMpl , expressed in megakaryocytes , were upregulated 4- and 2-fold , respectively , in mutant versus wild-type erythroblasts ( p≤0 . 05 ) ( Figure 3D ) . These results indicate that Gata2 reactivation is coupled with aberrant GATA-2 target gene expression . Given the dysregulation of genes associated with early progenitor proliferation , erythroid maturation , and alternate lineage fate , it is likely that these factors contribute in aggregate to the block in erythroid development . In contrast to the E12 . 5 fetal liver , erythroid progenitors isolated from the bone marrow of adult wild-type and Δ-1 . 8 mutant mice ( Stage II–IV ) did not reveal differences in Gata2 expression ( data not shown ) , indicating that Gata2 transcription is differentially regulated during fetal and adult erythropoiesis . Adult erythropoiesis has several unique attributes relative to the fetal process , including differences in proliferative capacity and rate of transit through the differentiation program [28] , [29] . Such differences might explain the ontogenic specificity of Gata2 reactivation . We reasoned that stress erythropoiesis in the adult , which resembles fetal liver erythropoiesis [28]–[32] , might shift the regulation of Gata2 expression to a state mimicking that in the fetus . To establish stress erythropoietic conditions , peripheral anemia was induced through phenylhydrazine-mediated red blood cell lysis . Examination of erythropoietic recovery in Δ-1 . 8 mice revealed no differences in hematocrit , implying that there is no deficiency in recovery from acute anemia in these mice ( data not shown ) . However , analysis of erythropoietic progenitor production in the bone marrow during recovery revealed that the absolute number of Stage III erythroid progenitors was significantly increased in Δ-1 . 8 mice ( p≤0 . 05 ) ( Figure 4A ) , while the number of Stage IV erythroid progenitors was similar , again indicating a block in the transition from Stage III to Stage IV . The increased number of Stage III cells is likely an indirect effect due to the increased sensitivity of Δ-1 . 8 mice to stress-induced ineffective erythropoiesis [27] . Expression analysis of sorted populations from the bone marrow of these mice showed that Gata2 transcription is increased significantly ( p≤0 . 05 ) in Stage IV cells from Δ-1 . 8 mice ( Figure 4B ) . These results mimic those obtained with E12 . 5 fetal liver ( Figure 2B , D ) , indicating that the −1 . 8 kb site controls Gata2 expression in both stress and fetal erythropoiesis . Gata2 is transcribed from two alternate promoters , termed 1S and 1G , leading to two transcripts with different first exons [33] . To determine whether the loss of Gata2 repression in Δ-1 . 8 erythroid cells ( Figure 5A ) reflects increased transcripts derived from one or both of the promoters , we used primers specific for mature forms of the 1S and 1G transcripts . The majority of Gata2 transcripts expressed in Stage I were derived from the 1G promoter and were repressed in Stage II–IV similarly in wild-type and Δ-1 . 8 cells ( Figure 5B ) . mRNA expression from the 1S promoter was increased nearly 8-fold ( p≤0 . 05 ) in Δ-1 . 8 Stage IV cells relative to Stage I cells . While wild-type cells exhibited increased 1S-derived mRNA at Stage IV relative to Stage I , this increase was significantly smaller ( Figure 5C ) . Quantitation of primary , unspliced transcripts derived from the 1S promoter revealed an even more striking increase in 1S-derived transcript from Δ-1 . 8 Stage IV cells ( ∼10-fold relative to Stage I ) compared to wild-type cells from the same stage , which did not demonstrate any appreciable increase ( p≤0 . 05 ) ( Figure 5D ) . Together , these results demonstrate that loss of the −1 . 8 kb site selectively reactivates transcription from the 1S promoter in erythroid cells . As expected , quantitative chromatin immunoprecipitation ( ChIP ) analysis of E14 . 5 fetal liver cells demonstrated reduction of GATA-1 occupancy at the −1 . 75 kb ( used as a surrogate for measuring occupancy at the deleted −1 . 8 kb site ) and the −2 . 8 kb sites ( p = 0 . 057 and 0 . 058 respectively ) and the proximal GATA-binding regions at −3 . 9 kb , ( p≤0 . 01 ) ; occupancy was not significantly altered at the distal −77 and +9 . 5 kb sites ( Figure 6A ) . ChIP analysis of Pol II demonstrated significantly increased occupancy at all sites examined upon mutation of the −1 . 8 kb site , with notable increases at the −77 kb enhancer ( p≤0 . 01 ) , the −1 . 75 kb site ( p≤0 . 01 ) and the 1G promoter ( p≤0 . 01 ) ( Figure 6B ) . Importantly , Pol II occupancy of a distant gene ( RPII215 ) did not change upon loss of the −1 . 8 kb site ( Figure 6B ) , providing evidence for locus specificity . ChIP analysis of GATA-2 occupancy yielded signals near background levels , consistent with GATA-2 expression being below the limit of detection in this assay ( data not shown ) . Average preimmune values for the wild-type and Δ-1 . 8 cells were 0 . 0018±0 . 00027 and 0 . 0041±0 . 0017 , respectively . To analyze histone modifications within the erythroid populations in which we observed an altered phenotype on removal of the −1 . 8 kb site ( Figure 2 ) , we performed quantitative ChIP on sorted fetal liver Stage III and IV cells . We quantitated dimeH3K4 and trimeH3K27 , two marks shown to be associated with repression at the Gata2 locus [17] , [21] , [34] . dimeH3K4 was significantly reduced at the −1 . 75 kb site , neighboring proximal regulatory regions , and the 1S promoter in both Stage III and Stage IV Δ-1 . 8 cells ( p≤0 . 05 ) ( Figure 7A ) . The repressive mark trimeH3K27 was decreased to a small extent at the promoters in Stage III Δ-1 . 8 cells ( p≤0 . 05 ) ( Figure 7B ) . Preimmune values were similar between wild-type and −1 . 8 samples ( Figure 7C ) . These results in primary erythroid progenitors provide direct evidence that the −1 . 8 kb cis element contributes to the maintenance of the dimeH3K4 mark in erythroid cells . Contribution of the dimethylH3K4 modification to transcriptional regulation is incompletely understood [35]–[37] . By contrast , the trimethylH3K4 mark is thought to play a critical role in promoting gene activation [38] , [39] . Also , recent attention has been focused on the monomethylH3K4 mark as an important regulator of enhancer elements [38] , [40] . We reasoned that loss of dimethylH3K4 might play an indirect role by providing a substrate for increases in the mono- or trimethyl forms of H3K4 . However , ChIP using E14 . 5 whole fetal liver cells revealed that the levels of trimeH3K4 were unchanged at all sites examined ( Figure S3A ) . Even more strikingly , the levels of monomeH3K4 were reduced at the −2 . 8 ( p≤0 . 05 ) and −1 . 75 kb sites ( p≤0 . 01 ) , as well as the 1G promoter ( p≤0 . 05 ) ( Figure S3B ) , similar to the reduction in dimeH3K4 observed in whole fetal liver cells ( data not shown ) and in sorted cells ( Figure 7 ) . Total H3 and preimmune values for ChIP using whole fetal liver cells were similar between wild-type and Δ-1 . 8 samples ( data not shown ) . These data indicate that loss of GATA-1 binding at the deleted −1 . 8 kb cis element leads to decreased GATA-1 occupancy at sites up to several kilobases away , reductions in dimeH3K4 and monomeH4K4 marks in the regulatory regions , and increased RNA Pol II occupancy . We propose a model in which this altered nucleoprotein structure favors a transcriptionally active locus , thereby permitting Gata2 reactivation . The Gata2 locus contains four CpG islands [17] located at the −2 . 8 kb GATA-binding region , both the 1S and 1G promoters , and an unclassified region between these promoters ( Figure 8A ) . Stable repression at loci characterized by CpG-rich promoters is thought to depend , in part , upon methylation of these promoters [4] , [41] . In addition , tissue specific gene silencing of Gata2 has been correlated with promoter methylation in some tissues [42] , [43] . Thus , we tested whether methylation of the 1S promoter is important for stable repression in erythroid cells and whether the −1 . 8 kb cis element maintains repression through such a mechanism . Bisulfite sequencing was utilized to quantitate promoter methylation of a 3′ section of the 1S CpG island within sorted populations of Stage II–IV erythroid progenitors from wild-type and mutant mice . In wild-type mice , the CpG island located at the Gata2 1S promoter was largely unmethylated in Stage II , Stage III , and Stage IV progenitors , with an average methylation of 5 . 2% , 8 . 9% , and 7 . 1% , respectively ( Figure 8B ) . As no specific residues were hypermethylated ( Figure S3C ) , these data imply that methylation of the 1S CpG island is not important for maintenance of repression in these cells . In Δ-1 . 8 mice , the 1S CpG island displayed similar levels of methylation in Stage II , Stage III , and Stage IV progenitors ( 5 . 9% , 8 . 2% , and 7 . 1% , respectively , Figure 8B ) . Thus , the stable repression of Gata2 does not require DNA methylation of the 1S CpG island , and the −1 . 8 kb site maintains repression in Stage IV erythroblasts through other mechanisms including regulation of transcription factor occupancy , histone modifications , and Pol II access .
We have described a loss of function strategy in mice to establish definitively whether one of the cis elements previously implicated in the control of Gata2 [17] , [19] , [21] functions independently or collectively with other cis elements to regulate Gata2 transcription in vivo . Unexpectedly , the endogenous −1 . 8 kb site is dispensable for activation of Gata2 and the initiation of repression , but instead selectively maintains Gata2 repression in terminally differentiating cells . Deletion of the −1 . 8 kb site reactivates Gata2 expression resulting in an erythroid maturation block , likely due to improper regulation of the reciprocally controlled Gata1 , genes involved in globin synthesis , genes expressed earlier in erythropoiesis , and genes associated with other hematopoietic lineages . While one or more additional GATA-binding sites in the locus must contribute to the initiation of repression , we propose that maintenance of repression is mediated through GATA-1 binding at the −1 . 8 kb site of the Gata2 1S promoter . In a wild-type setting , transcription factor binding and histone modifications lead to Pol II expulsion in a locus-wide manner to establish stable repression ( Figure 8C ) . In the absence of the −1 . 8 kb cis element , GATA-1 occupancy is lost at this site . Our results demonstrate that locus-wide Pol II expulsion requires maximal GATA-binding at the 5′ proximal regulatory regions , highlighting a critical role for the −1 . 8 kb site in regulating Pol II occupancy . The loss of GATA-1 occupancy in the absence of the −1 . 8 kb site results in a reduction in one of the marks associated with repression at this locus , dimeH3K4 , while having minimal effects on another repressive mark , trimeH3K27 . Intriguingly , dimeH3K4 decreases in a manner consistent with expectations from studies in cultured cells [17] , [21] . While this mark is commonly associated with activation in most contexts , recent genome-wide analysis studies have implicated this mark in both activation and repression [38] , [39] , and therefore our understanding of the functional consequences of this mark seems incomplete [35]–[37] . Two possibilities may account for the similarity of the dimethylH3K4 level in Δ-1 . 8 cells between repressed ( III ) and reactivated ( IV ) stages . First , dimethylH3K4 may not be the critical modification mediating maintenance of repression . Alternatively , other stage-specific factors in the nuclear milieu may lead to differential sensitivities to dimethylH3K4 between the repressed ( III ) and reactivated ( IV ) stages . Substantial reduction in both dimeH3K4 and monomethylH3K4 were observed upon loss of the −1 . 8 kb site without a concomitant increase in trimethylH3K4 . These findings suggest that the methylation states of H3K4 are regulated independently and locally through complexes recruited to the −1 . 8 kb GATA-binding site . These observations are in accordance with the finding that dimeH3K4 positive , trimethylH3K4 negative , marks are present at a subset of developmentally regulated hematopoietic genes [44] . Thus , our data highlight a potential role for these H3K4 marks in regulating transcription . It is interesting to note that the trimethylK27 mark , associated with GATA-1-mediated repression of the Gata2 locus [34] , is not affected by the −1 . 8 kb GATA-binding site . In addition , reduction of H3K27 trimethylation , widely accepted as a repressive mark [45] , does not appear to be required to reactivate gene expression at the Gata2 locus , perhaps indicating that it is involved selectively in the initiation of repression . Recent genome-wide analysis has also shown that H3K27 methylation is not merely present or absent , but rather increases quantitatively as the activity of the gene decreases [38] , [40] , suggesting that the level of transcriptional reactivation observed is within the range allowed by the H3K27 methylation level at this locus . Finally , in many cases , CpG rich promoters require hypermethylation of associated CpG islands for stable repression [4] , [37] . We find that the CpG island at the Gata2 1S promoter lacks high levels of methylation during stable repression , and that loss of the −1 . 8 site does not affect methylation levels . This data further supports a model in which −1 . 8 kb site-dependent histone marks maintain stable repression . We propose therefore that loss of GATA-1 binding and key repressive marks , including dimethyl- and monomethyl-H3K4 , result in a locus permissive for Pol II occupancy and reactivation of transcription . This model predicts that a specific protein or proteins are recruited by GATA-1 to the −1 . 8 kb site to maintain repressive chromatin structure . GATA-1 is known to interact with CBP [46] , HDACs 1 and 2 [47] , [48] , LSD1 [49] , BRG1 [50] , and polycomb repressive complex 2 ( PRC2 ) [34] . As no GATA-1-interacting proteins have been reported that possess the requisite methyltransferase activity to establish the dimeH3K4 histone mark that is lost in the −1 . 8 kb mutant , novel GATA-1-containing complexes may be required to maintain the −1 . 8 kb site-dependent histone marks . Ongoing genetic ablation studies examining the contribution of the other known GATA-binding regions to Gata2 regulation and local chromatin architecture will be important for understanding the control of this complex locus . Studies in multiple systems have led to a model of sequential gene repression during development [4] , separable into distinct phases . Reversible repression is replaced by epigenetic mechanisms that alter the chromatin structure at the locus though modifications of histones , and in some cases DNA , to maintain stable repression . The results described herein support such a model and characterize molecular mechanisms associated with the selective maintenance of repression of an endogenous target gene by an individual cis element to confer normal developmental control .
All animals were handled in strict accordance with good animal practice as defined by the relevant national and/or local animal welfare bodies , and the appropriate committee approved all animal work . Briefly , to generate the −1 . 8 kb knock-in allele , we replaced the palindromic GATA sites ( AGATAAGGCTTATCA ) with two SalI sites in order to clone in a Neo resistance cassette flanked by loxP sites . Once the neo cassette is removed , the locus contains a single loxP site flanked by SalI sites . The new sequence does not contain any binding motifs known to be involved in hematopoietic development . In more detail , we first inserted a HpaI site into pBSK between NotI and SacI with an oligo . Then , we cloned a −7 . 2 kb to intron 1S fragment of the Gata2 locus into pBSK with KpnI and HpaI . We then replaced the two palindromic WGATAR sites with a Sal I site via PCR and replaced the wild-type XbaI to NdeI fragment with this mutated version . Then , we cloned HSV-TK cloned into the SacII site of pBSK . Following this , we cloned a second SalI site into the XbaI site of pflox21 with an oligo and used the flanking SalI sites to clone this loxP-PGKneo-loxP cassette into the SalI site of the Gata2-containing pBSK ( Figure S1A ) . We screened targeted CJ7 ES clones by PCR and confirmed correct targeting by Southern blotting . We used standard blastocyst injection techniques to generate chimeric mice and screened F1 pups for germline transmission using Southern blotting ( Figure S1B , C ) . In some mice , the loxP-neomycin resistance gene was deleted by crossing with Gata1-Cre mice , which were of CD1/Swiss-Webster background . We confirmed Cre-mediated excision of neo from these mice using PCR and all further genotyping was performed by PCR ( Figure 1C ) . Mice were backcrossed onto a C57/Bl6 background for a minimum of 6 generations and were housed in a specific pathogen-free animal facility . Fetal liver cells were obtained from embryos at E12 . 5 and E14 . 5 after timed matings . Mouse bone marrow cells were obtained from 8- to 12-week-old animals by crushing femurs and tibias with either Iscove modified Dulbecco medium ( IMDM ) or Phosphate Buffered Saline ( PBS ) supplemented with 2% fetal calf serum ( Mediatech , Herndon , VA ) . Single cell suspensions of fetal livers and spleens were made by passage through 70 micron nylon mesh ( Sefar America , Kansas City , MO ) in PBS supplemented 2% fetal calf serum ( Mediatech , Herndon , VA ) . Cells were kept on ice until use and counts were performed using a Beckman Coulter AcT10 hematological analyzer . RNA was prepared from the described populations using the Trizol Kit ( Invitrogen , San Diego , CA ) , DNAseI treated by RQ1 RNase-Free DNase ( Promega , Madison , WI ) and quantified . cDNA was synthesized using 1 µg of RNA with the iScript cDNA Synthesis Kit ( Biorad , Hercules , CA ) . Typically , 1 µl of cDNA was then used as a template for quantitative PCR using the iQ SYBR Green Supermix ( Biorad , Hercules , CA ) in an iCycler thermocycler ( Biorad , Hercules , CA ) . Primer sequences can be found in Text S1 . Triplicate data sets were generated and results were normalized to β-actin reactions run in parallel . Whole PB was analyzed on a Beckman Coulter AcT10 hematological analyzer . White blood cell and progenitor subsets were analyzed from peripheral blood by staining with Gr-1 and Mac1 or CD3 and B220 after red blood cell lysis using ACK ( NH4Cl ) lysis buffer . All antibodies for FACS were obtained from Pharmingen ( San Diego , CA ) or eBiosciences ( San Diego , CA ) , and the following clones were used; Ly-76 ( Ter-119 ) , CD71 ( C2 ) , CD117 ( 2B8 ) . Antibodies to surface markers of interest were used at 1∶60 dilution and after 30–60 minutes unbound antibody was washed away . In the case of biotinylated antibodies , streptavidin conjugated to various fluorochromes was added for the last 15–30 minutes of antibody incubation at 1∶100 dilution . For cell sorting experiments of erythroid progenitor subsets , fetal liver cells were stained with antibodies to CD71 and Ter119 , and 7AAD was added to allow for exclusion of dead cells during sorting . For examination of enucleation , cells were stained with CD71 and Ter119 as above and incubated with Draq5 ( Biostatus Limited , Leicestershire , United Kingdom ) as per manufacturers instructions before analysis . Rabbit anti-GATA-1 and anti-GATA-2 polyclonal antibodies have been described previously [16] , [21] , [51] . Rabbit anti-Pol II ( N-20 , sc-899 ) was from Santa Cruz Biotech . Rabbit anti-acetyl-histone H3 ( #06-599 ) , anti-trimethyl-histone H3 ( Lys 9 ) ( #07-442 ) , anti-trimethyl-histone H3 ( Lys 27 ) ( #07-449 ) and anti-dimethyl-histone H3 ( Lys 4 ) ( #07-030 ) were from Millipore . Real-time-PCR-based quantitative chromatin immunoprecipitation ( ChIP ) analysis was conducted as described [52] . Single-cell suspensions were isolated from E14 . 5 wild-type and Δ-1 . 8 fetal liver cells , respectively , and crosslinked with 1% formaldehyde . Samples were analyzed by real-time PCR ( ABI Prism 7000 ) using primers designed by PrimerExpress™ 1 . 0 software ( PE Applied Biosystems ) to amplify regions of 75–150 bp that overlap with the appropriate motif . Product was measured by SYBR Green fluorescence in 20 µl reactions , and the amount of product was determined relative to a standard curve generated from titration of input chromatin . Calculations were derived using percentage of input and were normalized using relative units which were determined by defining 9% input sample as 1 . 0 . Analysis of dissociation-curves post-amplification showed that primer pairs generated single products . Bisulfite treatment of genomic DNA was performed as previously described using the Qiagen EpiTect Bisulfite Kit as per the manufacter's instructions . Sequence-specific PCR of the bisulfite-treated DNA was performed using primers specific to the murine Gata2 1S promoter ( outside primers: F , 5′- TTGTGTGGTGAGGGTGTAG-3′ , R , 5′- CAAATTTCTTTCCCTATTTTCT-3′; inside primers: F , 5′- TAGGTGGGGGAGAGTGTAG -3′ , R , 5′- CAAATTTCTTTCCCTATTTTCT -3′ . The PCR fragments were sub-cloned into the pCR®2 . 1-TOPO® vector ( Invitrogen ) and transformed into DH5α E . coli cells . Miniprep plasmid DNA was verified by EcoRI digestion and positive clones were sequenced using M13 forward ( −20 ) or reverse primers . Data are presented as mean ± SEM . Statistical significance was assessed by two-sided Student's t-test . | Different cell types are formed and maintained by proteins called transcription factors that directly bind to specific DNA sequences to activate or repress gene expression . While numerous DNA sequences bound by transcription factors are established , many questions remain unanswered regarding how they function at specific sites located at distinct chromosomal regions . As a model to study this process , we examined the regulation of a gene controlling red blood cell development , Gata2 , by the transcription factor GATA1 . In the DNA sequence upstream of Gata2 , there are several sites that GATA1 is known to bind to; however , it is unclear whether these binding sites work together or independently to control expression of Gata2 . To study this , we engineered mice to specifically remove one of these GATA1-binding sites . We found that removal of this single site reactivated expression of Gata2 in a specific stage of red blood cell development where Gata2 is normally not expressed , caused a block in differentiation of these cells , and changed the histone modification pattern specifically in the region upstream of Gata2 . This work supports a model in which individual transcription factor binding sites within regions of multiple binding sites can independently and distinctly regulate gene expression during development . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"genetics",
"and",
"genomics/epigenetics",
"developmental",
"biology/developmental",
"molecular",
"mechanisms",
"molecular",
"biology/transcription",
"initiation",
"and",
"activation",
"genetics",
"and",
"genomics/gene",
"expression"
] | 2010 | A Single cis Element Maintains Repression of the Key Developmental Regulator Gata2 |
Crimean-Congo hemorrhagic fever ( CCHF ) is a severe tick-borne disease well recognized through Europe and Asia where diagnostic tests and medical surveillance are available . However , it is largely neglected in Africa , especially in the tropical rainforest of Central Africa where only sporadic human cases have been reported and date back to more than 30 years . We describe here an isolated human case that occurred in the Democratic Republic of the Congo in 2008 and performed phylogenetic analysis to investigate whether it resulted from a regional re-emergence or from the introduction of a novel virus in the area . Near complete segment S and partial segment M sequences were characterized . Bayesian phylogenetic analysis and datation were performed to investigate the relationship between this new strain and viral strains from Africa , Europe and Asia . The new strain is phylogenetically close to the previously described regional genotype ( II ) that appears to be specific to Central Africa . Phylogenetic discrepancy between segment S and M suggested genetic exchange among local sublineages , which was dated back to 130–590 years before present . The phylogenetic analyses presented here suggest ongoing CCHF virus circulation in Central Africa for a long time despite the absence of reported human cases . Many infections have most probably been overlooked , due to the weakness of healthcare structures and the absence of available diagnostic procedure . However , despite the lack of accurate ecological data , the sporadic reporting of human cases could also be partly associated with a specific sylvatic cycle in Central Africa where deforestation may raise the risks of re-emergence . For these reasons , together with the high risk of nosocomial transmission , public health authorities' attention should be drawn to this etiological agent .
Crimean-Congo hemorrhagic fever virus ( CCHFV , family Bunyaviridae , genus Nairovirus ) is a tick-borne virus . It causes severe illness throughout Africa , Asia , Southeast Europe and the Middle East , with case fatality rates ranging from 3% to 30% . Its worldwide distribution closely matches that of its main arthropod vector , ixodid ticks belonging to the genus Hyalomma . Human infection occurs through tick bites , contact with infected livestock , or nosocomial transmission . The CCHFV negative-stranded RNA genome is divided into a small ( S ) , medium ( M ) and large ( L ) segment . Previous phylogenetic analysis of the S segment clustered strains into 6 to 7 distinct phylogeographic groups: West Africa in group I , Central Africa ( Uganda and Democratic Republic of Congo ( DRC ) ) in group II , South Africa and West Africa in group III , Middle East and Asia ( that may be split into 2 distinct groups Asia 1 and Asia 2 [4] ) in group IV , Europe and Turkey in group V , and finally Greece in group VI [1]–[5] . However , some of these phylogenetic lineages include strains separated by large spatial distances ( such as South Africa and West Africa ) suggesting viral migration , most likely via migratory birds transporting infected ticks , or secondary introductions following importation of commercial livestock . Comparative phylogenetic analysis revealed , with a few exceptions , parallel clustering of the S and L segments , while M segment reassortment seems more frequent [1] , [4]–[6] . During the last 60 years , CCHFV outbreaks have been described in Asia , the Middle East and the Balkans , where the virus has become endemic and caused several thousand human cases . During the last decade , CCHFV has caused human disease in previously unaffected countries ( Turkey 2002 , Iran 2003 , Greece 2008 , Georgia 2009 ) and has re-emerged in countries located southwest of the Russian Federation after an absence of nearly 30 years [7] . By contrast , fewer than 100 cases have been recorded in Africa [8] , most of them in South Africa [9] , [10] . In East and West Africa , enzootic CCHFV circulation has been shown by serological surveys of cattle and virus isolation from ticks since the 1970s [11] , [12] but until the outbreaks in Mauritania in 2004 [13] and Sudan in 2008 [14] , only sporadic human cases had been reported . In Central African Republic ( CAR ) , limited serological evidences of CCHFV circulation in Zebu cattle has been provided [15] and three viral strains were isolated from ticks between 1973 and 1976 , one of which lead to accidental infection of a laboratory worker [11] . Subsequent isolations from ticks were performed in the 80's [16] but no human case was reported . Despite the early identification of human CCHFV infection in DRC ( Kisangani , 1956 ) , CCHFV occurrence in Central Africa has not been much described and only sporadic human cases have been reported . One month after having isolated the first CCHFV strain ( strain Congo 3011 ) in newborn mice , Dr . Courtois became infected and this was the last notified case in DRC , from which the strain Congo3010 was isolated [17] , [18] . The virus was next identified in Uganda between 1958 and 1978 . Fifteen CCHFV strains were isolated from febrile patients , of which nearly half were laboratory workers having handled infectious samples [11] , [17] , [18] . From the geographic information associated with the other patients , it can be inferred that CCHFV was present both in the Entebbe area and in the Arua district ( previous West Nile district ) located 350 km North , near the border of Sudan . Three CCHFV strains were also isolated from ticks and an early serological survey suggested cattle infection [11] . No other epidemiological or ecological information is available on CCHFV in Central Africa or its borders , and no further cases have been recorded . In 2008 , CIRMF ( Centre International de Recherches Médicales de Franceville , Gabon ) identified CCHFV in a serum sample received for etiological diagnosis of a case of hemorrhagic fever in DRC . This is the only identification of CCHV in DRC for more than 50 years . To determine whether it was due to introduction of a novel virus or to re-emergence of a local genotype , we determined the phylogenetic relationships between this virus ( hereafter referred to as Beruwe-2008 ) and previously described isolates . Phylogenetic analysis showed that the Beruwe-2008 strain belonged to the genotype previously identified in this area and thus suggested that it had re-emerged . Local CCHFV persistence may have been supported by a sylvatic natural cycle specific to Central Africa , indicating that countries subject to major deforestation may note an increasing number of human infections .
Laboratory investigations were performed subsequently to the WHO request for surveillance and early alert of hemorrhagic fever outbreak in Central Africa . Because of the emergency settings associated with the suspicion of such acute illnesses , no ethics committee approval or written consent was deemed necessary . The blood sample was taken by national healthcare workers of the Lubutu hospital where the patient came for medical care . He was informed that his blood sample will be further used for diagnostic investigation and gave his verbal consent . The patient described here is anonymous . The blood sample was next sent to the Institut National de Recherche Biomédicale ( Kinshasa , DRC ) and then to CIRMF upon WHO authorities . The study was approved by the scientific committee of CIRMF . The patient was a 26-year-old man living in Beruwe ( Nord Kivu province ) in DRC , 325 km from Kisangani ( Figure 1 ) . He became ill in the mining area where he worked . He complained of fever and headache on day 1 and developed moderate bloody diarrhea on day 2 . Epistaxis , oral bleeding and hematuria occurred on day 3 . He treated himself with ibuprofen and paracetamol during the first three days . On day 4 after onset he additionally took quinine and finally presented with severe asthenia and persistent bleeding to Lubutu hospital , where the serum sample was taken . At this stage the patient was subicteric , with bleeding at the venipuncture site , but had only low-grade fever ( 37 . 6°C ) . He declared no contact with wild animals during the previous three weeks but he had slept in the forest . No further information on his outcome was available . The patient's serum was manipulated in biosafety level 4 ( BSL-4 ) conditions . The serum was first tested for Ebola and Marburg viruses . As results were negative , investigations were next performed for CCHFV . RNA was extracted with the QIAamp viral RNA mini kit ( Qiagen , Courtaboeuf , France ) according to manufacturer's instructions . Reverse transcription ( RT ) and real-time PCR amplification were performed with the High Capacity cDNA RT kit and Taqman universal PCR master mix ( Applied Biosystems - Life Technologies Corporation , Carlsbad , California ) , and previously reported primers and probes [19] . Conventional one-step RT-PCR was performed with CCHFV primers as previously reported [20] and with SuperScript III one-step RT-PCR system and Platinum Taq DNA polymerase ( Invitrogen -Life Technologies Corporation , Carlsbad , California ) . This yielded a 536-nucleotide fragment in the S segment , sequencing of which confirmed CCHFV identification . As viral isolation on Vero cells was unsuccessful , viral RNA was extracted from the patient's serum as described above , and was used for RT-PCR amplification with Platinum Taq DNA polymerase ( Invitrogen ) . Primers were derived from nucleotide alignments ( Table 1 ) . Three overlapping PCR products allowed near-complete characterization of the S segment coding sequence ( GenBank accession number HQ849545 ) and partial characterization of the M segment ( GenBank accession number HQ849546 ) . Amplification of the L segment was unsuccessful , being limited by the sample quantity . A total of 44 complete sequences for segment S and 38 complete sequences for segment M were retrieved from GenBank ( Table S1 in online supporting information ) . Nucleotides were aligned according to the amino-acid profile using the Translation Align algorithm implemented in Geneious software [21] . Initial phylogenetic analyses were performed with MrBayes V3 . 1 [22] , [23] using a GTR+gamma+invariant site substitution model for 4 million MCMC chain iterations sampled every 100 generations , corresponding to 40 000 trees ( data not shown ) . Following confirmation of the tree topology from MrBayes , the tip-dated coding alignments were submitted to Bayesian inference of node ages by using BEAST V1 . 4 . 7 [24] under the assumption of a codon-based substitution model ( SRD06 ) and an uncorrelated relaxed lognormal molecular clock and expansion , exponential and constant population growth models . The Expansion model yielded the best results , as indicated by ESS statistics and Bayes factor analysis of the posterior probability trace in TRACER . Sixty million generations were sampled every 1000 states , corresponding to 60 000 trees , that were annotated with TreeAnnotator and visualized with FigTree V1 . 3 . 1 from the BEAST package .
In 2008 we received a serum sample for etiological diagnosis of a case of hemorrhagic fever in DRC . The patient' serum was handled under BSL-4 facilities for RNA purification and tested positive for CCHFV by real-time PCR and conventional amplification with previously described detection systems [19] , [20] . The patient became ill in Beruwe , approximately 325 km from Kisangani , where the only 2 previously reported cases of CCHFV in DRC occurred in 1956 ( Figure 1 ) . The patient worked in a mining area near a forest environment and didn't seem linked to agro pastoral activities . As this was the only identified case of CCHV in DRC for more than 50 years , we performed a phylogenetic analysis to determine whether it was due to introduction of a novel virus or re-emergence of a local genotype . Virus isolation in Vero cells was unsuccessful , presumably owing to virus degradation subsequently to difficulties and delays of transportation . Genetic characterization was thus based on RT-PCR of RNA extracted from the patient's serum . As reassortment usually affects the M segment , priority was given to sequencing segments S and M , while segment L amplification was limited by sample quantity and was unsuccessful . Near-complete characterization of the segment S coding sequence was achieved , yielding 1501 contiguous nucleotides; the 5′ end was missing , presumably owing to RNA degradation . A 1001-nucleotide fragment was generated for segment M , corresponding to nucleotide positions 2382 to 3380 of the Congo3010-1956 glycoprotein coding sequence ( DRC strain ) . Pairwise nucleotide comparison of the Beruwe-2008 segment S sequence with those of the most closely related strains Congo3010-1956 ( DRC ) and Semunya-1958 ( Uganda ) – showed 92 . 4% and 92 . 0% similarity , respectively . In segment M the pairwise identities were 96 . 1% and 93 . 8% respectively . Identity between the Beruwe-2008 strain and strains belonging to other genetic groups ranged from 82 . 2% to 87 . 6% in segment S and from 72 . 5% to 81 . 3% in segment M ( Table 2 ) . Bayesian phylogenetic analysis with a molecular clock assumption was applied to segment S ( Figure 2A ) and M ( Figure 2B ) datasets . Both methods yielded tree topologies largely matching the phylogeographic groups previously defined from complete segments S and M [1]–[3] . In both segments , and with posterior probabilities reaching 1 , the Beruwe-2008 sequence grouped with the aforementioned DRC and Uganda strains forming lineage II ( Central Africa group ) . Although we cannot rule out the possibility of segment L reassortment , the Beruwe-2008 strain most likely belongs to the genotype previously identified in Central Africa , thus representing viral re-emergence rather than introduction of another genotype . In addition , the phylogenetic position of the Beruwe-2008 strain inside this Central African clade differed between the two segments , lying at the most ancestral branch in segment S while sharing a more recent common ancestry with the Congo3010-1956 strain from Kisangani in segment M . This is highly suggestive of intra-genotypic reassortment , thus implying co-circulation of these two DRC sub-lineages at this time . However , though it may be less probable , we cannot exclude definitively recombination between the two strains . Our dating analysis of the S segment resulted in time estimations slightly more recent than previously reported , but nonetheless within the same range and in keeping with an ancient origin of CCHF viruses [2] . The MRCA ( most recent common ancestor ) for the whole CCHFV species was estimated to have arisen 2518 years before present ( BP ) ( 95% High Posterior Density ( HPD ) : 820-5281 ) , the lineage II split-off was dated to 1484 years BP ( 95% HPD: 583-3389 ) and the MRCA of the three Central African strains was estimated at 587 years BP ( 95% HPD: 200-1327 ) . In the M segment , the MRCA estimates were slightly more recent , most probably owing to the use of partial rather than complete coding sequences and to different evolution of the two genes . This resulted in MRCA estimates of 1955 years BP for the whole species ( 95%HPD: 886-3844 ) , 221 years BP for the three Central African strains ( 95%HPD: 114-407 ) and 129 years BP ( 95% HPD: 75-228 ) for the two DRC strains . The genotype II split-off was estimated to have occurred 646 years BP , but the differences in the tree topologies prevented a true node age comparison with segment S . CCHV genotype II has been identified only in DRC and Uganda , while different CCHV lineages have been identified in neighboring countries to the north . Multiple genotypes have been identified in CAR , belonging to groups IV and III [2] , [20] , the latter also being encountered in Sudan [14] . By contrast no other genotype has been identified in Central Africa , for which reports on CCHFV are scarce and date back to 30 years . Hence , the data currently available suggest that genotype II is specific to central Africa . In DRC , CCHV has been reported only once , 50 years ago , but our data strongly suggest that the same genotype is still actively circulating . Of note , the MRCA estimates presented here are in agreement with ancient divergence of this lineage ( around 1000 years ago ) , but whether or not this split-off was linked to virus adaptation to Central Africa cannot be assessed . However , as the MRCA of the three strains was dated back to 683 to 243 years BP ( Figure 2A and B respectively ) , one might reasonably assume that the association of genotype II with this area goes back to this time period and thus did not result from very recent introduction . In addition , the co-circulation of different sub-lineages supports the possibility that ongoing CCHFV circulation occurred in the same area for some time . However , as the reassortment event would have taken place approximately 120 years BP , there is no evidence that CCHFV has been permanently circulating inside the Beruwe microhabitat , and we cannot exclude the possibility that this virus was very recently ( re ) introduced . In addition to the CCHFV genotypic specificity for Central Africa , its occurrence in the tropical rainforest contrasts strongly with the ecological characteristics of other areas in which CCHFV has been isolated [11] . Indeed , the enzootic distribution of CCHFV mostly coincides with temperate to dry or semi-dry climates in the forests , steppes and savannahs of Eurasia and West , East and South Africa . In these environments , domestic animals and their associated ticks are major agents of rural enzootic cycles affecting nearby human populations [11] , [25] . Despite the lack of accurate ecological data , the occurrence of CCHFV in Central Africa and its apparent genotypic specificity may suggest a distinctive sylvatic natural cycle in the deep tropical forest characterized by high rainfall , specific wildlife species , and a low density of domestic animals . Interestingly , co-speciation or long-term association with specific tick species has been previously suggested to explain the geographical distribution of CCHV genetic variants in Russia and Central Asia [26] . Such a sylvatic cycle , involving specific vectors and hosts with few contacts with human populations , could partly explain the lack of outbreaks and the sporadic nature of recorded human cases . In addition , as CCHFV is known to have been present in Central Africa for decades , and as human populations often live in isolated villages , many human infections may have been overlooked . However increasing invasion and destruction of rainforest habitats may lead to a higher risk of human CCHFV cases in future . Hence , despite 30 years without a single reported case , the data presented here suggest that CCHFV continues to circulate in Central Africa . More information on the epidemiology and the natural cycle of CCHFV in this ecosystem is required to assess its potential for emergence , notably in Gabon and Republic of the Congo . However health authorities and medical staff should be aware of the possibility of viral ( re ) emergence and of the high risk of nosocomial transmission . | Crimean-Congo hemorrhagic fever virus ( CCHFV ) is transmitted to humans through tick-bite or contact with infected blood or tissues from livestock , the main vertebrate hosts in a peri-domestic natural cycle . With numerous outbreaks , a high case fatality rate ( 3%–30% ) and a high risk for nosocomial transmission , CCHFV became a public health concern in Europe and Asia . However virus surveillance in Africa is difficult due to the limited sanitary facilities . Especially , CCHFV occurrence in Central Africa is very poorly described and seems highly in contrast with the temperate to dry environments to which the virus is usually associated with . We described a single human infection that occurred in Democratic Republic of the Congo after nearly 50 years of absence . The phylogenetic analysis suggests that CCHFV enzootic circulation in the area is still ongoing despite the absence of notification , and thus reinforces the need for the medical workers and authorities to be aware of the outbreak risk . The source of infection seemed associated with a forest environment while no link with the usual agro-pastoral risk factors could be identified . More accurate ecological data about CCHFV enzootic cycle are required to assess the risk of emergence in developing countries subjected to deforestation . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"and",
"Discussion"
] | [
"emerging",
"infectious",
"diseases",
"virology",
"emerging",
"viral",
"diseases",
"biology",
"microbiology",
"viral",
"evolution"
] | 2011 | Re-Emergence of Crimean-Congo Hemorrhagic Fever Virus in Central Africa |
Cell cycle control is modified at meiosis compared to mitosis , because two divisions follow a single DNA replication event . Cyclin-dependent kinases ( CDKs ) promote progression through both meiosis and mitosis , and a central regulator of their activity is the APC/C ( Anaphase Promoting Complex/Cyclosome ) that is especially required for exit from mitosis . We have shown previously that OSD1 is involved in entry into both meiosis I and meiosis II in Arabidopsis thaliana; however , the molecular mechanism by which OSD1 controls these transitions has remained unclear . Here we show that OSD1 promotes meiotic progression through APC/C inhibition . Next , we explored the functional relationships between OSD1 and the genes known to control meiotic cell cycle transitions in Arabidopsis . Like osd1 , cyca1;2/tam mutation leads to a premature exit from meiosis after the first division , while tdm mutants perform an aberrant third meiotic division after normal meiosis I and II . Remarkably , while tdm is epistatic to tam , osd1 is epistatic to tdm . We further show that the expression of a non-destructible CYCA1;2/TAM provokes , like tdm , the entry into a third meiotic division . Finally , we show that CYCA1;2/TAM forms an active complex with CDKA;1 that can phosphorylate OSD1 in vitro . We thus propose that a functional network composed of OSD1 , CYCA1;2/TAM , and TDM controls three key steps of meiotic progression , in which OSD1 is a meiotic APC/C inhibitor .
Meiosis is a key step in the life cycle of sexually reproducing eukaryotes , such as the majority of flowering plants . At meiosis ploidy is reduced by two , leading to the production of typically haploid gametes whose fusion during fertilization restores diploidy of the next generation . This is achieved by the modification of the meiotic cell cycle , compared to mitosis , allowing two rounds of chromosome segregation – meiosis I and meiosis II - after a single DNA replication event . Thus , a central question when analyzing the meiotic cell cycle is how three key transitions are controlled , i . e . entry of the meiocytes into meiosis I after prophase , transition from meiosis I to meiosis II and exit from meiosis II . The main driving force of cell-cycle progression , at both meiosis and mitosis , is the activity of cyclin-dependent kinases ( CDKs ) , in association with their regulatory partners , the cyclins . Entry into division phase requires high CDK activity that peaks at metaphase . Anaphase progression is regulated by a gradual degradation of CDK activity and mitotic exit requires low CDK activity [1] , [2] . CDK activity is regulated by the anaphase-promoting complex/cyclosome ( APC/C ) , a conserved multisubunit E3 ubiquitin ligase that triggers the degradation of multiple substrates , including cyclins , during mitosis and meiosis . The APC/C is activated by Cdc20/Fizzy and Cdh1/Fizzy-related proteins that also confer substrate specificity ( the latter is known as CCS52s in plants ) [3]–[5] . Precisely how the mitotic machinery is modified for the purpose of meiosis is unclear . The currently available knowledge that originates from studies carried out in unicellular fungi , Xenopus laevis and mouse oocyte systems , points towards a meiosis specific regulation of the APC/C as one of the key cell cycle modifications between meiosis and mitosis [2] , [3] . In oocytes , meiosis is driven by Cdc2/Cyclin B complexes . At the end of meiosis I , Cyclin B is only partially degraded and the residual , low level of Cdc2/CyclinB activity is essential for entry into meiosis II [6] . Partial Cyclin B degradation is obtained through temporally controlled inhibition of the APC/C by the Erp1/Emi2 protein [7] , [8] . In Schizosaccharomyces pombe , the Mes1 protein also partially restrains cyclin degradation through inhibition of the APC/C , thereby allowing entry into meiosis II [9]–[11] . In Saccharomyces cerevisiae , a meiosis specific APC/C activator ( Ama1 ) , and its inhibitor Mnd2 , are both required for meiotic progression [3] . Very little is known about control of the meiotic cell cycle in plants . It is largely unknown which one ( s ) of the Arabidopsis cyclins ( which include 10 A-type-cyclins and 11 B-type-cyclins ) constitute , with CDKA;1 [12]–[14] and possibly other CDKs , the core CDK complex that is necessary for meiosis . To date , only four genes involved in the three meiotic cell cycle transitions have been isolated in Arabidopsis thaliana . Two of them , TAM ( TARDY ASYNCHRONOUS MEIOSIS , also known as CYCA1;2 ) and OSD1 ( OMISSION OF SECOND DIVISION ) are essential for the meiosis I/meiosis II transition . The mutation of CYCA1;2/TAM or of OSD1 leads to a premature exit from meiosis after meiosis I , and thus to the production of diploid spores and gametes [15]–[18] . These two genes are also involved in the prophase/meiosis I transition as their concomitant loss leads to a premature exit from meiosis after prophase I , before any division [15] . CYCA1;2/TAM encodes one of the 10 Arabidopsis A-type cyclins [18] and OSD1 encodes a plant-specific protein , with additional functions in suppressing ectopic endomitosis via APC/C inhibition [15] , [16] , [19] . The third one , TDM ( THREE-DIVISION MUTANT ) , is essential for meiotic exit , as its mutation leads to entry into an aberrant third division of meiosis after regular first and second divisions [13] , [20] , [21] . Finally , SMG7 ( SUPPRESSOR WITH MORPHOGENETIC EFFECTS ON GENITALIA 7 ) is also required at the end of meiosis , as its mutation leads to cell cycle arrest at anaphase II [13] , [22] . Epistasis analysis suggest that SMG7 and TDM act in the same pathway [13] . Here we explored the meiotic molecular function of OSD1 and CYCA1;2/TAM , and the functional relationship between OSD1 , the APC/C , CYCA1;2/TAM and TDM to control meiosis progression .
OSD1 depletion leads to a premature exit from meiosis at the end of meiosis I , a phenotype reminiscent of the vertebrate Erp1/Emi2-depleted oocytes and the mes1 fission yeast mutant . While this work was in progress , evidence was found that OSD1 ( also named GIGAS CELL 1 , GIG1 ) negatively regulates the APC/C to control mitotic progression [19] . Yet , while the OSD1 protein has been shown to act as a mitotic APC/C inhibitor [19] and is well conserved in all plants , it does not appear to be conserved over other eukaryotes and notably does not show global similarity with other known APC/C inhibitors [16] , which conversely do not seem to have homologues in plants . However , closer examination of the OSD1 sequence revealed that OSD1 shares multiple features with Mes1: OSD1 has the same three putative cell-cycle-related domains in the same order on the protein ( Figure 1 ) . These three domains are very well conserved over OSD1 homologues ( Figure S1 ) [16] . Two of these domains are putative APC/C degradation motifs: a D-box ( residues 104–110 , RxxLxx[LIVM] ) and a GxEN/KEN-box ( residues 80–83 , GxEN in eudicotyledon and KEN in monocotyledon OSD1 homologues ) . The corresponding two motifs have been shown to be important for the Mes1 function [10] . OSD1 also has a C-terminal MR-tail in common with Mes1 ( the two last amino-acids of the protein are a methionine and an arginine ) . This MR-tail has not been functionally tested in Mes1 . However the MR-tail of Nek2a , a kinase that is involved in mitotic regulation via APC/C inhibition , has been described as being a docking domain of Nek2a on the APC/C , being thus essential for its binding and inhibition activities [23] . Similarly , the C-terminal RL-tail of Emi2 is essential for inhibition of the APC/C at meiosis [24] . These observations prompted us to propose that OSD1 might also promote meiotic progression by regulating the APC/C activity through these three domains . Using yeast 2-hybrid ( Y2H ) experiments Iwata et al [19] recently showed that OSD1 ( also called GIG1 ) interacts with the APC/C activator CDC20 . 1 , CDC20 . 5 , CCS52A1 and CCS52B , but not with the core APC/C components they tested ( APC2 , APC7 , APC10 , CDC27a , and HBT ) . We independently used Y2H experiments to test interaction of OSD1 with different APC/C subunits ( Figure 2A ) . Corroborating and extending Iwata et al results , OSD1 did not interact with any of the APC/C core subunits tested ( APC2 , CDC27a , HBT , APC4 , APC5 , APC6 , APC7 , APC8 , APC10 , APC11 ) . Concerning the activators , our result confirmed the interaction with CCS52A1 but did not reveal interaction with the other activators tested , including CDC20 . 1 that was scored positively by Iwata et al . As a negative result in Y2H experiments could be due to protocol and material variations , we used a complementary approach . Tandem affinity purification ( TAP ) experiments , using APC/C core components and the activators CCS52A2 , CCS52B and CDC20 . 1 as baits , previously identified OSD1 by mass spectrometry [25] . As mass spectrometry can fail to identify all proteins in a sample , we used an anti-OSD1 antibody ( Figure S2 ) on TAP purified samples using CDC20 . 1 , CDC20-3 and the three CDH1 homologues ( CCS52A1 , CCS52A2 , CCS52B ) as bait [25] , to test to presence of OSD1 . OSD1 was revealed in the CDC20 . 1 TAP ( but not CDC20-3 ) and the three CCS52 TAPs ( Figure 2B ) . Altogether our and Iwata et al results suggest that OSD1 can interact with a range of APC/C activators , including CDC20 . 1 , CDC20 . 5 , CCS52A1 , CCS52A2 and CCS52B . Next we asked whether the D-box , GxEN-box and MR tail represent true APC/C interaction motifs ( Figure 2C ) . For both the D-box and GxEN-box , the amino acid residues essential for APC/C binding were substituted to alanine ( ΔD , RxxL→AxxA; G , GxEN→AxAA ) whereas the MR tail was deleted ( ΔMR ) . We also mutated a putative second D-box motif ( ΔD′ , RxxL Aa 34–37 ) in OSD1 that is not conserved among the different plant proteins ( Figure S1 ) . All the OSD1 proteins were stably expressed in yeast ( Figure S3 ) . Mutation of the conserved D-box completely abolished the Y2H interaction with CCS52A1 . Deletion of the MR tail diminished , but not abolished the interaction with the APC/C activator . In contrast , mutation of D′ or of the GxEN-box did not reduce the interaction with CCS52A1 ( Figure 2C ) . To investigate the in vivo role of the APC/C interaction motifs revealed above , we created several versions of the genomic OSD1 gene ( including OSD1 promoter and terminator ) with a GxEN-box mutation ( OSD1ΔGxEN , GxEN→AxAA ) , a D-box mutation ( OSD1ΔD , RxxL→GxxV ) , a MR-tail mutation ( OSD1ΔMR , MR→* ) or combination of two or all of these mutations . None of these constructs modified the plant phenotype when introduced in wild type plants ( data not shown ) . We then introduced them in the osd1-3 mutant ( Figure 3 ) . The wild type genomic clone was able to restore normal meiosis ( i . e formation of tetrads ) of the osd1-3 mutant ( number of independent transformants n = 8 , 8/8 100% tetrads ) . In contrast , OSD1ΔMR could not restore tetrad formation ( n = 6 , 0% tetrad ) whereas OSD1ΔD barely complemented ( n = 6 , 0 to 15% tetrads ) . Albeit we cannot exclude that the introduced mutations destabilize the protein in planta ( though the modified OSD1 proteins accumulated at equal level when expressed in yeast and mouse oocytes ( Figure S3 , Figure 4A ) ) , these results indicate that the OSD1 D-box and MR-tail are important for OSD1 function . Correspondingly , the OSD1ΔDΔMR allele could not restore tetrad formation in osd1-3 ( n = 5 , 0% tetrads ) . In contrast , OSD1ΔGxEN almost completely reverted the osd1-3 mutant phenotype ( n = 3 , 82 to 93% tetrads ) , suggesting that the GxEN-box is not essential for the OSD1 function in planta . Strikingly , the OSD1ΔGxENΔD allele could complement osd1 mutants ( n = 4 , 83 to 94% tetrads ) , showing that deleting the GxEN-box in OSD1ΔD restored OSD1 function . OSD1ΔGxENΔMR and OSD1ΔDΔGxENΔMR did not complement osd1-3 ( n = 2 and n = 4 ) , showing that the MR tail is required in all situations ( Figure 3 ) . To further confirm that OSD1 is an APC/C inhibitor , we took advantage of the fact that - while OSD1 is not conserved in mammals - the APC/C and its activators are extremely well conserved . Thus , expression of OSD1 in a mammalian system - such as mouse oocytes- should equally interfere with APC/C activity and thereby disturb meiotic progression . OSD1 was stably expressed in mouse oocytes ( Figure 4A ) . Oocytes injected with mRNAs encoding OSD1 , but not control-injected oocytes , arrested at metaphase I with aligned chromosomes ( visualized through simultaneous injection of H2B-RFP ) ( Figure 4B ) . Chromosome spreads reveal the presence of bivalents indicative of a metaphase I arrest ( Figure 4C ) . This shows that OSD1 can inhibit the APC/C and prevent progression through meiosis I . Expression of OSD1ΔMR , OSD1ΔGxEN or OSD1ΔD ( Figure 4A ) did not provoke the metaphase arrest , showing that these three motifs are required for the APC/C inhibition by OSD1 in mouse oocytes ( Figure 4B and 4C ) . Only a few genes involved in control of the male meiotic cell cycle have been described in plants . Two mutants provoke premature exit before meiosis II – osd1 and cyca1;2/tam [15]–[17] . In contrast , tdm mutation prevents exit from meiosis and provokes entry into a third round of division ( meiosis III ) after meiosis II [13] , [20] , [21] ( Figure 5 ) . The tdm-3 mutant is a newly described T-DNA allele which has the same phenotype as previously described tdm mutants [20] , [21] . We studied the epistatic relationship between osd1-3 , tam-2 and tdm-3 during male meiosis ( Figure 6 , Figures S4 and S5 ) , completing prior work [13] , [15] . As previously described with different alleles [13] , the tam-2/tdm-3 double mutant had the same phenotype as tdm-3 , with a third division of meiosis ( Figure 6 and Figure S4 ) and complete male sterility ( Figure S4 ) . In clear contrast , the double mutant osd1-3/tdm-3 was male fertile ( Figure S5 ) and meiocytes exited meiosis before meiosis II , like in the single osd1-3 mutant ( Figure 6 and Figure S4 ) . Thus , depletion of TDM enables entry into meiosis II in tam but not osd1 mutants . SMG7 , which controls meiosis II exit through TDM [13] , exhibits the same epistatic relationship with CYCA1;2/TAM and OSD1 as TDM ( i . e osd1-3 is epistatic to smg7-1; which is epistatic to tam-2 [13] , data not shown ) . As we described previously with different alleles [15] , meiocytes in the osd1-3/tam-2 double mutant exit meiosis after a normal prophase I , without entering the first division ( Figure 6 and Figure S4 ) . In the triple mutant osd1-3/tam-2/tdm-3 all male meiocytes progressed through meiosis I but arrested at telophase I , before cytokinesis ( Figure 6 and Figure S4 ) leading to male sterility ( Figure S5 ) . Thus , mutating TDM allows osd1-3/tam-2 to enter and progress into meiosis I . Notably , in contrast to the situation for the single tam-2 mutant , mutating TDM in the osd1-3/tam-2 double mutant does not completely suppress the tam-2 defect , as the triple mutants are sterile and arrest at telophase I ( no cytokinesis ) whilst osd1-3 plants are fertile and exit from meiosis after telophase I . OSD1 contains 7 predicted CDK phosphorylation sites ( 4 [S/T]P and 3 [S/T]Px[R/K] ) , five of them being well conserved ( Figure S1 ) , suggesting that it could be the target of a CDK . Co-precipitation assays from E . coli expressed proteins showed that CYCA1;2/TAM binds to all three kinases: CDKA;1 , CDKB1;1 and CDKB2;2 , although CDKA;1 appears to have a higher affinity to CYCA1;2/TAM than the others ( Figure 7 ) . However , a subsequently performed kinase assay revealed that only CYCA1;2/TAM-CDKA;1 but not CYCA1;2/TAM-CDKB1;1 or CYCA1;2/TAM-CDKB2;2 is active against both OSD1 and the generic substrate histone H1 . These results suggest a regulatory interaction between CYCA1;2/TAM-CDKA;1 and OSD1 in meiosis . CYCA1;2/TAM is likely not the sole cyclin promoting meiosis progression because male meiosis continues until the end of the first division in tam-2 mutants . Female meiosis is less affected than male meiosis , as 60% of the female gametes are haploid , produced by a complete meiosis [15] . Among the 10 Arabidopsis A-type cyclins , CYCA1;1 is the most similar to CYCA1;2/TAM [26] and therefore a good candidate to have similar , possibly redundant , functions to those of CYCA1;2/TAM . We thus characterized an Arabidopsis line carrying a T-DNA insertion in CYCA1;1 ( cyca1;1-1 , see M and M ) which displayed no defects during meiosis and produced normal diploid progeny . Further , the double mutant cyca1;1-1/tam-2 exhibited the same meiotic phenotype and produced similar frequencies of haploid/diploid gametes as tam-2 ( 70% triploids and 30% tetraploids among the progeny of selfed double mutant ) . Hence , CYCA1;1 does not appear to have a meiotic function . Like many cyclins , CYCA1;2/TAM possesses a D-box [18] , a domain essential for cyclin destruction by the APC/C . We thus created a genomic version of the CYCA1;2/TAM gene , including endogenous promoter and terminator , with a D-box mutation ( TAMΔD , RxxL→GxxV ) . The corresponding wild type construct rescued the tam-2 meiotic defect ( n = 5 , 100% tetrads ) . In contrast , the introduction of TAMΔD in either wild type or tam-2 mutant ( n = 8 ) , generated a dominant effect on male and female meiosis . Plants containing the TAMΔD transgene produced only monads and were completely male and female sterile ( Figure 8 ) . No somatic phenotype was observed , strongly suggesting that CYCA1;2/TAM functions specifically at meiosis . Meiotic chromosome spreads showed that meiosis in TAMΔD plants progressed through meiosis I and meiosis II , up to telophase II ( ( Figure 8A–8D ) . But then , meiosis entered into an aberrant third division of meiosis , with stretched chromosomes dispersed throughout the cell , and no cytokinesis ( Figure 8E–8F ) . Immunolocalization of tubulin , confirmed that meiocytes expressing TAMΔD entered a third meiotic division , with the formation of four spindles ( Figure 9 ) , like previously shown for the tdm mutant [13] . When TAMΔD was introduced in tdm-3 ( n = 5 ) , meiosis progressed through meiosis I and meiosis II , and entered into the third division of meiosis typical of tdm or TAMΔD ( not shown ) . In contrast , when TAMΔD was introduced into osd1-3 ( n = 5 ) , meiosis progressed through meiosis I and arrested at telophase I , without entering meiosis II ( Figure 10 ) . Unlike single osd1-3 , cytokinesis did not occur ( Figure 10C ) , the meiotic product did not develop into pollen grains , and the plants were sterile ( Figure 10D ) . When UVI4 , the OSD1 paralogue , is mutated , an increase of somatic endoreduplication and no meiotic phenotype is observed [27] . It has also been recently shown that mutation of OSD1 , in addition to its meiotic consequences , triggers ectopic endomitosis [19] . To determine the interaction between OSD1 and UVI4 , we aimed to produce a double osd1/uvi4 mutant . However no double mutant was recovered from self-pollinating populations of osd1-1+/− uvi4+/− , osd1-2+/− uvi4+/− , or osd1-2+/− uvi4−/− ( 92 double mutants expected in total , Table S1 ) . This distortion indicates that the mutation of both OSD1 and UVI4 leads to gametophyte and/or embryo lethality . Reciprocal crosses between osd1-2+/− uvi4+/− or osd1-2+/− uvi4−/− plants and wild-type plants showed that transmission of osd1 and uvi4 through male gametophyte is regular but that the transmission of the double osd1/uvi4 mutant allele through female gametophyte is reduced by 80% ( Table S1 ) . We observed female gametophyte development in osd1-2+/− uvi4−/− plants , in which 50% of the gametophytes are expected to inherit the double osd1/uvi4 mutation . Wild type female gametophyte development includes three haploid mitotic events , leading to the formation of eight nuclei ( Figure 11A and 11C ) . In osd1-2+/− uvi4−/− , 58% ( n = 371 ) of the gametophytes showed wild type-like development , the others being blocked at a 1 cell ( 37% ) or 2 cell stage ( 5% ) ( Figure 11B ) . These arrested cells had a very large nucleus , with an increased DNA content ( compare Figure 11C and 11D ) , suggesting a defect in mitotic cell cycle . Both genetic and cytological data suggested that some osd1/uvi4 female gametophytes may be viable , prompting us to look further for double mutant plants . Among approximately 13 , 000 seeds of an osd1-2+/− uvi4−/− plant sown in vitro , 25 very abnormal plants were identified and confirmed by genotyping to be osd1-2−/− uvi4−/− . These plants , due to strongly affected growth , measured at most 2 cm after 5 weeks ( Figure 11E ) , while wild type plants of the same age were fully developed and about 30 cm high . Altogether , these results show that OSD1 , beyond its meiotic function has an essential function , redundantly with UVI4 , in gametophyte and somatic growth .
Our and previously published data suggest that a functional network between OSD1 , CYCA1;2/TAM and TDM controls three key transitions of meiosis ( prophase-meiosis I , meiosis I-meiosis II and meiosis II-exit ) : ( i ) OSD1 and CYCA1;2/TAM act in a synergetic manner to promote the transition from prophase to meiosis I , as the double mutant fails to enter meiosis I . TDM , appears to repress this transition as its mutation restores the entry into meiosis I of the osd1-3/tam-2 double mutant . ( ii ) OSD1 and CYCA1;2/TAM are crucial for the meiosis I-meiosis II transition as both single mutant exit meiosis before meiosis II . TDM also appears to repress this transition as its mutation allows the tam-2 mutant to enter meiosis II . Interestingly , OSD1 appears to be absolutely essential for the entry into meiosis II as osd1 mutants never entered this phase in all the backgrounds we tested . ( iii ) TDM and CYCA1;2/TAM are also involved in the exit from meiosis II , to prevent entry into a third meiotic division . Indeed , the mutation of TDM or the expression of a non-destructible version of CYCA1;2/TAM provokes the entry into a third division of meiosis . It is unclear if OSD1 could also be involved in this transition , as osd1 mutants never reach this stage . Further investigation is required to understand how this network is fine tuned to allow entry into a division after prophase and after meiosis I but to allow exit after meiosis II . Control of APC/C activity is fundamental for regulation of cell cycle progression . The APC/C is an E3 ubiquitin ligase that triggers the degradation of multiple proteins , including cyclins , at meiosis and mitosis [3] . Several APC/C inhibitors with crucial functions at mitosis or meiosis have been identified in various eukaryotes ( e . g . EMI1 and EMI2/ERP in vertebrates , Mes1 in S . pombe , Mnd2 and Acm1 in S cerevisiae ) , but these proteins are not conserved between kingdoms . An independent study recently showed that OSD1 negatively regulates the APC/C to prevent endo-mitosis during somatic development [19] . Here we propose that OSD1 , similar to Mes1 in S . pombe and Emi2 in vertebrates , promotes meiotic progression through APC/C inhibition . Indeed , OSD1 interacts directly with activator subunits of the APC/C , as shown by both TAP and Y2H experiments ( ours and Iwata et al results [19] ) . In addition , expression of OSD1 in mouse oocytes provokes meiotic arrest at metaphase I , consistent with APC/C inhibition . Similar to Mes1 , OSD1 contains three APC/C interaction domains , a D-box , a GxEN/KEN-box and a MR-tail . Both the interaction with the APC/C and the in planta meiotic function of OSD1 are dependent on its D-Box domain and its MR tail , suggesting that OSD1 inhibits the APC/C through direct binding with its active site or by sequestering its activators . Remarkably , the OSD1 GxEN-box is not required for OSD1 function , but its mutation allows the OSD1 protein mutated in its D-Box to fulfill its function . However , the three domains ( D-box , GxEN/KEN-box and a MR-tail ) are required to provoke meiotic arrest when OSD1 is overexpressed in mouse oocytes . The modulation of the cell cycle machinery that permits the entry into a second division without an intervening replication at meiosis seems to be fulfilled in various eukaryotes thanks to apparently evolutionary unrelated APC/C inhibitors , Mes1 in S . pombe , Emi2 in vertebrates and OSD1 in plants . Interestingly , OSD1 and Emi2 both have a paralogue in their respective genomes , UVI4 and Emi1 respectively , that play roles in the mitotic cell cycle through APC/C regulation [3] , [28] . However , contrary to Mes1 , OSD1 has also a somatic function as revealed by the mitotic phenotype of the single osd1 mutants [19] and the strong gametophytic and somatic defects of the osd1/uvi4 double mutant . The molecular function of TDM is unknown . However , four TPR domains ( AA 61–191 ) were predicted with high probability in TDM ( P = 8 . 0E-12 ) [29] . Further , using remote similarity searches via HHpred [30] , we found that CUT9 ( a TPR-containing APC/C component , appeared as the first hit ( Protein Data Bank entry 2xpi_A , E = 0 . 00095 ) . APC16 , another TPR-containing APC/C component also appeared among the first hits ( 3hym_B , E = 0 , 0022 ) . This raised the possibility that TDM may interact with or may be a component of the APC/C , and thus promotes meiotic exit via APC/C-mediated cyclin destruction . CDKA;1 appears to be a major cyclin-dependent kinase that drives meiotic progression in plants [12] , [13] . However , the cyclin ( s ) forming ( with CDKA;1 ) the predicted core cyclin/CDK meiotic oscillator has/have not been identified yet . Two cyclins have been shown to have an essential role at meiosis , CYCA1;2/TAM and SDS . However , SDS has been shown not to affect meiotic progression , but to regulate the choice of the partner of homologous recombination [31] , [32] . CYCA1;2/TAM being the sole known cyclin whose mutation affects meiotic progression , appeared to be a good candidate to fulfill part of this function . Indeed the fact that tam null mutants exit prematurely from meiosis supports this hypothesis . We also showed here that TAM is an active cyclin as it can form an active complex with CDKA;1 . However , prior evidence suggests that CYCA1;2/TAM may not be the core CDK oscillator that drives meiotic divisions [13] . Correspondingly , we showed here that the expression of a non-destructible CYCA1;2/TAM does not provoke a meiotic arrest at metaphase/anaphase I as may be expected for the core CDK oscillator , but induces entry into a third division . Strikingly , the phenotypes induced by the tdm mutation or the expression of non-destructible CYCA1;2/TAM appear identical , suggesting that TDM and TAM act in an antagonist manner to promote and prevent exit from meiosis , respectively . In addition , tdm is epistatic to tam-2 . Two hypotheses may account for these results ( Figure S6 ) . ( i ) CYCA1;2/TAM could be a negative regulator of TDM , which itself promotes meiotic exit , maybe through APC/C activation . ( ii ) Alternatively , another cyclin ( s ) ( distinct from CYCA1;1 as shown here ) might , together with CYCA1;2 , promote directly meiosis progression in a dose-dependent manner . The function of TDM could be to negatively regulate these cyclins , possibly through activation of the APC/C that would clear the cell from the remaining cyclins at the end of the meiotic program . Further work is required to discriminate between these hypotheses ( Figure S6 ) . We also showed that the CYCA1;2/TAM-CDKA;1 complex phosphorylates OSD1 , at least in vitro . This suggests that TAM could regulate OSD1 to prevent precocious meiotic exit . Alternatively , phosphorylation by TAM could inactivate OSD1 and thus allow exit from meiosis I . Interestingly , the activity and stability of Emi2/Erp1 - the vertebrate meiotic APC/C inhibitor - is regulated by phosphorylation [33] , [34] . Further functional analysis of the OSD1 putative phosphorylation sites is required to establish the role of this CYCA1;2/TAM-CDKA;1-mediated phosphorylation in meiotic cell cycle progression . These are the early days of meiotic cell cycle studies in plants , and already a complex regulatory network has emerged . Further studies are required to understand the control of meiotic progression in plants , and notably one of the next important goals is to establish which cyclin ( s ) constitutes the core meiotic progression oscillator and which activators of the APC/C ( among the five putative CDC20 and three CDH1 ) are involved in this complex variation of the cell cycle .
Arabidopsis plants were cultivated in greenhouse as previously described [35] or in vitro on Arabidopsis medium [36] at 21°C , under a 16-h to 18-h photoperiod and 70% relative humidity . For epistasis studies , we used osd1-3 , tam-2 and tdm-3 , alleles that are all in the same genetic background ( Col-0 ) , to prevent any genetic complication . The T-DNA of the tdm-3 mutant ( SALK_034202 ) is inserted in the second exon ( ATG+609 pb ) . In the cyca1;1-1 ( pst18025 ) , the T-DNA insertion is in the fifth exon ( ATG+1415 pb ) . The tdm-3 and cyca1;1-1 mutants were genotyped by PCR by two primer pairs . The first pair is specific to the wild type allele and the second to the left border of the inserted sequence . tdm-3: N534202U ( 5′- GGAGATCGAGTTGATAGTGC-3′ ) & N534202L ( 5′-ATACTAGGGAACTTGGGCT-3′ ) ; N534202U & LbB1 ( 5′-GCGTGGACCGCTTGCTGCAACT-3′ ) ; cyca1;1-1 : pst18025U ( 5′-TTGATTTGCTTGGTATTGCAG-3′ ) & pst18025L ( 5′–TGGTCGTCTTGTTGGGTCTAG-3′ ) ; pst18025L & Ds5-2a ( 5′-TCCGTTCCGTTTTCGTTTTTTAC-3′ ) . The primers used to genotype tam-2 , osd1-1 , osd1-2 and osd1-3 were previously described [15] , [16] , [28] . The uvi4 mutant ( pym ) [27] was genotyped by CAPS using the primers: pymU ( 5′-GGAGTGCTCTTCATTTTCTG-3′ ) , pymL ( 5′-TCTCATTTTGGATTTGTCTG-3′ ) and the restriction enzyme BsuRI ( 419 pb+158 pb for the mutant versus 286 pb+133 pb+158 pb for the wild type allele ) . HHpred searches were performed on user defined query alignment , without automatic PSI-BLAST enrichment of the query set and by using otherwise default settings [29] , [30] . The alignment of OSD1 proteins was performed with T-Coffee using default settings [29] , [37] . Tandem affinity purification constructs were generated and purified as described previously [38] . UVI4 , OSD1 , CDC20 . 1 , CDC20 . 3 , CCS52A1 , CCS52A2 and CCS52B tandem purified baits were separated by SDS-PAGE and probed with an anti-OSD1 antibody after western blotting . Yeast 2-hybrid interaction testing using OSD1 as bait ( pDEST32 ) with different APC/C subunits as prey ( pDEST22 ) was performed by mating , as described previously [28] . For mutant allele interaction screening , OSD1 mutant alleles were tested as bait ( pDEST32 ) against CCS52A1 as prey ( pDEST22 ) and introduced in the yeast PJ69-4a strain by cotransformation . GV stage oocytes were harvested from 9–16 weeks old CD-1 ( Swiss ) mice ( Janvier ) , injected and analyzed by live imaging essentially as described in [39] . H2B-RFP ( gift from Z . Polanski , Cracow , Poland ) mRNA was synthesized with the T3 mMessage Machine kit ( Ambion ) according to the manufacturer's instructions . For live imaging , a motorized inverted Nikon TE2000E microscope ( Plan APO 20x/0 , 75NA objective ) with PrecisExite High Power LED Fluorescence ( LAM 1: 400/465 , LAM2: 585 ) , equipped with a temperature chamber ( Life Imaging Services ) , Märzhäuser Scanning Stage , CoolSNAP HQ2 camera , and controlled by Metamorph software was used . Timepoints were taken every 20 minutes . Images were treated with ImageJ software . Immunofluorescence studies on formaldehyde fixed prometaphase I oocytes with anti-Osd1 antibody ( 1∶150 ) , and chromosome spreads of metaphase II oocytes were performed as described in [39] . OSD1 cDNA was amplified by sequential PCR first using attB1Ad-OSD1_s ( 5′-AAAAAGCAGGCTTCATGCCAGAAGCAAGAGATCG-3′ ) , attB2Ad-OSD1_as ( 5′-AGAAAGCTGGGTCTCATCGCATAGTCATTAAAGTCCG-3′ ) followed by attB1 adapter primer ( 5′-GGGGACAAGTTTGTACAAAAAAGCAGGCT-3′ ) , attB2 adapter primer ( 5′-GGGGACCACTTTGTACAAGAAAGCTGGGT-3′ ) . The PCR product was cloned , by Gateway ( Invitrogen ) , into the pDONR223 vector ( Invitrogen ) . A recombination reaction was performed between the resulting entry clone and a destination vector pHMGWA [40] . E . coli SoluBL21 cells ( AMS Biotechnology ) were transformed with the resulting destination clone , and grown in LB medium containing 100 µg/ml ampicillin until OD600 = 0 . 6 at 37°C . The culture was transferred to 18°C and grown for 30 min . The production of the fusion protein was induced by adding 0 . 3 mM IPTG ( isopropyl-β-d-thiogalactopyranoside , Thermo scientific ) overnight at 18°C . Cells were harvested by centrifugation and re-suspended in Ni-NTA binding buffer ( 50 mM NaH2PO4 , 100 mM NaCl , 10% ( v/v ) glycerol , 25 mM imidazole , pH 8 . 0 ) , and lysed by sonication . After addition of Triton X-100 to 0 . 2% ( w/v ) , the cell slurry was incubated at 4°C then clarified by centrifugation . The supernatant was passed through a column packed with Ni-NTA resin ( Qiagen ) , which was washed sequentially with Ni-NTA binding buffer followed by kinase buffer ( 50 mM Tris-HCl , pH 7 . 5 , 10 mM MgCl2 , 1 mM EGTA ) containing 150 mM NaCl , and eluted with kinase buffer containing 150 mM NaCl and 200 mM imidazole . CYCA1;2/TAM cDNA was cloned into pHMGWA as described above by using primers attB1Ad-TAM_s ( 5′-AAAAAGCAGGCTTCATGTCTTCTTCGTCGAGAAATCTATC-3′ ) and attB2Ad-TAM_as ( 5′-AGAAAGCTGGGTCTCAGAGGAAAAGCTCTTGCG-3′ ) followed by attB1 adapter primer and attB2 adapter primer . CYCA1;2/TAM-CDK complexes were prepared from E . coli as described [14] . Kinase reactions were performed as described in [14] using kinase buffer . In the case of OSD1 , kinase buffer containing 150 mM NaCl was used . Observation of final male meiotic products and chromosomes spreads were carried out as previously described [31] , [41] and observed with a ZEISS AxioObserver microscope . Observation of developing ovule by DIC and confocal microscopy was performed as described by Motamayor et al [42] . Alexander staining was performed according to [43] . Inflorescence were fixed in ethanol∶acetic acid ( 3∶1 ) and digested for 1 h as described in [41] . Meiocytes were squashed and immobilized on polysin slides as described in [44] , digested again for 30 min at 37°C in the digestion medium described in [41] and subsequently incubated one hour in PBS 1% Triton at room temperature . After 2 rinses with PBS 0 . 1% Triton , slides were incubated overnight at 4°C in primary antibodies ( mouse anti-tubulin ( Sigma T5168 ) diluted at 1/300 in PBS , 1% BSA , then washed in PBS , 0 . 1% Triton 5 times for 10 min . After 2 h of incubation at 37°C with the secondary antibodies in PBS 1% BSA , slides were washed in PBS 0 . 1% Triton 5 times for 10 min and mounted in Vectashield antifade medium ( Vector Laboratories ) with 80 µg/ml propidium iodide . Images were acquired with Zeiss Apotome . An anti-OSD1 antibody was raised against a full-length recombinant protein as described in [35] . OSD1 and CYCA1;2/TAM genomic fragment were amplified by PCR using OSD1 U ( 5′-CATATAAGCCTTGACCCTCTTTC-3′ ) , OSD1 L ( 5′-AGAAACCACCGAACTTGTGAAGA-3′ ) and TAM U ( 5′-CCAGTCACCACAATACACAC-3′ ) , TAM L ( 5′-GCGGTTTGGGTTGGTTTTTGTTT-3′ ) . The amplification for OSD1 covered 1603 nucleotides before the ATG and 170 nucleotides after the stop codon . The amplification for CYCA1;2/TAM covered 1495 nucleotides before the ATG and 493 nucleotides after stop codon . The PCR product was cloned , by Gateway ( Invitrogen ) , into the pENTR vector ( Invitrogen ) , to create pENTR-OSD1 and pENTR-TAM , respectively , on which directed mutagenesis was performed using the Stratagene Quickchange Site-Directed Mutagenesis Kit . The mutagenic primers used to generate the OSD1ΔD , OSD1ΔGXEN and OSD1ΔMR mutations were ( 5′-GCCTTCTTGGTATCCAGGAACACCTGTACGCGACATAAC-3′ ) , ( 5′- GATTGCCACAGGCAAGAGCGGCTATGCCCATAG-3′ ) and ( 5′- GGTGCGGACTTTAATGACTTAGCGATGATCTTTACTTAGG-3′ ) respectively . The mutagenic primers used to generate the TAMΔD were ( 5′- GTTGGAAACCGTGGTGCTCCCGTCGGCGACATCACAAATC-3′ ) . To generate binary vectors for plant transformation , an LR reaction was performed with the binary vector for the Gateway system , pGWB1 [45] . The resulting binary vectors , pOSD1 , pOSD1ΔD , pOSD1ΔGXEN , pOSD1ΔMR , pOSD1ΔDΔGXEN , pOSD1ΔDΔMR , pOSD1ΔDΔGXENΔMR and pTAM , pTAMΔD , were transformed using the Agrobacterium-mediated floral dip method [46] , on plant populations segregating for the osd1-3 , tam-2 or tdm-3 mutation . Transformed plants were selected on agar plates containing 50 mg/L kanamycin for OSD1 constructs and 20 mg/L hygromycin for CYCA1;2/TAM constructs , respectively . | In the life cycle of sexual organisms , a specialized cell division—meiosis—reduces the number of chromosomes from two sets ( 2n , diploid ) to one set ( n , haploid ) , while fertilization restores the original chromosome number . Meiosis reduces ploidy because it consists of two cellular divisions following a single DNA replication . In this study , we analyze the function of a group of genes that collectively controls the entry into the first meiotic division , the entry into the second meiotic division , and the exit from meiosis in the model plant Arabidopsis thaliana . We revealed a complex regulation network that controls these three key transitions . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"plant",
"science",
"plant",
"biology",
"genetics",
"biology",
"molecular",
"cell",
"biology",
"genetics",
"and",
"genomics"
] | 2012 | OSD1 Promotes Meiotic Progression via APC/C Inhibition and Forms a Regulatory Network with TDM and CYCA1;2/TAM |
Protein-RNA complexes formed by specific recognition between RNA and RNA-binding proteins play an important role in biological processes . More than a thousand of such proteins in human are curated and many novel RNA-binding proteins are to be discovered . Due to limitations of experimental approaches , computational techniques are needed for characterization of protein-RNA interactions . Although much progress has been made , adequate methodologies reliably providing atomic resolution structural details are still lacking . Although protein-RNA free docking approaches proved to be useful , in general , the template-based approaches provide higher quality of predictions . Templates are key to building a high quality model . Sequence/structure relationships were studied based on a representative set of binary protein-RNA complexes from PDB . Several approaches were tested for pairwise target/template alignment . The analysis revealed a transition point between random and correct binding modes . The results showed that structural alignment is better than sequence alignment in identifying good templates , suitable for generating protein-RNA complexes close to the native structure , and outperforms free docking , successfully predicting complexes where the free docking fails , including cases of significant conformational change upon binding . A template-based protein-RNA interaction modeling protocol PRIME was developed and benchmarked on a representative set of complexes .
About three quarters of the human genome could be transcribed into RNA , including 4 , 693 miRNAs [1] and 105 , 255 long noncoding RNAs [2] . The function of most of these RNAs is unknown . RNAs never act alone . One hypothesis is that the long noncoding RNA are molecular scaffolds for protein binding [3 , 4] . Several hundreds of novel RNA-binding proteins ( RBP ) were discovered by high-throughput sequencing [5 , 6] . Protein-RNA complexes play an important role in gene regulation , mRNA degradation and many other biological processes . High-throughput experimental techniques ( HITS-CLIP [7] , PAR-clip [8] , RIP-chip [9] ) and computational methods [10–18] have been developed to characterize protein-RNA interactome . These methods identify and characterize protein-RNA interactions , but do not provide the structure of protein-RNA complexes , which is important for understanding the molecular function . An increasing number of experimentally determined protein-RNA structures in PDB are still a fraction of all identified protein-RNA interactions , due to the inherent limitations of the experimental techniques . Thus this gap has to be filled by computational approaches [19] . The principles of protein-RNA interaction are based on structural and physicochemical complementarity [20–22] , and are similar to those of protein-protein interactions [23] . Thus the fundamental paradigms of structure prediction should be similar as well: free docking , for protein-protein [23] and protein-RNA complexes [24–29] , and the template-based docking , for protein-protein [30] and protein-RNA complexes ( investigated in this report ) . The accuracy of the template-based models is determined by the quality of the selected template , identified by sequence or structure alignment . Whereas the template-based paradigm in protein-protein modeling has been extensively studied and systematically validated/benchmarked [31] , similar investigation of template-based approach to protein-RNA complex structure prediction is still lacking ( although the approach has been applied to predicting RNA binding sites on proteins in SPOT-Struct-RNA[18] ) . We performed such investigation on a representative set of protein-RNA complexes . The analysis of all-to-all alignments in the set revealed a transition point between random and correct binding modes . The results showed that structural alignment significantly outperforms sequence alignment in identifying good templates , suitable for generating protein-RNA complexes with the ligand RMSD from the native structure < 10 Å . A template-based protein-RNA modeling protocol was developed and benchmarked on a representative set of complexes . The study provides a way for protein-RNA structure modeling on a genome scale .
Co-crystallized protein-RNA structures were downloaded from PDB ( 1 , 619 complexes in 2014-05-13 release ) . Structures with resolution better than 3 . 0 Å were retained . Multimeric complexes were split into binary ones , defined as one protein chain and one RNA chain . The minimal lengths of the protein and the RNA were 30 and 20 residues , respectively . The interface was defined by < 5 Å distance between any heavy atom of the protein and any heavy atom of the RNA . The minimal numbers of protein and RNA residues at the interface were 5 each . This resulted in 2 , 951 binary complexes , including 563 RNA chains and 2 , 721 protein chains . The RNA redundancy was removed by BLASTClust [32] with sequence identity cutoff 0 . 99 and coverage cutoff 0 . 99 . The 563 RNA chains were grouped into 288 clusters . The structure with the highest resolution in a cluster was designated as representative . This resulted in 633 binary complexes , which still included some short identical RNAs due to limitation in the default word size for nucleotides in BLASTClust . Thus CD-hit package [33] was used to further filter the RNA chains with sequence identity cutoff 0 . 99 . Finally , 439 non-redundant binary complexes ( NRBC439 ) were kept for all-to-all alignment and benchmarking . To determine the predictive power of our program , we split the NRBC439 set into two parts: 80% with an older deposit date were designated as the templates ( NRBC349 ) , and 20% with a newer deposit date were designated as targets ( bound set , NRBC90 ) . The performance of the template-based and free docking was also tested on the protein-RNA docking benchmark set [34] . To avoid modeling of targets on themselves , 26 complexes that were also part of the template set were excluded . Since in our implementation the template-based protocol can deal only with single-chain proteins and RNAs , the benchmark set was restricted to complexes with single-chain monomers . The length of the RNA chain was ≥ 10 nt according to the alignment procedure ( SARA [35] ) . Although the minimal 20 nt length was used previously [35] , in our study successful models were generated with the ≥ 10 nt threshold . The resulting set contained 49 complexes ( unbound set ) . In NRBC439 set all-to-all pairwise alignment was performed by three approaches . The first approach was local sequence alignment by fasta35 with default parameters [36] . Sequence identity of a complex was defined as the smaller sequence identity of the two monomers . The coverage of the complexes alignment was defined as the lowest coverage of the four chains in the two aligned complexes . The second approach was global sequence alignment by needle in the EMBOSS package [37] , also with the default parameters . The complex sequence identity and the coverage were defined as in the first approach . The third approach was structural alignment . For the structure alignment of RNA we chose SARA [35] , based on the reported performance characteristics [38] and availability . A newer version , SARA-coffee , is a structure-based multiple RNA aligner , which integrates SARA with R-coffee framework . For pairwise alignment , used in our study , the results of SARA-coffee and SARA are the same . For the structure alignment of proteins , we used TM-align [39] , following our previous studies of protein-protein complexes [31 , 40–44] . The output of TM-align is TM-score , which varies from 0 for completely dissimilar structures , to 1 for identical structures . The output of SARA is a score , which depends on the RNA size . To establish a similar description of structural similarity of proteins and RNAs , the SARA score was normalized by the score value of the RNA aligned to itself , resulting in the score interval 0–1 , similar to the protein alignment . As with the complex sequence identity , the complex structural score was defined as the minimum of TM-score and the normalized SARA score . The aligned atoms ( Cα in protein and C3' in RNA ) were used to calculate interaction RMSD ( IRMSD ) similarly to the one proposed for protein-protein complexes [45] , which numerically characterizes binding mode similarity of complexes of different monomers . It was shown previously to correlate well with the traditional ligand and interface RMSDs for complexes of same monomers in different binding modes ( cannot be applied to the complexes of different monomers ) [31] . The three alignment approaches were applied to NRBC439 to test the ability to detect a good template . Binary complexes in NRBC90 were queries for the template set NRBC349 . After a template was selected , the target protein was superimposed on the template protein by TM-align and the transformation matrix was saved . The target RNA was superimposed on the template RNA by SARA . Since SARA does not output the transformation matrix , it was reproduced by superimposing the RNA from SARA's output onto the original query RNA . The ligand RMSD ( RMSD of RNA C3' atoms ) between the model and the native structure was calculated . The quality of the model was measured by the ligand RMSD . In protein-RNA docking , a prediction was defined as "acceptable" [28] ( elsewhere called "native-like" [26 , 29] ) for the ligand RMSD ≤ 10 Å from the native structure of the complex , and a more accurate "medium" for the ligand RMSD ≤ 5 Å . These definitions correlate with the ones in protein-protein docking field [46] , and the corresponding docking models are generally considered within the intermolecular energy funnel [47] and thus subject to refinement by local optimization .
A previous study on template-based protein-protein docking determined strong dependence of the binding mode similarity on the structural similarity of the participating proteins , with the phase transition from dissimilar modes to the similar ones at TMm = 0 . 4 [31] . In the current study we asked a question: do protein-RNA complexes behave in a similar way ? We performed all-to-all pairwise comparison of protein-RNA binary complexes in NRBC439 set . The similarity of the monomers was measured by the sequence alignment ( fasta35 and needle for local and global alignment , correspondingly ) and by the structure alignments . Fig 1 shows the results of such comparison for local and global sequence alignments . For the local alignment , the 0 . 3 coverage threshold is used . The 0 . 3 value was the optimal , minimizing the noise from the lower threshold alignments ( results with no threshold for the coverage were largely random ) , while retaining 420 of 438 binary complexes for the analysis . The dip in cumulative fractions near 0 . 8 threshold value may be random , due to low sampling at this data range . One can also speculate that some of RBPs may have close homologs , with sequence ID near this value , whereas recent analysis showed that most RBPs are more diverse [48] . As the figure shows , the transition to similar binding modes occurs near the complex sequence identity 0 . 3 . The results of such comparison obtained by the structure alignment approach are shown in Fig 2a . The transition point on the alignment distributions was used as a cutoff for selecting good templates . S1 Fig shows that the success rate of detecting templates begins to decrease near the transition point ( complex structural score 0 . 45 ) . To distinguish the role of the protein in detecting a good template for a protein-RNA complex , the target/template similarity was also measured only for the protein component ( Fig 2b ) . This distribution is similar to the one in Fig 2a , indicating an important role of the protein . However , the role of the RNA is evident at the higher end of the structural similarity ( > 0 . 7 ) , where it eliminates multiple alternative binding modes . Thus the similarities of both protein and RNA are needed for an accurate identification of a good template for the complex . Overall , correlation of the protein-RNA structural similarity with the binding mode is weaker than that of the protein-protein complexes [31] because of the greater RNA flexibility [49 , 50] . Structural similarity vs . sequence identity of the protein-RNA complexes is plotted in Fig 3 . The plot is divided into four areas by the lines x = 0 . 45 ( transition point for structural similarity ) , and y = 0 . 25 ( transition point for sequence similarity ) . The correlation of structure and sequence similarity in protein-RNA is similar to that in protein-protein complexes [31] . The structure and sequence are dissimilar in the lower left quadrant , which contains 98 . 4% of the alignments . This points to the diversity of sequences and structures in NRBC439 set ( supported by observation that 1 , 542 RBPs formed 1 , 111 families in human RBPome [48] ) . The upper right quadrant contains 1 . 02% of the alignments , and 69 . 53% of those with the structural score ≥ 0 . 45 , where structure and sequence are similar , suggesting that both approaches can find a good template . The alignments with similar structure and dissimilar sequence are in the lower right quadrant , containing 0 . 45% of alignments , and 30 . 47% of those with the structural score ≥ 0 . 45 . This suggests that structural alignment approach could find good templates for about 1/3 of cases when sequence alignment cannot . Last , the top left quadrant shows similarity detected by the sequence , but not the structural alignment . It is almost empty , which means that the structural alignment finds most templates detectable from the sequence . A structure alignment-based docking was implemented in a procedure PRIME ( Protein-RNA Interaction ModEling ) . Fig 4 shows the outline of the approach . Docking was systematically benchmarked on NRBC90 targets using NRBC349 templates . For each target docking models were generated by PRIME , ranked separately by the complex structural score and by the TM-score . The success rates of different approaches are shown in Fig 5 . The success rate for predicting "acceptable" model almost reaches the highest value after top 4 . This suggests that for the docking , the complex structural score , which accounts for both TM-score for proteins and SARA score for RNA , is better than just the TM-score for proteins in top 1 , top 2 , and top 3 . The TM-score outperformed or tied with the complex structural score when considering more top models . The TM-score detected the template for three complexes , for which the complex structural score could not . The reason was that when the normalized SARA score was counted in , the complex structural score decreased below the cutoff ( score values 0 . 37 , 0 . 016 , and 0 . 15 ) . The improvement of the success rates for top 1 , top 2 , and top 3 predictions with the complex structural score was largely due to the reduction of noise after the transition point in Fig 2 ( by moving it to the left of the transition point ) . For example , the alignment of the target complex 3umy , chains A and B , and the template complex 2hw8 , chains A and B , had IRMSD = 28 . 04 Å , but the TM-score 0 . 90 , ranked 2 by the TM-score alone . At the same time , the corresponding complex structural score is 0 . 29 , ranked 51 , moving the complex to the left of the transition point , and thus reducing the noise for the high scored complexes . Fig 6 shows the distribution of the best models according to ligand RMSD . The distribution is bimodal , pointing to the existence of alternative binding modes , similar to protein-protein complexes [31] . The high-quality predictions ( 0–2 Å ) correspond to 30 targets ( 33% ) . Benchmarking of PRIME suggests that 65% of target models can be built successfully ( structural score-10 . 0 for top 10 predictions in Fig 5 ) . Ranked by the complex structural score , most models with "acceptable" quality are ranked at top 4 . Similar to protein-protein modeling , the template-based protein-RNA docking has a clear advantage over the free docking method , where scoring functions typically are struggling to pick the correct model from the multitude of docking poses [29] . The template-based method of course cannot be applied when a template is not found , in which case the free docking should be used . In our benchmark , templates were detected for 69 out of 90 targets . Fig 7 shows an example of the target with low protein sequence identity to the template , successfully modeled by the structure alignment . Still , structure similarity does not guarantee correct predictions . The alternative binding modes were observed in nine targets with high structural similarity to the templates . Although the complex structural scores of their alignment to the templates were larger than the transition point , the ligand RMSD of the models built on these templates were > 10 Å . For example , the TM-score , normalized SARA score and the complex structural score between the target 4lgt , chains A and E , and the template 2i82 , chains A and E , were 0 . 543 , 0 . 524 and 0 . 524 , respectively . However , the binding mode is different , with the model/native ligand RMSD 22 . 45 Å . To compare the performance of template-base and free docking method , we tested template-based PRIME and free docking RPDock on the unbound set ( see Methods ) . RPDock [29] is a protein-RNA rigid docking protocol , which takes into account protein/RNA geometric and electrostatic complementarity , and stacking interaction in the base of nucleotides with the aromatic rings of charged amino acids . All PRIME models were ranked by the complex structural score , and RPDock models were ranked by DECR-RP [29] . Fig 8 shows the docking results . Success rate is defined the number of those with at least one "acceptable" model divided by the total number of targets . The results show that the success rate of the template-based protein-RNA docking is significantly higher than that of the free docking , similarly to the previous results in protein-protein docking [41] ( although a broader assessment of the protein-protein category is still on-going [51 , 52] ) . The detailed data on benchmarking ( S1 Table ) indicates that the template-based approach significantly outperforms free docking , successfully predicting complexes where the free docking fails , including cases of larger bound/unbound RMSD ( see S2 Table , and an example of a successful template-based prediction of a complex with a significant conformational change on the protein component in S3 Fig ) . PRIME also runs ~ 5 times faster than RPDock ( S2 Fig ) , which is especially important for genome-scale studies . PRIME currently does not include a refinement protocol , which is still a challenging task in macromolecular docking [46] . The development of a dedicated refinement protocol is in our future plans . However , even a standard minimization by GROMACS ( v5 . 0 . 7 ) [53]with AMBER99 force field reduced the number of clashes in most complexes ( S4 Fig ) . Sequence and structure alignment approaches were compared in template-based modeling of protein-RNA complexes . All-to-all alignment of protein-RNA complexes detected a phase transition from random to similar binding modes , according to the degree of monomers similarity . The structure alignment showed to be significantly better than the sequence alignment in identifying correct templates . In systematic benchmarking , structure alignment-based docking had far better success rate than the free docking , successfully predicting complexes where the free docking failed , including interactions with significant conformational change upon binding . The findings are qualitatively similar to those observed earlier in structural modeling of protein-protein complexes [31] . Applicability of the prediction protocols to complexes of modeled monomers , rather than to experimentally determined structures of monomers , which typically have higher accuracy than models , was previously established for protein-protein interactions in systematic benchmarking studies on specifically designed sets of protein models[54 , 55] . Similar studies are needed to determine such applicability to modeled RNAs [56] . The structure alignment-based approach for protein-RNA modeling is implemented in PRIME software , publicly available at http://rnabinding . com/PRIME . html . | Structures of protein-RNA complexes are important for characterization of biological processes . The number of experimentally determined protein-RNA complexes is limited . Thus modeling of these complexes is important . Reliable structural predictions of proteins and their complexes are provided by comparative modeling , which takes advantage of similar complexes with experimentally determined structures . Thus , in the case of protein-RNA complexes , it is important to determine if similar proteins and RNAs bind in a similar way . We show that , similarly to the earlier published results on protein-protein complexes , such correlation of the protein-RNA binding mode and the monomers similarity indeed exists , and is stronger when the similarity is determined by structure rather than sequence alignment . The data shows clear transition from random to similar binding mode with the increase of the structural similarity of the monomers . On the basis of the results we designed and implemented a predictive tool , which should be useful for the biological community interested in modeling of protein-RNA interactions . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"and",
"Discussion"
] | [
"sequencing",
"techniques",
"protein",
"interactions",
"protein",
"structure",
"prediction",
"protein",
"structure",
"molecular",
"biology",
"techniques",
"rna",
"alignment",
"research",
"and",
"analysis",
"methods",
"sequence",
"analysis",
"rna",
"structure",
"protein",
"structure",
"determination",
"sequence",
"alignment",
"proteins",
"molecular",
"biology",
"protein",
"structure",
"comparison",
"biochemistry",
"rna",
"nucleic",
"acids",
"biology",
"and",
"life",
"sciences",
"macromolecular",
"structure",
"analysis"
] | 2016 | Template-Based Modeling of Protein-RNA Interactions |
Genomic time series data generated by evolve-and-resequence ( E&R ) experiments offer a powerful window into the mechanisms that drive evolution . However , standard population genetic inference procedures do not account for sampling serially over time , and new methods are needed to make full use of modern experimental evolution data . To address this problem , we develop a Gaussian process approximation to the multi-locus Wright-Fisher process with selection over a time course of tens of generations . The mean and covariance structure of the Gaussian process are obtained by computing the corresponding moments in discrete-time Wright-Fisher models conditioned on the presence of a linked selected site . This enables our method to account for the effects of linkage and selection , both along the genome and across sampled time points , in an approximate but principled manner . We first use simulated data to demonstrate the power of our method to correctly detect , locate and estimate the fitness of a selected allele from among several linked sites . We study how this power changes for different values of selection strength , initial haplotypic diversity , population size , sampling frequency , experimental duration , number of replicates , and sequencing coverage depth . In addition to providing quantitative estimates of selection parameters from experimental evolution data , our model can be used by practitioners to design E&R experiments with requisite power . We also explore how our likelihood-based approach can be used to infer other model parameters , including effective population size and recombination rate . Then , we apply our method to analyze genome-wide data from a real E&R experiment designed to study the adaptation of D . melanogaster to a new laboratory environment with alternating cold and hot temperatures .
There is a small but growing literature on the analysis of evolve-and-resequence data . Feder et al . [30] present a statistical test for detecting selection at a single biallelic locus in time series data . ( Although it is not a major focus , their method can also be used to estimate the selection parameter . ) Similar to our method , they model the sample paths of the Wright-Fisher process as Gaussian perturbations around a deterministic trajectory in order to obtain a computable test statistic . However , their aim is slightly different from ours in that they analyze yeast and bacteria data sets where the population size is both large and must be estimated from data . Here we focus on population sizes which are smaller and more typical of experiments performed on higher organisms , for example mice or Drosophila . We generally assume that the effective population size is known but also test our ability to estimate it from data . Also , because of the increased amount of drift present in the small population regime , we necessarily restrict our attention to selection coefficients which are somewhat larger than those considered by Feder et al . Finally , although Feder et al . do study the performance of their method when time series data are corrupted by noise due to finite sampling ( as in e . g . a next-generation sequencing experiment ) , they do not model this effect . Here we properly account for the effect of sampling by integrating over the latent space of population-level frequencies when computing the likelihood . Another related work is Baldwin-Brown et al . [31] , which presents a thorough study of the effects of sequencing effort , replicate count , strength of selection , and other parameters on the power to detect and localize a single selected locus segregating in a 1 Mb region . Results are obtained by simulating data under different experimental conditions and comparing the resulting distributions of allele trajectories under selection and neutrality using a modified form of t-test . Because it is not model-based , this method is incapable of performing parameter estimation . As a result of their study , Baldwin-Brown et al . present a number of design recommendations to experimenters seeking to attain a given level of power to detect selection . In a related work , Kofler and Schlötterer [32] carried out forward simulations of whole genomes to provide guidelines for designing E&R experiments to maximize the power to detect selected variants . Illingworth et al . [33] derive a probabilistic model for time series data generated from large , asexually reproducing populations . The population size is sufficiently large ( on the order of ∼ 108 ) that population allele frequencies evolve quasi-deterministically . The deterministic trajectories are governed by a system of differential equations describing the effect of a selected ( “driver” ) mutation on nearby linked neutral ( “passenger” ) mutations . Randomness arises due to the finite sampling of alleles by sequencing . The main difference between the setting of Illingworth et al . ’s and our own concerns genetic drift . While drift may be ignored when studying a large population of microorganisms , we show that it confounds our ability to detect and estimate selection in populations of order ∼ 103 . Thus , for E&R studies on ( smaller ) populations of macroscopic organisms , methods which assume that allele frequencies evolve deterministically may not perform as well as those which explicitly take drift into account . Topa et al . [34] present a Bayesian model for single-locus time series data obtained by next-generation sequencing . In each time period , the allele count is modeled as a draw from a binomial distribution with number of trials equal to the depth of sequencer coverage , and success probability equaling the population-level allele frequency . The posterior allele frequency distribution is used to test for selection by comparing a neutral model to one in which unobserved allele frequencies to depend on time . In the non-neutral case , a Gaussian process is used to allow for directional selection acting on the posterior allele frequency distributions . Finally , Lynch et al . [35] derive a likelihood-based method for estimating population allele frequency at a single locus in pooled sequencing data . The method allows for the possibility of sequencing errors as well as subsampling the population prior to sequencing . Using theoretical results as well as simulations , the authors give guidelines on the ( subsampled ) population size and coverage depth needed to reliably detect a difference in allele frequency between two populations . Unlike the other methods surveyed here , the approach of Lynch et al . is not designed to analyze time series data . Hence the data requirements needed to reliably detect allele frequency changes using their method—for example , sequencing coverage depth of at least 100 reads—are potentially greater than for methods are informed by a population-genetic model of genome evolution over time . Our method differs from the above-mentioned approaches in several regards . To the best of our knowledge , ours is the first method capable of analyzing time series data from multiple linked sites jointly . We find that this is advantageous when studying selection in E&R data . Furthermore , it enables us to analyze features of these data which cannot be studied using single-locus models , such as local levels of linkage disequilibrium and the effect of a recombination hotspot . Additionally , because our model is based on a principled approximation to the Wright-Fisher process , it can numerically estimate the selection coefficient , dominance parameter , recombination rates , and other population genetic quantities of interest . In this way it is distinct from the aforementioned simulation-based methods [31 , 32] , methods which only focus on testing for selection [30 , 31 , 34] , or methods based on general statistical procedures which are not specific to population genetics [34 , 35] . Source code implementing the method described in this paper is included in S1 Code . The experimental data analyzed in Analysis of a real E&R experiment data are from Franssen et al . [36] and are available on the Dryad digital repository http://dx . doi . org/10 . 5061/dryad . 403b2 .
We consider the following model of an E&R experiment . A sexually reproducing population of N diploid individuals is evolved in discrete , non-overlapping generations . Pooled DNA sequencing [37 , 38] is performed T times at generations t1 < t2 < ⋯ < tT . At each segregating site in the resulting data set , we assume that there are two alleles , denoted A0 and A1 . ( As will be seen below , up to a change in the sign of the selection coefficient associated with each site , the model is agnostic to which allele is called A0 or A1 . ) Let L and R denote the number of loci and the number of experimental replicates , respectively . The array D ∊ [0 , 1]T×L×R counts relative frequency with which the A1 allele was observed for each combination of generation , locus and replicate . Given D and a vector of underlying population-genetic parameters θ , let ℙ ( D∣θ ) denote the model likelihood . In an idealized E&R experiment , generations are discrete and non-overlapping , mating is random , and the population size is fixed , so the likelihood is well approximated by the classical Wright-Fisher model of genome evolution [39]: ℙ ( D ∣ θ , G 0 ) = ∑ G 1 ∊ 𝒢 ⋯ ∑ G T ∊ 𝒢 ℙ ( D ∣ G 0 , ⋯ , G T ) ℙ θ ( G T ∣ G T - 1 ) ⋯ ℙ θ ( G 1 ∣ G 0 ) , ( 1 ) where ℙθ ( Gi∣Gi−1 ) is the transition function of the discrete , many-locus Wright-Fisher Markov chain from genomic configuration Gi−1 to Gi given parameters θ , 𝒢 is the set of all possible genotypic configurations in a diploid population of size N , and ℙ ( D∣G0 , … , GT ) is the probability of the sequencer emitting D conditional on G0 , … , GT . ( Here , G0 represents the haplotypic configuration of the founding experimental population . In order to take advantage of linkage information we assume that this is known , although as described in Methods this is not necessary in order to use a single-locus version of our model . ) For typical problems , evaluating ( 1 ) is intractable since ∣𝒢∣ is very large and the transition density ℙθ ( Gi∣Gi−1 ) is difficult to compute and store . Asymptotic ( i . e . , diffusion ) approximations to the transition density may be inaccurate if the population size N and/or scaled generation time 2Nt are small , as is common in an E&R experiment . Hence , alternative approximations to ℙ ( D∣θ ) are needed to perform inference . The approximation we make is as follows . Let X ≡ ( Xijk ) ∊ [0 , 1]T×L×R denote the ( unobserved ) population frequency of the A1 allele at each data point . Conditional on knowing X , and assuming that the DNA sequencer samples each site independently and binomially from the population , we have Dijk ∼ Binomial ( cijk , Xijk ) where cijk is the depth of sequencing coverage observed at this site . ( Although sequencer coverage is random , we assume that it is independent of all other variables in the experiment and treat it as constant conditional on the observed data . ) Marginalizing over X , we obtain 𝓛 ( D ∣ θ ) = ∫ [ 0 , 1 ] T × L × R ∏ i , j , k 𝓑 ( D i j k ; c i j k , x i j k ) 𝑝 X ( x ∣ θ ) d x , ( 2 ) where 𝓑 ( d;c , x ) = ( cd ) xd ( 1−x ) c−d is the probability mass function of the binomial distribution and 𝑝 X ( x ) is the density of X . Note that if each cijk is large , as when the samples have been deeply sequenced , then the likelihood is ( approximately ) proportional to the density of X , i . e . , 𝓛 ( D∣θ ) ∝ 𝑝 X ( x ) , and the integral in ( 2 ) does not need to be evaluated . This computational savings can be useful when performing simulations . To perform inference we must approximate the density 𝑝 X , which represents the joint distribution of all allele frequencies observed in the experiment . In Methods , we provide the details of the approximation we use . Briefly , it is as follows: we assume that , conditional on the initial genome configuration G0 , the underlying allele frequencies Xijk are distributed according to a Gaussian process: X ∣ G 0 , θ ∼ 𝒩 ( μ ( G 0 , θ ) , Σ ( G 0 , θ ) ) ( 3 ) where the first- and second-order moment functions μ ( ⋅ ) and Σ ( ⋅ ) are obtained by considering Wright-Fisher models on a small number of loci . For example , the terms of Σ ( ⋅ ) correspond to the covariance between a pair of linked sites ( potentially at different time points in the experiment ) under the Wright-Fisher model . To compute this we can “marginalize out” the remaining loci in the model and study simpler Wright-Fisher model on only two loci . ( A slightly more elaborate approximation is needed in the case when there is a nearby selected locus , as detailed in Methods . ) Thus , we are essentially approximating the complex joint distribution of allele frequencies using a sequence of simpler one- and two-locus distributions . This approximation enables us to capture the correct mean and covariance structure in the random variable X while omitting higher order correlations . Using this approximation we can perform tractable , likelihood-based inference while capturing salient aspects of the linkage-induced correlation present in the data . Indeed , by ( 2 ) , ( 3 ) and the preceding discussion we have 𝓛 ( D ∣ θ ) ≈ ∫ [ 0 , 1 ] T × L × R ∏ i , j , k 𝓑 ( D i j k ; c i j k , x i j k ) ϕ θ ( x ) d x = : 𝓛 ˜ ( D ∣ θ ) , ( 4 ) where ϕθ denotes the density function of the Gaussian distribution in equation ( 3 ) . This expression may then be maximized over θ to perform inference . Alternatively , by placing a prior on θ a Bayesian approach may be adopted , but we do not explore that in this work . We tested our method on simulated data designed to capture the essential features of an E&R experiment . See Methods for the details on simulation . Briefly , it consisted of cloning a set of F homozygous founder lines ( whose haplotypes are assumed to be known ) to form an experimental population of N diploid organisms , which were then simulated forwards in time for T generations according to the Wright-Fisher random mating model . The experiment was repeated using the same starting conditions to form R experimental replicates . After the simulation terminated , the frequency of allele A1 was recorded for each combination of segregating site , time period and replicate , possibly with introduced sampling error; this setup mimics pooled sequencing . The input to the model consisted of these time series allele frequency data along with the haplotypes of the founder lines . Certain aspects of the simulation were varied to test different aspects of the model; these changes are described more fully in their respective sections below . Unless otherwise noted , the simulations were performed using F = 200 founder lines , census population size N = 1000 , sampling at generations ti ∊ {10 , 20 , 30 , 40 , 50} , R = 3 experimental replicates and a region of size L = 105 sites . These values were chosen to reflect a typical E&R experiment and we refer to them in the sequel as the “default” parameter values . Expected sequencing coverage depth is denoted by C , with C = ∞ corresponding to having perfect knowledge of the population allele frequencies . We use C = ∞ in the default parameter setting to upper bound the performance of our method , but also consider C ∊ {10 , 30} to investigate the effect of uncertainty in allele frequency estimation . In these scenarios , coverage at each site was Poisson distributed with mean C . Lastly , scenarios with coverage “Ĉ” denote simulations in which each segregating site had a random level of coverage drawn from the empirical coverage depth distribution observed in actual E&R sequencing data ( see Analysis of a real E&R experiment data for further details . ) The average coverage depth observed in this experiment was 84 short-reads , but the distribution has a heavy left tail which leads to a small percentage of sites having little to no coverage ( S1 Fig ) . A common objective in E&R experiments is to detect genetic adaptation . For example , a population may be partitioned , with one subgroup placed in a new environment . Upon running an E&R experiment , one wishes to 1 ) determine whether a fitness difference exists between the control and subject groups; 2 ) find the alleles responsible for the adaptation; and 3 ) estimate the strength of selection acting on these alleles . To test our model’s ability to perform each of these tasks , we simulated E&R experiments in which a segregating site in the founding population was chosen uniformly at random and placed under selection . The relative fitnesses of A0/A0 and A1/A1 homozygote genotypes are respectively given by 1 and 1+s , while the relative fitness of the heterozygote A0/A1 is 1+hs . In what follows , we assume h = 1/2 unless stated otherwise . Let si denote the coefficient of selection at segregating site i = 1 , … , K , where K is the total number of segregating sites in the region being considered . We wish to test the following null and alternative hypotheses: H 0 : s 1 = ⋯ = s K = 0 , versus H A : s j ≠ 0 for some j , ( 5 ) which can be implemented using a standard likelihood-ratio ( LR ) test . As the number R of experimental replicates grows large , the distribution of the test statistic under the null hypothesis tends to a χ2 distribution . However , since R was set to a realistic ( i . e . , small ) value in our experiments , we found that the test performed better if the null distribution was determined empirically . The null distribution was calculated by performing additional simulations under neutrality ( s = 0 ) , computing the maximum likelihood estimate ŝ for each simulation , and then using these estimates to compute the empirical null distribution of the LR test statistic - 2 log 𝓛 ˜ ( D ∣ s = 0 ) - sup u log 𝓛 ˜ ( D ∣ s = u ) , ( 6 ) where 𝓛 ˜ ( D | s = u ) is defined in ( 4 ) . Using the default parameter settings mentioned earlier , Fig . 1 displays the test’s estimated receiver operating characteristic ( ROC ) curve for various strengths s of selection and various numbers of founding haplotypes ( F ) . Larger values of F correspond to increased haplotypic diversity in the start of the E&R experiment . Each curve was estimated from 200 simulations . Some overall trends are apparent: stronger selection is easier to detect than weaker selection , and increased haplotypic diversity makes it more difficult to confidently reject the null hypothesis of neutrality . With a small number of initial haplotypes ( F = 20 ) , strong selection ( s = 0 . 1 ) is easily distinguished from neutrality . Moderate selection ( s = 0 . 05 ) is more challenging to detect , but the test still has 75% power with a false positive rate of ∼ 6% . Weaker selection ( s = 0 . 02 ) poses more of a challenge; in this case achieving 50% power would entail a false positive rate of approximately 30% . As the number of founding lineages increases , it becomes harder to test for selection . This occurs because many sites are segregating at low initial frequencies , increasing the chance that some are lost due to drift . Detecting weakly selected variants may be confounded by genetic drift , which can cause low-frequency alleles to be lost even if they are under positive selection . One option for improving sensitivity to weaker selection is to reduce the effect of drift by increasing the effective population size used in the experiment . To study how this influences our ability to detect weaker selection , we ran additional simulations with larger population sizes N ∊ {2000 , 5000} while holding the remaining experimental parameters fixed . Results from these experiments are shown in Fig . 2 . The top panel ( N = 1000 ) is reproduced from the middle panel of the preceding figure for ease of comparison . We see that reducing the amount of genetic drift in the data improves the performance of the test , particularly when it comes to distinguishing weak selection ( s = 0 . 02 ) . Once selection has been detected in a region , it is desirable to map the selected site as accurately as possible . An obvious estimator in this case is to declare the site with the highest likelihood-ratio ( versus a neutral model ) from the preceding test to be the selected site . Table 1 shows how this estimation procedure performed for different strengths of selection . We also studied how varying the number of founding lines affected the ability to precisely locate the selected site by allowing F to take on the values F ∊ {20 , 200 , 2000} . Since the minimum minor allele frequency ( MAF ) in an E&R experiment is 1/F , a low number of founding lines ensures that sites are segregating at intermediate frequencies , while a large value of F decreases LD and improves the ability to map the selected site accurately . Note that under our default parameter regime , setting F = 2000 amounts to sampling each founder from a panmictic population of size , so that the patterns of diversity reflect what would be seen in a ( neutrally evolving ) region in nature . Two measures of the accuracy are displayed in Table 1 . The first set of columns examines the distribution of the distance ( in base pairs ) between the estimated and true selected site . The second set of columns examines the distribution of the rank of the true selected site when all segregating sites in the region are sorted according to their likelihood ratio . As the table shows , selection becomes easier to localize as it becomes stronger and as the number of founder haplotypes grows . With strong selection ( s = 0 . 1 ) and 20 founding haplotypes , the method correctly pinpointed the exact location of the selected site in over 50% of the simulations . Additionally , the correct selected site was among the top four in 75% of the simulations . With F = 200 founder lines , the true selected site ranked among the top two overall in over half the simulations . The top rows of Table 1 indicate that weak selection ( s = 0 . 02 ) is difficult to localize precisely using this method; the median estimated distance from the true selected site was 27–29 kb in these cases . Since increasing the number F of founder lines diminishes linkage disequilibrium , it may seem counterintuitive that our results suggest that localizing selection actually becomes more difficult as F increases . In S1 Table , we have displayed the same statistics as Table 1 for the restricted subset of simulations where the selected site was segregating at an initial frequency of at least 0 . 1 . Compared to the unrestricted data set , these sites are more likely to rise in frequency by the action of positive selection , and less likely to be lost due to drift . Here we see that increasing F does improve the ability to map the selected site for s ∊ {0 . 02 , 0 . 05}; for strong selection ( s = 0 . 1 ) , essentially all cases of F performed equally well . Interestingly , an intermediate number of founding lineages ( F = 200 ) seems to outperform both other regimes , suggesting that there is a trade-off between improving localizability by increasing F and limiting the number of segregating sites which must be considered by decreasing the number of founding lineages . We also studied how coverage depth affects the ability to map the selected site . For F = 200 , Table 2 repeats the analysis of Table 1 when the data are sampled at simulated coverage depths of 10 and 30 short-reads , as well as from the empirical coverage distribution discussed above . Comparing the two tables , we see that the additional noise introduced by sequencing makes the problem of localizing the selected site more difficult; the modal estimate is often separated from the true site by tens of kilobases . Nevertheless , in more than half the trials performed we observed that a strongly selected site would be among the top five segregating sites ( in terms of likelihood ratio; see Table 2 , last two rows ) . For medium selection , increasing coverage depth from 10 to 30 short-reads improved our ability to map the selected site by several kilobases , and more than halved the number of segregating sites we would need to examine before encountering the selected site . Weaker selection , already difficult to detect without sampling , is even more so when noise is introduced . Once a selected site has been located , it is desirable to numerically quantify the fitness of the A1 allele . Table 3 describes the distribution of these estimates for various combinations of selective strength , coverage depth , and model complexity ( i . e . , the number of loci in the Gaussian process approximation ) . For each of the simulations above we estimated s by maximum likelihood . To separate the ability of our model to estimate selection from its ability to locate the selected site , we assumed that the selected site was already known when performing these estimates . Aside from varying selection strength , we also examined how coverage depth and the number of loci used for estimation affected the quality of the estimates . For each parameter combination , the table displays the mean , median and inter-quartile range ( IQR ) of the distribution of the maximum likelihood estimate ŝ of s . Several interesting features emerge from the table . Inter-quartile range is of roughly the same order across scenarios , so that estimation error shrinks relatively as selection become stronger . For one-locus models , IQR shrinks as coverage depth increases . For multi-locus models the effect of increasing the number of sites used to perform estimation is interesting . When the data are observed without noise , we saw little improvement in the accuracy of ŝ when using a single-locus model fit only to data from the selected site versus a multi-locus model which also took the trajectories of linked sites into account . In fact , in several cases this cause the estimates to become more dispersed as the trajectory of the selected allele had relatively less weight in the likelihood calculation . On the other hand , when allele frequencies are sampled with noise we see that estimates ŝ obtained from a five-locus model generally have smaller IQR , particularly in the low-coverage-depth case C = 10 . These findings are confirmed in Fig . 3 , which displays density estimates for the residual s−ŝ for each of these cases presented in the table . Compared with the one-locus model , the five-locus model which takes additional data from linked sites into account produces estimates which are more concentrated around the true parameter value . Thus , when the data are noisy ( i . e . , when C is small ) , the trajectories of nearby linked sites provide useful information concerning the ( unobserved ) population frequency of the selected allele as it evolves over time . We observed a slight negative bias for weaker selection and a slight positive bias for medium and strong selection , which can be attributed to loss or fixation of the selected allele . Indeed , estimated selection may be negative when a weakly selected allele segregating at low frequency is lost due to drift; similarly , there is a tendency to overestimate the strength of selection acting on a high-frequency allele which fixes quickly . It is also interesting to consider the effect of study design on estimation accuracy . In Table 4 we examine how parameter estimates are affected by sequencing effort and experimental duration . We focus on the limited-coverage case ( C = 10 ) since it is most sensitive to adding or removing sequence data from additional generations . For ease of comparison , the first set of rows reproduces data from Table 4 , where generations {10 , 20 , 30 , 40 , 50} were sequenced . The next subsection examines the case when sequencing effort is reduced to two time periods {25 , 50} . The final subsection studies estimation quality when the experimental duration is halved , and only one round of sequencing is performed at generation 25 . In all cases we see that the estimators are approximately unbiased , 𝔼 ( ŝ ) ≈s , but that their dispersion about the true parameter value is greatly affected by data availability . Sampling genomic data at just a single time period t = 25 roughly doubles the IQR of the estimator in each case . Interestingly , with two time periods ( t ∊ {25 , 50} ) performance is improved , and the estimator is only somewhat less precise than when sampling at every tenth generation . Finally , as in the previous table we see again that , at least for data sampled at low coverage , estimation performance is unilaterally improved by fitting a multi-locus model versus a single-locus model . In the preceding discussion , the dominance parameter was fixed at h = 1/2 , so that selection acted additively . Our method is capable of handling general diploid selection . In our experiment , we tested our method’s ability to estimate the effect of overdominance , in which case heterozygotes are fitter than either homozygote . We simulated populations under the conditions h > 1 and s ≪ 1 such that heterozygotes had a relative fitness of 1+hs where hs ∊ {0 . 02 , 0 . 05 , 0 . 10} . Thus , heterozygotes have a fitness advantage of the same order as that which we were able to detect in the additive case . Results for jointly estimating h and s are shown in Table 5 . A fixed value of s = 0 . 01 was used for fitness in all cases , while h was varied . We found that estimating overdominance is difficult when both alleles are initially segregating near their limiting frequency of ½ , since the resulting allele trajectories appear very similar to those generated by a neutral model with drift . The results in the table are therefore conditioned on the initial allele frequency residing outside of the interval [0 . 4 , 0 . 6] . When considered individually , the estimators ĥ and ŝ are highly variable ( see Table 5 , columns 3–6 ) . This behavior is expected since , as witnessed in the previous subsections , small values in s ( specifically , s = 0 . 01 ) are difficult to detect in experimental data . Encouragingly , a different picture emerges when we consider the product estimator ĥ⋅ŝ ( see Table 5 , columns 7–8 ) . The estimator is close in expectation to the true value hs ( column 2 ) and also more tightly concentrated around that value . Density estimates of the product estimator ĥŝ are shown in Fig . 4 and confirm this finding . Each density estimate has a mode at the true parameter value hs and is reasonably concentrated around that value . Our multi-locus model can also be used to study phenomena which alter covariance between linked alleles . For example , in a region containing a recombination hotspot , covariance decreases markedly as increased recombination breaks down linkage disequilibrium . Using the same likelihood-based approach as above , the recombination rate within the hotspot can be estimated from E&R data . To test this , we simulated a region of length L = 100 kb in which the middle 2 kb region had an elevated recombination rate rH = α ⋅ r , where r = 10−8 is the background recombination rate and α ∊ {10 , 102 , 103} . For simplicity , we focused on the case of C = ∞ and assumed that the hotspot boundaries are known . For each simulation , a 30-locus model was fit using 10 randomly-selected loci from within the hotspot and 20 outside of it . Density estimates for the residual log10 ( r̂H ) −log10 ( rH ) are shown in Fig . 5 . In all cases , the mode of the density occurs close to zero . A 3-order increase in the recombination rate is easily detected in experimental data , and a 2-order increase can also be estimated to well within an order of magnitude of accuracy . Increasing the recombination rate by only a factor of 10 leads to a fairly dispersed estimator , and it would be difficult to detect using the default experimental parameters . As a final application of our method , we consider estimating the effective population size Ne from experimental data . Up to now we have assumed that the ( census ) size N of the experimental population is fixed at a known value . In practice , the effective and census population sizes may differ due to various factors , including nonrandom mating and population structure . It could be interesting to quantify this effect by estimating Ne in experimental data using the same likelihood-based procedures described above . Since our model approximates the Wright-Fisher process , in which Ne = N , and simulations were carried out also assuming the Wright-Fisher model , we expect our estimate N̂e to be close to N . Fig . 6 shows a scatter plot of N̂e versus N for 1 , 000 simulated E&R experiments . In each experiment , the population size N was chosen uniformly at random from the interval [10 , 104] . We see that the estimator is quite accurate for small population sizes and becomes more variable as N grows . This is expected since N̂e is essentially measuring genetic drift , which is of order O ( 1/N ) as N grows . Thus , the inverse map taking drift to population size is well-conditioned for small N and becomes ill-conditioned as N grows . Finally , we tested our method on data from an actual E&R experiment of D . melanogaster adapting to a new laboratory environment involving an alternating cycle with 12-hrs of cold ( 18∘C ) and 12-hrs of hot ( 28∘C ) temperature conditions . The experiment has been described previously [25 , 36] , so we give only a brief summary here . The experiment consists of three D . melanogaster populations each of N ≈ 1000 individuals . The populations were founded by gravid females from isofemale lines , and then evolved forward in discrete generations . Pooled sequencing was performed at generations 15 , 37 , and 59 on three experimental replicates . The observed coverage distribution for a selected data point ( replicate 4 , generation 59 ) is shown in S1 Fig . The distribution has fairly high average coverage depth , but a significant number of sites have little or no coverage . After read-mapping and filtering sites to have sufficient coverage and quality , 1 . 46 million segregating sites remained in the data set . In order to maximize the accuracy of our model , we further filtered the data to include sites segregating only at intermediate frequencies ( MAF ≥ 0 . 1 ) , resulting in a total of 414 , 049 sites . The distribution of coverage for each filtered pool-seq data point is plotted in S2 Fig . In addition to pooled sequencing data , whole-genome haplotype sequences were collected for 29 founder individuals ( see [36] for details ) . This enabled us to estimate local linkage disequilibrium for use in the multi-locus model . We employed a two-pass approach to analyze the data . In the first pass , we performed a genome-wide scan of the entire data set using the single-locus implementation of our model . Using the results of this first pass , we identified regions of the genome for which there was strong evidence of non-neutrality . We then fit more computationally demanding 3- , 5- , and 7-locus models in these genomic regions in order to localize and estimate the strength of selection . Further details of our analysis procedure are provided in Methods . Total run-time for the one-locus portion of the analysis was 8 hours 43 minutes for the entire genome ( ≈ 0 . 07 seconds per site ) , using a parallel implementation on a 16-core machine . For the multi-locus models , the average running time per site was 0 . 94 seconds ( 3 loci ) , 2 . 54 seconds ( 5 loci ) and 4 . 96 seconds ( 7 loci ) . Memory consumption for the multi-locus models averaged around 40 GB , although this can be reduced at the expense of greater run-time by disabling result caching features built into our software . The first pass identified the following 16 intervals ( in Mb ) for further analysis: Chr X: ( 1 . 6 , 1 . 7 ) ; Chr 2L: ( 15 . 0 , 16 . 0 ) , ( 16 . 5 , 18 . 5 ) , ( 19 . 0 , 20 . 7 ) ; Chr 2R: ( 20 . 9 , 21 . 1 ) ; Chr 3L: ( 2 . 3 , 3 . 0 ) , ( 6 . 6; 6 . 7 ) , ( 8 . 6 , 8 . 8 ) , ( 13 . 0 , 14 . 5 ) , ( 15 . 2 , 16 . 0 ) , ( 18 . 0 , 18 . 9 ) , ( 20 . 2 , 20 . 8 ) ; Chr 3R: ( 14 . 3 , 14 . 7 ) , ( 15 . 7 , 16 . 1 ) , ( 18 . 4 , 19 . 0 ) , ( 26 . 2 , 26 . 4 ) . Focusing on these regions , we computed the LR test statistic at about 37 , 000 SNPs in total for each multi-locus model . Because of long-range linkage disequilibrium and hitchhiking effects [36] , all models produced rather large LR statistics for numerous sites . However , compared to the one-locus model , multi-locus models generally produced more distinctive peaks in the LR statistic . For example , Fig . 7 illustrates a 200 kb region of chromosome arm 3R for which the one-locus analysis resulted in several distant SNPs with comparably high LR values , while all multi-locus models highlighted two nearby SNPs ( illustrated in red ) in the 14 . 615–14 . 619 Mb region with pronounced LR peaks . S3 Fig is another example of size 800 kb for which every multi-locus model yielded a distinctive peak ( shown in red ) near 18 . 205 Mb of chromosome arm 3L , while the one-locus model did not single out any particular SNPs in the region . To deal with variable results across different multi-locus models , we used the following strategy: For each of 3- , 5- , and 7-locus models , we first ranked the SNPs according to their LR statistic and took the top 100 SNPs . This corresponds to the LR statistic being greater than 8 . 741 , 9 . 525 , and 11 . 310 for the 3- , 5- , and 7-locus model , respectively . ( Shown in S4 Fig are empirical cumulative distributions of the LR statistic for each multi-locus model; the 99th percentile for the 3- , 5- , and 7-locus models are 6 . 883 , 7 . 330 , and 8 . 257 , respectively . ) Then , we took the intersection of the resulting three top 100 lists . This led to thirteen SNPs , nine of which belong to five coding genes ( one SNP in CG42334 and two SNPs each in CG9726 , CG33991 , CG17697 , and CG7720 ) . In particular , gene CG7720 actually resides in the region illustrated in Fig . 7 , and the two distinctive SNPs mentioned in the previous paragraph are in fact the two top ranking SNPs contained in CG7720 . Allele frequency trajectories of the thirteen identified SNPs are illustrated in S5 Fig; they generally display an increasing trend over the time course of the experiment . A brief description of the five genes is provided in Table 6 . It is well known that temperature affects the cell membrane composition [40] , and it is interesting that one of the five genes we identified is involved in transmembrane transport . It is also interesting that two of the remaining genes are related to cytoskeleton ( reorganization and coordination ) . Using the same data , Franssen et al . [36] recently studied the evolving pattern of linkage disequilibrium and identified 17 haplotype-blocks putatively under selection . Interestingly , three of the five genes mentioned above—namely , CG33991 , CG17697 , CG7720—are contained in that set of haplotype-blocks .
In this paper we have presented a model for analyzing time series data generated by evolve-and-resequence experiments . Our model is designed to analyze multiple recombining sites evolving in a moderately-sized population and potentially affected by measurement error . On data obtained from simulated E&R experiments combined with pooled sequencing , we have shown that it is possible to detect , localize and estimate the strength of selection in the range s ∊ [0 . 01 , 0 . 10] in a population of moderate size ( N ∼ 103 ) and using a moderate number ( R = 3 ) of experimental replicates . We have also explored the effect of the founding population composition ( in terms of the number of founders ) and sequencer effort ( coverage depth , number of sampling time points , and time intervals between sampling ) on the quality of these estimates . Finally , we have shown that our method can also be applied to study other phenomena of interest , including overdominance and effective population size; in particular , our work suggests that E&R data can be used to estimate recombination rates in putative hotspots in model organisms inferred by previous studies [5 , 41 , 42] . Space and time considerations have necessarily prevented us from considering many other combinations of experimental parameters which could be informative when designing E&R experiments . To enable other researchers to explore these options , we have made the computer code used in this study publicly available . We have also applied our method to analyze genome-wide data from a real E&R experiment of D . melanogaster adapting to a new laboratory environment over tens of generations . Because of the small population size involved in that particular E&R experiment , LD does not break down fast enough over the time scale of the experiment , and long-range correlation between distant sites and hitchhiking effects pose challenges to localizing the true sites under selection . In our work , we have observed that combining information from several multi-locus models may produce improved results . We have employed a heuristic ensemble approach in this paper; further statistical work on this problem would be worthwhile to pursue in the future . In a given multi-locus model , we have noticed that choosing appropriate SNPs to include in the model is important for producing cleaner signals . Specifically , we recommend choosing SNPs for which the allele frequency does not get too close to the boundary ( 0 or 1 ) and that are sufficiently far apart ( e . g . , > 100 kb apart for the particular E&R data we considered ) . Our analysis of the E&R data has identified five genes in D . melanogaster ( Table 6 ) which may be involved in adaptation , and some of these genes reside in haplotype-blocks recently identified as candidate regions of selection [36] . Further , some of the genes we have identified are involved in related biological processes , in particular concerning cytoskeleton and transmembrane transport . It would be interesting to investigate this thread of observations further . We note that we have employed a rather conservative approach in our analysis , so it is likely that we missed several other regions potentially under selection . Experience has shown that the running time of our model is dominated by the recursive procedure used to calculate covariances between pairs of sites ( see Methods ) . Thus , to fit a K-locus model sampled at T time points has computational complexity of order O ( K2 T2 ) . When performing the large number of simulations needed to benchmark our model , this quadratic scaling in the model size K prevented us from fitting models jointly using many more sites . Since our results suggest that estimation precision can be improved ( in particular , at low coverage ) by exploiting linkage information between sites , it could make sense in practice to expend additional computation time in order to add more sites into the model . It is interesting to compare our findings with existing results . Feder et al . [30] suggest that power to detect selection is maximized when ( positively ) selected alleles are sampled as they rise in frequency , but before they have fixed . By a simple modification of their argument , the expected strength of selection required for a mutation in our simulated E&R experiments to achieve frequency xf in T time periods is given by s fix ( T ) = 1 H F - 1 ∑ k = 1 F - 1 1 k T log x f 1 - x f · F - k k , ( 7 ) where H n := ∑ i = 1 n 1 / i is the harmonic series . Above we generally chose T = 50 and F = 200; for xf = 0 . 95 we find that sfix ( T ) ≈ 0 . 11 which roughly agrees with our finding ( Fig . 1 ) that medium and strong selection ( s = 0 . 1 ) could be reliably detected , while weaker selection was fairly difficult to detect . Our findings are somewhat more optimistic than those of Baldwin-Brown et al . [31] , whose simulation results suggest that E&R experiments require a fairly large number of experimental replicates ( R ≥ 25 ) , founder haplotypes ( F ≥ 500 ) and strong selection ( s ≥ 0 . 1 ) in order to reliably detect and localize selected sites in a 1 Mb region . Since we used a smaller region for simulation ( L = 100 kb ) , the results we report are not directly comparable; nevertheless , it is interesting that with many fewer replicates and haplotypes ( R = 3 and F = 20 ) we could reliably detect the selected site in at least 50% of trials ( Table 1 ) . With sampled data the problem becomes harder , but we found that average coverage depth 30 still sufficed to discover the selected site from among the top four segregating sites in 50% of trials ( Table 3 ) . Several extensions to our model could potentially be of use . In our simulations we assumed that sequencer coverage depth is Poisson distributed . However , some studies have noted that coverage depth is overdispersed relative to the Poisson distribution , in which case an alternative distribution such as the negative binomial is preferred . For multi-locus estimation problems , our model requires that the haplotypic structure of the founding experimental population be known . In cases where this information is not known exactly , a Bayesian approach could be adopted in which model results are weighted by a prior on the space of initial haplotypic configurations . Such a procedure could allow the researcher to trade sequencing effort for computation time by decreasing the burden of initial sequencing that must be performed in order to establish the haplotypes of the founding lineages . The other extreme of sequencing effort is to obtain haplotype data for a sample of individuals at each sampling generation , rather than to use pooled sequencing to infer only marginal allele frequencies . ( Indeed , there is a discussion on the utility and power of pooled sequencing [37 , 43–45] . ) The same multi-locus model underlying our approach can be applied to develop a method for analyzing haplotypic time series data , and we will explore incorporating such an extension into our method . Our approximation to the multi-locus Wright-Fisher process relies on a system of recursions which describe the evolution of neutral sites conditional on the presence of a linked selected site ( see Methods ) . The process of generating those recursions has been automated [46] to handle more general scenarios including population structure and interaction between multiple selected sites . Our model could therefore be extended to handle these more complex scenarios at the expense of ( potentially significantly ) greater computational effort and data requirements . For datasets consisting of a small number of time intervals , or which are sampled at low coverage , allele frequency trajectories may be very noisy , making it difficult to reliably detect the presence ( or absence ) of selection . In these cases , it could be useful to decrease the variance of our estimates by including many more segregating sites into the model in hopes of “averaging out” the noise . The quadratic time complexity of our method makes this difficult to achieve , but alternatives could be explored . These could include approximating the covariance matrix used in the model by something which is faster to compute , ( for example , using the Matérn covariance function ) , or using an ensemble approach whereby a large number of small models are fit simultaneously to the same putative selected site and at various linked neutral sites .
As described above , in the case of neutrality it suffices to consider covariances between pairs of sites in a two-locus haploid model . The one-generation transition function of the neutral two-locus Wright-Fisher model with recombination fraction r is f : Δ 3 → Δ 3 Z t ↦ Z t + r C t ϵ ( 8 ) where ϵ ≡ ( −1 , 1 , 1 , −1 ) and C t ≡ Z t ( 1 ) Z t ( 4 ) − Z t ( 2 ) Z t ( 3 ) is the linkage disequilibrium at time t . Thus , conditional on Zt we have that 2N × Zt+1 is multinomially distributed according to f ( Zt ) : 2 N Z t + 1 ∣ Z t ∼ Multinomial ( 2 N , f ( Z t ) ) . ( 9 ) ( Note that the multinomial distribution which arises in this equation is due to the random sampling of gametes to form generation t + 1 , and is different from the binomial sampling scheme described earlier ( equation 2 ) which was resulted from sampling biallelic sites using sequencer . ) Using equation ( 9 ) , we can derive an accurate approximation to the evolution of the covariance of the Zt process . In what follows we let π = ( z ( 1 ) , z ( 2 ) , z ( 3 ) , z ( 4 ) ) and c0 = z ( 1 ) z ( 4 ) − z ( 2 ) z ( 3 ) denote the initial distribution and linkage disequilibrium of the Wright-Fisher process under consideration . Lemma 1 . To order O ( r + 1 2 N ) , 𝔼 π Z t ( i ) = z ( i ) + ϵ i t r c 0 1 - t - 1 4 N 𝔼 π ( r Z t ( i ) Z t ( j ) ) = r 2 N z ( i ) z ( j ) ( 2 N - t ) + t z ( i ) 1 { i = j } 𝔼 π ( r Z t ( i ) C t ) = r 2 N z ( i ) c 0 2 N - 3 t + t 2 ( 1 - ϵ i ) z ( 1 ) z ( 4 ) - ( 1 + ϵ i ) z ( 2 ) z ( 3 ) . Corollary 2 . To order O ( r + 1 2 N ) , 𝔼 π Z t ( i ) Z t ( j ) = z ( i ) z ( j ) + ϵ i ϵ j t r c 0 ( ϵ i z ( i ) + ϵ j z ( j ) ) + t 2 N - z ( i ) z ( j ) 1 { i ≠ j } + z ( i ) ( 1 - z ( j ) ) 1 { i = j } r t 2 N { 1 2 t + 1 - | ϵ i - ϵ j | z ( 1 ) z ( 4 ) + z ( 2 ) z ( 3 ) - ϵ i ϵ j c 0 ( 2 t - 1 ) ( ϵ i z ( i ) + ϵ j z ( j ) ) - 1 8 | ϵ i + ϵ j | c 0 ( ϵ i + ϵ j ) ( t + 1 ) 1 { i ≠ j } + 4 t ( ϵ i + 1 ) z ( 2 ) z ( 3 ) + ( 1 - ϵ i ) z ( 1 ) z ( 4 ) } . Proofs of the above results are given in S1 Text . These results can be combined to give an O ( r + 1 2 N ) approximation to the within-generation covariance cov π ( Z t ( i ) , Z t ( j ) ) . Using the same approach , we can also approximate the covariance between generations . Indeed , by Lemma 1 and the Markov property , 𝔼 π Z t + u ( i ) ∣ Z t = 𝔼 Z t Z u ( i ) = Z t ( i ) + ϵ i u r C t 1 - u - 1 4 N . Hence , 𝔼 π ( Z t + u ( i ) , Z t ( j ) ) = 𝔼 π Z t ( i ) Z t ( j ) + ϵ i u r Z t ( j ) C t 1 - u - 1 4 N and each of the expectations on the right-hand side is given to order O ( r + 1 2 N ) by the preceding results . Remark . The constants subsumed in the O ( r2+1 ( 2N ) 2 ) terms in the above expressions increase as t increases; in particular , we would not expect the approximation to be accurate if tr ∊ O ( 1 ) . For our application typically t ≪ 1/r . Computations in the non-neutral case are more involved because the transition operator f ( Zt ) is a rational function of its arguments . This results in moments of Zt+1 depending on all moments of Zt . To illustrate the issues involved , consider first the simplest possible example of a one-locus Wright-Fisher model with diploid selection and no mutation [39] . The relative fitnesses of A0/A0 and A1/A1 homozygote genotypes are given by 1 and 1 + s , respectively , whereas the relative fitness of the A0/A1 heterozygote is 1 + hs . The frequency of the A1 allele at time t is denoted Xt . Conditional on Xt , 2N×Xt+1 has a binomial distribution with 2N trials and success parameter f ( Xt ) , where f ( x ) = x + s [ h + ( 1 - 2 h ) x ] x ( 1 - x ) 1 + s x [ 2 h + ( 1 - 2 h ) x ] . ( 10 ) We cannot apply the method described in the preceding subsection due to the appearance of x in the denominator of ( 10 ) . Hence , a different form of approximation is required . First , we formally decompose Xt as Xt=X¯t+δXt , where X¯t=f ( X¯t−1 ) equals the deterministic trajectory that would be followed by Xt in the infinite-population limit , and δXt is a random disturbance away from the deterministic path due to genetic drift . Next , we expand 𝔼 ( Xt ) in a Taylor series about this deterministic path: 𝔼 ( X t ) = 𝔼 ( f ( X t - 1 ) ) = 𝔼 ( f ( X ¯ t - 1 + δ X t - 1 ) ) ≈ f ( X ¯ t - 1 ) + d f d x X ¯ t - 1 × 𝔼 ( δ X t - 1 ) + 1 2 d 2 f d x 2 X ¯ t - 1 × 𝔼 [ ( δ X t - 1 ) 2 ] . This yields a recursion for computing 𝔼 ( Xt ) in terms of moments of the disturbance term in the preceding time period , 𝔼[ ( δXt−1 ) u] , u = 1 , 2 . Since also 𝔼 ( X t ) = X ¯ t + 𝔼 ( δ X t ) = f ( X ¯ t - 1 ) + 𝔼 ( δ X t ) , these terms themselves obey the recursion 𝔼 ( δ X t ) ≈ d f d x X ¯ t - 1 × 𝔼 ( δ X t - 1 ) + 1 2 d 2 f d x 2 X ¯ t - 1 × 𝔼 [ ( δ X t - 1 ) 2 ] which is a recursion for computing 𝔼 ( δXt ) in terms of the moments of δXt−1 . Inductively assuming that we can compute 𝔼[ ( δXt ) u] for u = 1 , 2 , this enables us to compute 𝔼 ( Xt ) and var ( Xt ) = var ( δXt ) . This approach was previously employed by Barton et al . [47] to obtain order O ( 1/N ) approximations to these moments . Here we have used the same idea but automated the symbolic algebra and code generation needed to generate the recursions to higher orders of accuracy . The above idea can be extended to multiple loci in a straightforward manner . ( As we describe in the next subsection , we only require models of size up to L = 3 for our purposes , but we state it in full generality here . ) Recall Z t = ( Z t ( 1 ) , … , Z t ( 2 L ) ) ∊ Δ 2 L − 1 . Conditional on Zt , the vector 2N × Zt+1 is multinomially distributed with success probabilities f ( Zt ) . The form of f:Δ2L−1 → Δ2L−1 varies according to the underlying model; we describe our choice of f in the following subsection . As in the one-locus case , write Z t ( i ) = Z ‾ t ( i ) + δ Z t ( i ) where Z ‾ t ( i ) is the deterministic trajectory which would be followed by Z t ( i ) in the infinite-population limit , and δ Z t ( i ) is a random disturbance . ( Note that in general , 𝔼 ( δ Z t ( i ) ) ≠ 0 for t > 1 . ) For u , v non-negative integers , we have 𝔼Z t ( i ) u Z t ( j ) v = 𝔼 ( Z ¯ t ( i ) + δ Z t ( i ) ) u ( Z ¯ t ( j ) + δ Z t ( j ) ) v = 𝔼 ( Z ¯ t ( i ) + δ Z t ( i ) ) u ( Z ¯ t ( j ) + δ Z t ( j ) ) v - δ Z t ( i ) u δ Z t ( j ) v + 𝔼 δ Z t ( i ) u δ Z t ( j ) v . ( 11 ) From the conditional distribution 2N Zt∣Zt−1 ∼ 𝓑 ( 2N , f ( Zt−1 ) ) , we have ( 2 N ) u + v · 𝔼 Z t ( i ) u Z t ( j ) v ∣ Z t - 1 = g i j ( f ( Z t - 1 ) ) = g i j ( f ( Z ¯ t - 1 + δ Z t - 1 ) ) , where gij ( z ( 1 ) , … , z ( 2L ) ) is a polynomial in z ( 1 ) , … , z ( 2L ) which can be computed using the moment generating function of the multinomial distribution . By performing a Taylor expansion of hij ≡ gij ∘ f about the deterministic path Z ‾ t − 1 and taking expectations , we get another formula for 𝔼[ ( Zt ( i ) ) u ( Zt ( j ) ) v] in terms of moments of δ Zt−1: 𝔼 Z t ( i ) u Z t ( j ) v ≈ h i j ( Z ¯ t - 1 ) + ∑ l ∂ h i j ∂ z ( l ) Z ¯ t - 1 𝔼 ( δ Z t - 1 ( l ) ) + 1 2 ∑ l , m ∂ h i j ∂ z ( l ) ∂ z ( m ) Z ¯ t - 1 𝔼 ( δ Z t - 1 ( l ) δ Z t - 1 ( m ) ) . ( 12 ) For u + v ≤ 2 , comparing ( 11 ) and ( 12 ) yields a recursion for computing 𝔼[ ( δZt ( i ) ) u ( δZt ( j ) ) v] in terms of moments of δ Zt of total degree strictly less than u + v , and moments δ Zt−1 of total degree at most u + v . The latter feature is important for computation because it implies that we only need to compute a bounded number of terms in each recursive step , which would not be the case if we had instead expanded the function hij ( ⋅ ) about zero with respect to model parameters ( for example , selection or mutation ) . The recursive nature of the above algorithm lends itself to computing moments of the form cov ( δ Z t + m ( i ) , δ Z t ( j ) ) . Stopping the recursion m time steps into the past , we obtain an expression of the form 𝔼 ( δ Z t + m ( i ) ∣ δ Z t ) = 𝑝 i m ( δ Z t ) , where 𝑝im ( z ( 1 ) , … , z ( 2L ) ) is a polynomial . Hence , 𝔼 ( δ Z t + m ( i ) δ Z t ( j ) ) = 𝔼 ( δ Z t ( j ) 𝑝 i m ( δ Z t ) ) is again a recursion involving moments of δ Zt which can be solved using the techniques described above . When selection is acting on a nearby linked site , some additional care is needed in computing the first- and second-order moments for neutral sites . For example , the hitchhiking effect will cause these moments to be different from they would be in the absence of linked selection . Consider a three-locus model with Xt = ( Xt , 1 , Xt , 2 , Xt , 3 ) , where Xt , j denotes the marginal allele frequency at time t at locus j . Suppose the site corresponding to Xt , 1 is under positive selection , and the remaining sites are neutral and under positive LD with the selected site . Computing 𝔼Xt , 2 using a one-locus neutral model as described above will produce an underestimate since linkage with site 1 will cause site 2 to rise in frequency faster than what is expected under neutrality . A similar effect can be seen when computing 𝔼 ( Xt , 2 Xt , 3 ) . To capture this effect it is necessary to condition on the presence of a linked selected site when performing the moment calculations discussed earlier for neutral sites . To carry this out we utilize a three-locus model which describes the evolution of two neutral and one linked selected site over time . This model was derived by Stephan et al . [48] using the general framework of Kirkpatrick et al . [46] . In the notation of the preceding subsection , we let L = 3 and obtain the transition function f using the system of recursions presented in equations ( 1 ) – ( 11 ) of [48] . This system can then be expanded in terms of the random disturbance δ Zt to yield the system of recursions ( 11 ) and ( 12 ) . The differentiation steps needed to perform the expansion involve a very large number of terms , and are too complex to perform by hand . Instead , we automated these computations using the symbolic algebra package Maple . Code to automatically generate these recursions is included in the source code accompanying this paper . Our procedure for simulating an E&R experiment was the following . To generate realistic patterns of standing variation , a set of F founder lines was sampled from the coalescent with recombination using the program ms [49] . ( The exact ms command-line used for each simulation was: ms <F> 1 -t <4μLNe> -r <4Ne ( L − 1 ) r> <L> , where the variables in angled brackets are computed using the values described in the text . ) Recombination and mutation rates and the effective population size were set to biologically plausible values for D . melanogaster , a common model organism used in E&R studies ( r = 2 × 10−8/bp/gen , μ/2 = 10−9/bp/gen , N = 106 ) [50] . Each founder line was cloned 2N/F times to generate an initial diploid population of size N . This replication step is intended to mimic the practice using of ( nearly- ) homozygous recombinant inbred founder lines to initialize an E&R experiment . Next , the experimental population of size N was simulated forward in time using the discrete-time simulator simuPOP [51] . Finally , alleles were sampled binomially and independently at each locus and time point to simulate next-generation sequencing . Parameters for the forward simulation and sampling were varied from scenario to scenario as described in the main text . The output of the simulation consisted of the haplotypes of the initial founder lines and the frequency of each segregating site ( potentially after sampling ) at each time point . All simulations were performed on a machine with 2 × 2 . 5 GHz AMD Opteron 6380 processors ( 32 cores total ) and 256 GB of memory . In our model , we used an effective population size of 200 , as previously estimated for the E&R data we considered [25] . To prevent our estimates from becoming confounded by the action of genetic drift , we restricted our analysis to only those sites which were segregating at intermediate frequencies throughout the experiment . Specifically , we only considered sites which were segregating at frequencies in the interval [0 . 1 , 0 . 9] for all generations and replicates . A total of 414 , 049 sites remained after filtering . First , we computed the one-locus likelihood-ratio statistic at each of the 414 , 049 sites , comparing the fitted model to the null ( neutral ) model . Then , we partitioned the genome into non-overlapping windows of a fixed size ( we considered various window sizes , including 5 kb , 10 kb , 50 kb , 100 kb , 200 kb , 500 kb , and 1 Mb ) and computed the average one-locus LR statistic over the SNPs in each window . By visually inspecting plots of these quantities , we identified regions of the genome which were enriched for SNPs that potentially behaved non-neutrally . For each region identified , a multi-locus model was then estimated for each segregating site within the region . Specifically , we fit a model in which each site in the region was posited to be under selection , and the trajectories of linked neutral sites were affected due to hitchhiking . To choose which linked neutral to include in the model , we identified SNPs which were segregating at multiples of approximately 250 kb from the midpoint of the region . For example , to analyze the region 6 . 6–6 . 7 Mb on chromosome 3L using a five-site model , we first fixed four SNPs segregating at intermediate frequencies near positions 6 . 15 Mb , 6 . 4 Mb , 6 . 9 Mb and 7 . 15 Mb . For each site between 6 . 6 Mb and 6 . 7 Mb , we then estimated the strength of selection s using the five-locus model containing the selected site plus the four fixed neutral sites . | A growing number of experimental biologists are generating “evolve-and-resequence” ( E&R ) data in which the genomes of an experimental population are repeatedly sequenced over time . The resulting time series data provide important new insights into the dynamics of evolution . This type of analysis has only recently been made possible by next-generation sequencing , and new statistical procedures are required to analyze this novel data source . We present such a procedure here , and apply it to both simulated and real E&R data . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] | [] | 2015 | Multi-locus Analysis of Genomic Time Series Data from Experimental Evolution |
Elevated levels of acute-phase serum amyloid A ( A-SAA ) cause amyloidosis and are a risk factor for atherosclerosis and its clinical complications , type 2 diabetes , as well as various malignancies . To investigate the genetic basis of A-SAA levels , we conducted the first genome-wide association study on baseline A-SAA concentrations in three population-based studies ( KORA , TwinsUK , Sorbs ) and one prospective case cohort study ( LURIC ) , including a total of 4 , 212 participants of European descent , and identified two novel genetic susceptibility regions at 11p15 . 5-p13 and 1p31 . The region at 11p15 . 5-p13 ( rs4150642; p = 3 . 20×10−111 ) contains serum amyloid A1 ( SAA1 ) and the adjacent general transcription factor 2 H1 ( GTF2H1 ) , Hermansky-Pudlak Syndrome 5 ( HPS5 ) , lactate dehydrogenase A ( LDHA ) , and lactate dehydrogenase C ( LDHC ) . This region explains 10 . 84% of the total variation of A-SAA levels in our data , which makes up 18 . 37% of the total estimated heritability . The second region encloses the leptin receptor ( LEPR ) gene at 1p31 ( rs12753193; p = 1 . 22×10−11 ) and has been found to be associated with CRP and fibrinogen in previous studies . Our findings demonstrate a key role of the 11p15 . 5-p13 region in the regulation of baseline A-SAA levels and provide confirmative evidence of the importance of the 1p31 region for inflammatory processes and the close interplay between A-SAA , leptin , and other acute-phase proteins .
Serum amyloid A ( SAA ) is a sensitive marker of the acute inflammatory state . Its isoforms are expressed constitutively ( C-SAA ) and show a rapid ( up to 1000-fold ) increased expression in response to inflammatory stimuli such as trauma , infection , injury , and stress during the acute phase ( A-SAA ) [1] . The high inductive capacity along with a high conservation of genes and proteins throughout evolution of vertebrates and invertebrates suggests that A-SAA plays a key role in pathogen defence and probably functions as an immune-effector molecule [1] . Acute inflammation has mainly beneficial effects in restoring homeostasis . However , in recent years , clinical and epidemiological studies have gathered substantial evidence that A-SAA is associated with obesity [2] and that prolonged and recurrent chronic infection as well as inflammation is causally involved in the pathogeneses of amyloidosis [1] . Furthermore , it induces , promotes , or influences susceptibility to several chronic diseases such as atherosclerosis and its clinical complications [3]–[9] , type 2 diabetes [10] , [11] , and various malignancies [12] . The SAA gene family is located within 150 kb at chromosome 11 and comprises of four genes: SAA1 and SAA2 , the bona fide acute-phase SAA isoforms , SAA3 , a pseudogene in humans , and SAA4 , a low level expressed gene coding for C-SAA [13] , [14] . A-SAA expression is regulated by a variety of stimuli , including the pro-inflammatory cytokines TNF-α and IL-6 , as well as glucocorticoids [15] , [16] . Like other acute-phase proteins , A-SAA is expressed primarily by the liver [17] . However , extra-hepatic expression has been reported for different cell lines like epithelial cells , monocyte and macrophage cells , most endothelial cells , adipocytes , atherosclerotic lesions , and smooth muscle cells [17] . Twin studies suggest a substantial genetic contribution to baseline A-SAA concentrations with heritability estimates of 59% ( 95% confidence interval , 49–67% ) [18] . The identification of genetic variants that are predisposed to elevated levels of A-SAA could provide important clues to the immune response pathways involved in the regulation of A-SAA levels which might also be of relevance for related clinical entities . In the past , association analyses between genetic variants and A-SAA levels were limited and restricted to allelic variants of SAA genes and protein concentrations [19]–[21] . We therefore conducted the first genome-wide association study on baseline A-SAA concentrations . In a meta-analysis of four genome-wide scans ( KORA S4 , LURIC , TwinsUK and Sorbs ) we included 4 , 212 participants of European descent . Additionally , in order to account for known gender-specific differences in the regulation of A-SAA [22] , [23] we stratified the analysis by gender .
In the present meta-analysis of four genome wide scans 106 SNPs distributed across two regions showed genome-wide significant associations with p-values below the threshold of 5×10−8 ( Figure 1 , Table S1 ) . Table 1 shows study specific results for the top hits within the two regions and three identified subregions ( see below ) of the meta-analysis as well as an additional region for men in the gender stratified analysis . Genotypic mean levels are provided in Table S2 . Results of the single genome-wide studies were consistent across all four studies regarding the direction and magnitude of the effects . In addition , results were consistent between different genotyping technologies ( Table S3 ) . No deviations from the Hardy-Weinberg-Equilibrium were observed . The variable of inter-study heterogeneity ( I2 ) showed homogeneity at the 1p31 locus . At the 11p15 . 5-p13 locus we observed I2 values that indicated a more distinct heterogeneity . This reflects the relatively large and varying beta values and differences in the minor allele frequency ( Table S1 ) . However , taking into account that this locus was clearly significantly associated with A-SAA in all studies included in the meta-analysis , results of the meta-analysis are reported based on a fixed effect model . The first region ( 193 . 3 kb of length ) resides at 11p15 . 5-p13 and includes SAA1 one of the structure genes of A-SAA . Within this region the strongest association was found for two highly correlated intronic polymorphisms of the general transcription factor 2 H1 ( GTF2H1 ) gene , rs4150642 ( p = 3 . 20×10−111 ) and rs7103375 ( p = 3 . 26×10−111 ) ( Figure 2A ) . These two top hits show modest correlation ( r2≤0 . 376 ) with other significantly associated SNPs within this region . When the structure of correlation and explained variances within the region were analysed three mostly independent subregions were identified ( Table S4 , Figure 2B–2D ) . The first subregion encloses the 5′ end of SAA1 ( Figure 2B ) with strongest association for rs4638289 ( p = 2 . 77×10−53 ) . The other two subregions harbour the genes Hermansky-Pudlak Syndrome 5 ( HPS5 ) and GTF2H1 ( Figure 2C ) and lactate dehydrogenase A and C ( LDHA and LDHC ) ( Figure 2D ) with strongest associations for rs4353250 ( p = 1 . 68×10−51 ) and rs2896526 ( p = 4 . 12×10−22 ) , two intronic polymorphisms of HPS5 and LDHA , respectively . The second region was detected at 1p31 ( Figure 2E ) . All 38 significantly associated variants cluster around the 3′ end of the leptin receptor gene ( LEPR ) . The most significantly associated SNP , rs12753193 , ( p = 1 . 22×10−11 ) is located downstream of LEPR . All associations were consistent in the KORA S4 validation analyses ( Table 1 for the top hits , Table S5 for all SNPs selected for validation ) . The entire regression model ( including the top SNPs of the two genomic regions ( rs4150642 and rs12753193 ) , age , gender and BMI ) explains 19 . 32% of the total variation of A-SAA in our data . With an explained variance of 10 . 84% for the top SNP ( rs4150642 ) of the 11p15 . 5-p13 locus ( 5 . 57% for rs4638289 of the SAA1 subregion , 5 . 34% for rs4353250 of the HPS5/GTF2H1 subregion , and 2 . 37% for rs2896526 of the LDHA/LDHC subregion; Table S4 ) and 0 . 93% for the top SNP ( rs12753193 ) of the 1p31 locus the identified genomic regions account for a major part of such variance . When the analysis was stratified by gender , one additional SNP ( rs549485 ) located about 350 kb apart from the SAA1 subregion at 11p14 in the secretion regulating guanine nucleotide exchange factor ( SERGEF ) gene showed a borderline significant association with A-SAA levels in men ( p = 2 . 76×10−8 ) in the meta-analysis . In the validation analysis the association between two highly correlated SNPs within this region ( rs493767 and rs550659 , r2 = 0 . 961 ) and A-SAA levels was also borderline significant ( p = 8 . 50×10−3 and p = 1 . 65×10−2 , respectively ) . No significant differences between men and women were found within the regions identified in the overall meta-analysis ( data not shown ) .
Based on a meta-analysis of four genome-wide association studies including 4 , 212 participants of European descent two novel genetic susceptibility regions were identified to be associated with baseline A-SAA concentrations . With 11 . 68% explained variance in our data , which makes up 19 . 76% of the total estimated heritability of 59% , these two regions seem to have a major impact on baseline A-SAA concentrations . The region at 11p15 . 5-p13 accounts for most of the explained variance . Its SAA1 subregion contains part of a highly conserved region between the two bona fide acute-phase structure genes SAA1 and SAA2 , which consist of at least 5 and 2 allelic variants , respectively [1] , [24] . These two genes are concurrently induced during the acute-phase [1] , and cluster within 18 kb of each other in a head to head arrangement [25] . This study is the first presenting the complex genetic architecture of A-SAA levels at this locus . In the identified region , there has been evidence of regulatory elements like C/EBPalpha and C/EBPbeta ( http://genome . ucsc . edu ) , which are necessary for the full responsiveness to IL-1β and IL-6 either alone or in combination [1] . Our finding underlines the high functional potential for this region . The adjacent GTF2H1 is a basal transcription factor involved in nucleotide excision repair of DNA and RNA transcription by RNA polymerase II [26] . HPS5 encodes a protein , which is probably involved in organelle biogenesis associated with melanosomes and platelet dense granule , its mutations lead to a homonymous clinical entity [27] . And LDHA and LDHC , which are expressed in muscle tissue and in testes , respectively , encode for lactate dehydrogenase , an enzyme that catalyzes the interconversion of lactate and pyruvate [28] . Variants of the GTF2H1 gene have been recently found to be associated with lung cancer in a Chinese population [29] . Furthermore , it was demonstrated that LDHA is involved in tumour genecity and its reduction causes bioenergetic and oxidative stress leading to cell death [30]–[32] . Finally , Kosolowski et al . [33] found LDHC to be expressed in several types of tumour cell lines . It is thought , that recurrent or persistent chronic inflammation may play a role in carcinogenesis by causing DNA damage , inciting tissue reparative proliferation and/or by creating an environment that is enriched with tumour-promoting cytokines and growth factors [12] . Furthermore , SAA synthesis could be found in human carcinoma metastases and cancer cell lines [17] . As the approach taken in this study is observational in nature it is not possible to draw causal inferences . For that reason , it could be possible that not genes , but small regulatory elements may be responsible for the findings . This is most likely the case as the identified region contains one structure gene and the adjacent region . In any case , the major impact on baseline A-SAA concentrations demonstrates a key role of the 11p15 . 5-p13 region in the regulation of inflammation . Therefore , the identification of causal variants and their impact on diseases related to elevated baseline A-SAA concentrations represent promising targets for future functional and epidemiological studies . The second region was found on chromosome 1p31 , harbouring the LEPR gene locus . Leptin , an important circulating signal for the regulation of body weight , was found to be correlated with SAA concentrations independently of BMI , and both were expressed in adipose tissue [34] . In the KORA F3 study ( Text S1 ) a moderate but significant correlation was found between circulating A-SAA and leptin concentrations in blood in 181 participants with measurements of both proteins ( Spearman correlation = 0 . 25 , p = 7×10−4 ) . So far it is unclear whether leptin influences SAA expression directly or via the leptin stimulated cytokines , IL-6 and TNF-α [34] . LEPR is a single transmembrane receptor of the cytokine receptor family most related to the gp130 signal-transducing component of the IL-6 receptor , the granulocyte colony-stimulating factor ( GCSF ) receptor , and the Leukaemia Inhibitory Factor ( LIF ) receptor , all of which are thought to play an essential role in the inflammatory process [35] , [36] . Previous studies have provided evidence of an association of the LEPR gene locus with CRP and fibrinogen [37]–[40] , which were both correlated with A-SAA in the KORA S4 study ( Text S1 ) ( CRP: Spearman correlation = 0 . 58 , p = 3 . 22×10−155 , and fibrinogen: Spearman correlation = 0 . 31 , p = 3 . 89×10−41; N = 1734 ) . The finding gives confirmative evidence of the importance of the LEPR gene locus for inflammatory processes and the close relationship between leptin , A-SAA , CRP and fibrinogen . Furthermore , in the gender stratified analysis one region containing SERGEF was identified to be presumably associated with A-SAA in men . The adjacency of this identified region to the SAA gene family suggests that regulatory elements may be responsible for this signal . However , the association with A-SAA levels was only borderline significant in our study and therefore awaits replication . Two limitations of our study have to be mentioned . Firstly , due to the restrictions in laboratory methods our analyses were confined to the A-SAA isoforms and did not capture the constitutively expressed C-SAA isoform which might also be of interest , especially when analyzing baseline SAA levels . Secondly , the number of studies with genome-wide data and measured A-SAA levels was limited compared to other genome-wide association studies . Nevertheless , the study had enough power to detect two novel genetic susceptibility regions for A-SAA which explain 19 . 76% of the total estimated heritability already . Furthermore , results were consistent across all four studies and within different genotyping platforms , the regions are biologically highly plausible , and the results may contribute to future research on the regulation of inflammatory response and its role in related clinical entities . Taken together , the present meta-analysis is the first whole genome approach to identify genetic variants that are associated with baseline A-SAA concentrations . Two novel genetic susceptibility regions were identified to be associated with baseline A-SAA concentrations . The findings demonstrate a major impact of the 11p15 . 5-p13 gene region on the regulation of inflammation and suggest a close interplay between leptin , A-SAA , and other acute-phase proteins as well as a larger role of the LEPR gene locus in inflammatory processes as it has been assumed in the past .
The present meta-analysis combined data from four genome-wide scans: one survey of the Cooperative Health Research in the Region of Augsburg ( KORA S4 ) , the Ludwigshafen Risk and Cardiovascular Health study ( LURIC ) , the UK Adult Twin Register ( TwinsUK ) and a self-contained population from Eastern Germany ( Sorbs ) ( ) . Approval was obtained by each of the local Ethic Committees for all studies and written informed consent was given by all study participants . In total , the meta-analysis included 4 , 212 individuals ( 1 , 928 males , 2 , 284 females ) of European ancestry with measured baseline A-SAA concentrations . For validation analyses we used data of 2 , 136 participants of the KORA S4 sample , which were not included in the meta-analysis ( Text S1 ) . Sample sizes and characteristics of the study participants of the four genome-wide scans and the validation sample are displayed in Table S6 . In all four studies , study participants were fasting and EDTA plasma samples were analyzed by immunonephelometry on a BNAII device from Siemens , Germany , and well-validated automated microparticle capture enzyme immunoassays [10] , [41] . The inter-assay coefficients of variation were below 7% in all four studies . For genotyping different platforms as the Affymetrix 500K GeneChip array ( Sorbs ) , Affymetrix 6 . 0 GeneChip array ( KORA S4 , LURIC , Sorbs ) , Illumina HumanHap300 BeadChip ( 317K ) ( TwinsUK ) and Illumina Human 610K BeadChip ( TwinsUK ) were used . Quality control before imputation was undertaken in each study separately . Detailed information on genotyping and imputation is reported in Table S7 . Imputation based on the HapMap Phase 2 CEU population was performed using IMPUTE [42] in all studies . After imputation all genotype data had to meet the following quality criteria: a minor allele frequency ≥0 . 01 , a call rate per SNP ≥0 . 9 , and r2 . hat metrics ≥0 . 40 for imputed SNPs . In total , 2 , 593 , 456 genotyped or imputed autosomal SNPs were analyzed in the meta-analysis . For validation and comparison of genotyping platforms , we selected 27 of the most significantly associated SNPs . Genotyping of these SNPs was performed with the MassARRAY system using the iPLEX technology ( Sequenom , San Diego , CA ) in the KORA S4 study . The allele-dependent primer extension products were loaded onto one 384-element chip using a nanoliter pipetting system ( SpectroCHIP , Spectro-POINT Spotter; Sequenom ) , and the samples were analyzed by matrix-assisted laser desorption-ionization time-of-flight mass spectrometry ( Bruker Daltonik , Leipzig , Germany ) . The resulting mass spectra were analyzed for peak identification via the SpectroTYPER RT 3 . 4 software ( Sequenom ) . To control for reproducibility , 9 . 8% of samples was genotyped in duplicate with a discordance rate of less than 0 . 5% . In each study , linear regression models for all available SNPs have been calculated on ln-transformed A-SAA levels in mg/l . The genetic effect has been assumed to be additive . Adjustment has been made for age , gender , BMI , and study specific covariates , i . e . the Friesinger Score in the LURIC population [43] and a genotyping batch variable in the TwinsUK population . Additionally , this analysis was undertaken stratified by gender . The genome-wide scans were calculated with the analysis software SNPTEST ( http://www . stats . ox . ac . uk/~marchini/software/gwas/snptest . html ) ( KORA S4 , LURIC ) QUICKTEST ( http://toby . freeshell . org/software/quicktest . shtml ) ( Sorbs ) and Merlin ( http://www . sph . umich . edu/csg/abecasis/Merlin/ ) ( TwinsUK ) . The results of all four genome-wide scans were meta-analysed using a fixed-effects model applying inverse variance weighting with the METAL software ( www . sph . umich . edu/csg/abecasis/metal ) . Study specific results were corrected for population stratification using the genomic control method . For the overall meta-analysis , the inflation factor was 1 . 009 . No further correction was applied . P-values below the threshold of p = 5×10−8 , which corresponds to a Bonferroni correction for the estimated number of one million tests for independent common variants in the human genome of European individuals [44] , were considered to be significant . As a measure for between study heterogeneity I2 was calculated [45] . Deviations from Hardy-Weinberg-Equilibrium were tested for all identified SNPs by means of the exact Hardy Weinberg test . For the calculation of explained variances , we subtracted the multiple R2 value of the covariate model from those of the full model including covariates and top hits of the loci in every single study and assessed the weighted mean ( KORA S4 , LURIC , and the Sorbs ) . We tested adjacent regions for independency by analyzing the significance of their top SNPs in a joint model . The OMIM ( http://www . ncbi . nlm . nih . gov/omim ) accession numbers for genes mentioned in this article are 104750 for SAA1 , 607521 for HPS5 , 189972 for GTF2H1 , 150000 for LDHA , 150150 for LDHC , 601007 for LEPR , and 606051 for SERGEF . | An elevated level of acute-phase serum amyloid A ( A-SAA ) , a sensitive marker of the acute inflammatory state with high heritability estimates , causes amyloidosis and is a risk factor for atherosclerosis and its clinical complications , type 2 diabetes , as well as various malignancies . This study describes the first genome-wide association study on baseline A-SAA concentrations . In a meta-analysis of four genome-wide scans totalling 4 , 212 participants of European descent , we identified two novel genetic susceptibility regions on chromosomes 11 and 1 to be associated with baseline A-SAA concentrations . The chromosome 11 region contains the serum amyloid A1 gene and the adjacent genes and explains a high percentage of the total estimated heritability . The chromosome 1 region is a known genetic susceptibility region for inflammation . Taken together , we identified one region , which seems to be of key importance in the regulation of A-SAA levels and represents a novel potential target for the investigation of related clinical entities . In addition , our findings indicate a close interplay between A-SAA and other inflammatory proteins , as well as a larger role of a known genetic susceptibility region for inflammatory processes as it has been assumed in the past . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"genetics",
"and",
"genomics/gene",
"discovery",
"genetics",
"and",
"genomics/genetics",
"of",
"the",
"immune",
"system"
] | 2010 | Genome-Wide Association Study Identifies Two Novel Regions at 11p15.5-p13 and 1p31 with Major Impact on Acute-Phase Serum Amyloid A |
For ethical and logistical reasons , population-genetic studies of parasites often rely on the non-invasive sampling of offspring shed from their definitive hosts . However , if the sampled offspring are naturally derived from a small number of parents , then the strong family structure can result in biased population-level estimates of genetic parameters , particularly if reproductive output is skewed . Here , we document and correct for the strong family structure present within schistosome offspring ( miracidia ) that were collected non-invasively from humans in western Kenya . By genotyping 2 , 424 miracidia from 12 patients at 12 microsatellite loci and using a sibship clustering program , we found that the samples contained large numbers of siblings . Furthermore , reproductive success of the breeding schistosomes was skewed , creating differential representation of each family in the offspring pool . After removing the family structure with an iterative jacknifing procedure , we demonstrated that the presence of relatives led to inflated estimates of genetic differentiation and linkage disequilibrium , and downwardly-biased estimates of inbreeding coefficients ( FIS ) . For example , correcting for family structure yielded estimates of FST among patients that were 27 times lower than estimates from the uncorrected samples . These biased estimates would cause one to draw false conclusions regarding these parameters in the adult population . We also found from our analyses that estimates of the number of full sibling families and other genetic parameters of samples of miracidia were highly intercorrelated but are not correlated with estimates of worm burden obtained via egg counting ( Kato-Katz ) . Whether genetic methods or the traditional Kato-Katz estimator provide a better estimate of actual number of adult worms remains to be seen . This study illustrates that family structure must be explicitly accounted for when using offspring samples to estimate the genetic parameters of adult parasite populations .
Infectious disease research is rapidly adopting the tools of evolutionary biology and molecular ecology [1]–[5] . Molecular genetic data , evolutionary theory , and population genetic tools can provide methodology to uncover epidemiological processes that cannot be easily determined otherwise . Such processes include pathogen migration and gene-flow , strain divergence , and selection [6]–[9] . However , some pathogens can be difficult subjects for molecular studies because their adult stages cannot be ethically or pragmatically collected from their human hosts . Thus , researchers often rely on the collection of progeny to infer information about the adult population . Schistosome parasites are one such example . Schistosomes are dioecious blood flukes that become reproductively mature in the vasculature ( mesenteric veins or the veins of the bladder plexus ) of their hosts where they reside in primarily monogamous pairs [10] , [11] . The adults are inaccessible , but their offspring can be collected as eggs that are shed in urine or feces . Consequently , schistosome offspring are often used as a proxy for the adult population , typically to infer worm burdens , and genetic structure among host individuals , host species , and geographic locations [12]–[20] . One challenge associated with using samples of offspring to assay genetic structure is that a sample of offspring may be misrepresentative of the adult population , and can thus give biased estimates of parameters of the adult population [9] , [21] , [22] . Two types of biased parameter estimates can result when offspring are produced by a small effective number of breeders , Nb ( either because few adults were breeding and/or because those that did breed had highly skewed reproductive output ) . First , the sampling variance in population allele frequencies that arises from sampling the offspring of a small effective number of adult breeders will yield inflated estimates of genetic differentiation among hosts . Second , the strong sibling structure that will exist in a large sample of offspring having a small Nb will cause negative deviations from Hardy-Weinberg equilibrium ( i . e . downwardly biased estimates of FIS ) [e . g . 23] , [24] , [25] and inflated estimates of linkage disequilibrium ( LD ) among loci within hosts . The likelihood that these sampling artifacts will arise when sampling offspring depends on the number of breeding adults per host , the reproductive skew among those adults , and on the sample size of offspring collected . Furthermore , because it is relatively easy to collect large numbers of offspring , one can reach false conclusions with high statistical confidence . The distribution of reproductive output among individuals in natural populations of organisms is usually highly skewed , causing the ratio of the variance of reproductive output to the mean reproductive output , ( variance to mean ratio , VMR ) , of the adults in a population to be much greater than 1 [e . g . Table 10 . 2 in 26] . In laboratory infections of mice , schistosomes showed VMR ranging from 7 . 2 to 7 . 4 [10] . This reproductive skew produced a ratio of effective number of breeders ( Nb ) to census number of breeding adults ( Nc ) of 0 . 24 . VMR has not previously been measured in natural populations of schistosomes . Therefore , in order to accurately measure important epidemiological parameters , it is essential to determine how large of an effect the above sampling issue will have on population genetic studies of this parasite . Schistosomes are a substantial public health issue in tropical and developing countries . They are estimated to infect over 200 million people ( approximately 1 out of 35 ) worldwide [27] . Schistosomiasis is a chronic and debilitating disease with a life cycle that is difficult to control . Long-lived and recurrent infections present an ongoing inflammatory challenge that can result in anemia , severe portal hypertension , malnutrition , poor growth , impaired cognitive development , increased suseptibility to coinfection , and increased pathology in coinfection [28] . The schistosome life cycle involves a snail intermediate host and a mammalian definitive host . Eggs are released with the urine or feces of the mammalian host , hatch in water , and release free-swimming miracidia . Miracidia infect snails and undergo asexual reproduction resulting in thousands of clonal cercariae that emerge from the snail daily . Cercariae penetrate the skin of their definitive host and establish long lived infections averaging 6–11 years [29] . Estimating population genetic parameters such as F-statistics of adult schistosome populations is important because they can reveal local transmission patterns and the distribution of genetic variation within and among hosts and geographic regions . Genetic data might also be useful in providing measurements of worm burdens and their effective population size , parameters that can be difficult to measure otherwise [30] , [31] . Accurate estimates of these population genetic parameters are sorely needed for effectively targeting drug treatment efforts against schistosomes [32] and to ameliorate reduced drug susceptibility in schistosome populations , which has already been detected in a natural population [33] . The primary aim of this study was to investigate whether sampling artifacts are likely to influence population genetic studies of schistosomes from humans when offspring are sampled in lieu of adults . We collected data from 2 , 424 miracidia noninvasively sampled from 12 humans in western Kenya . To determine the amount of reproductive skew and family structure naturally present in samples collected from humans , we used microsatellite genotype data to cluster offspring into putative sibships . We investigated how family structure influenced LD , inbreeding coefficients ( FIS ) within hosts , and genetic differentiation ( FST and GST ) among hosts , and we developed a correction procedure to remove the bias introduced by family structure . We also investigated whether a single fecal sample taken from a human host would give an adequate representation of the genetic composition of worms in that host , or whether multiple fecal samples should be obtained from a host over several days .
Throughout the paper we use “infrapopulation” to refer to the adult worms in a host ( patient ) , and “component population” to refer to all the adult worms in all the hosts of a host population [34] . The term “sample” refers to a sample of miracidia from a host ( i . e . the offspring sampled from an infrapopulation ) . Initially , we measured population genetic parameters of adult and offspring schistosomes that were collected from mice as part of a prior study [10] . Because we detected the predicted biases in these samples from which both offspring and adult populations could be assayed ( See Online Supporting Information , S1 ) , we collected data from humans naturally infected with schistosomes so that we could determine if these sampling artifacts are relevant to samples from humans residing in a natural transmission zone . We obtained miracidia from human fecal samples from twelve participants enrolled in a longitudinal study [35]–[38] . As these samples were considered discarded medical waste they were viewed as “exempt” by the University of New Mexico Internal Review Board . As part of the longitudinal study , patients were monitored periodically , and if infected , were treated with praziquantel . Patients were adult males who work in Lake Victoria near Kisumu , Kenya , and were either car washers or sand harvesters . Car washers stand knee to ankle deep in the lake as they wash vehicles near the edge of the lake . Sand harvesters stand up to chest-deep , shoveling sand from the bottom of the lake into boats to sell to concrete manufacturers . Both groups of men were exposed to schistosome cercariae as they worked . These patients presumably were exposed to the same pool of cercariae in the lake ( spatially and temporally ) , such that we expect no spatial or temporal genetic subdivision among worm populations from different patients . In support of that expectation , no spatial genetic subdivision was found when cercariae were sampled from snails in this same region , nor was there evidence for LD or deviations from HWE in those cercarial samples [39] . Note that those cercarial samples were scored using the same microsatellite loci as used in this study [39] . To obtain miracidia , eggs were hatched using standard protocols [18] . The miracidia were lysed individually in the wells of a 96 well plate and genotyped at 12 microsatellite loci as described by Steinauer et al . [40] . GenBank accession numbers of the loci include the following: AF325695 , AF202965 , AF202966 , AF202968 , L46951 , AF325698 , AF325694 ( Multiplex panel P17 ) and M85305 , R95529 , AI395184 , BF936409 , AI067617 ( multiplex panel P22 ) . Only those individuals having data for at least 10 of the 12 loci were included in the analysis . We used sibship analyses to determine whether family structure was present in schistosomes collected from naturally infected human hosts . COLONY v . 2 . 0 [41] , [42] was used to partition individual miracidia into their probable sibling groups . COLONY implements a maximum likelihood approach that can incorporate genotyping errors to identify full-sibling and half-sibling families [42] . Using COLONY , we performed analyses with two different user-defined options , first using a monogamous mating system , and second using a polygamous mating system ( to include half sibships ) . Because there was no empirical support for the presence of half sibling families ( see below ) and designating a polygamous mating system resulted in suboptimal full sibling family partitions , we used only the COLONY results where monogamous mating was specified ( see online supplement S2 ) . Nevertheless , the overall results were very similar when using the samples generated with a polygamous mating system ( see Online Supporting Information , S2 , for details on the performance of COLONY on these samples ) . Analyses with COLONY were run with full likelihood , with no priors or known allele frequencies , and one short run per dataset . Our sibling partitions from COLONY were similar to those derived from alternative software packages ( see online supplement S3 ) . The results from COLONY were first used to calculate the number of families occurring in each offspring sample ( FSF = “full sib families” ) . This number included families with a sibship size of 1 . Because the total number of families identified in a sample should increase with the number of sampled miracidia , n , we also calculated FSF/n in order to compare among samples . This number gives a relative estimate of infrapopulation size . We employed a commonly used metric , the variance to mean ratio ( VMR ) for reproductive success , to quantify the amount of family structure in each sample of offspring . Ratios greater than one indicate a skewed distribution in reproductive output or a large variance in family sizes . It is important to point out that this metric will be downwardly biased when the sample size is much smaller than the true number of breeders , thus it is a minimum estimate . We also estimated the effective number of breeders [Nb; 43] within each patient as another relative measure of infrapopulation size using the Linkage Disequilibrium ( LDNE ) and Sibship Assignment ( SA ) [44] methods . The former was calculated using LDNE [45] , and the latter with COLONY v . 2 . To correct both empirical and simulated samples for family structure , we randomly sampled one individual per family to create a reduced data set that no longer contained any full-sib individuals . We have named this the “one-per-family” approach . Because there is some inherent stochasticity associated with randomly selecting a single individual from each family , we used custom scripts to automate the process , which allowed us to create a large number of “one-per-family” samples for calculations of genetic parameters . By creating a large number of samples , we can effectively sample all single individuals from a given family and capture the associated mean and variance ( See online supplement S4 for a discussion of the variance in these samples ) . We predicted that small Nb and large family structure within samples of miracidia would increase LD among loci , cause negative deviations from HWE ( more negative FIS ) , and inflate FST among hosts [9] . Therefore , these parameters were compared between the raw , uncorrected samples and the corrected samples from humans . To account for the reduction in sample size in the corrected samples after removing full siblings , we calculated Weir and Cockerham's estimation of FST ( theta ) , which is unbiased with respect to sample size [46] . We also calculated standardized FST ( FST of the sample relative to the maximum FST value possible given the dataset ) [47] using RecodeData v . 0 . 1 [48] . To calculate pair-wise theta , we used the Geneclust package in R [49] , [50] . We automated the process to iteratively ( 1 ) create a one-per-family dataset for each patient and ( 2 ) calculate all pair-wise FST values between patients . This process was repeated 1000 times , after which we calculated the mean and 95% CIs for the one-per-family FST values and compared them to the uncorrected data set . We repeated the above procedure to calculate a global ( as opposed to pair-wise ) value of Weir and Cockerham's FST using the R package Hierfstat [51] . We also calculated the corrected standardized GST ( G″ST ) [52] using GENODIVE [53] . GST is often used as an analog of FST because FST is dependent on within sample diversity . For assessment of our correction method , G″ST was also calculated in 10 randomly generated corrected datasets for comparison and tested for significance in each dataset using 10 , 000 permutations of the data . To calculate within-patient FIS for uncorrected ( raw ) and 1000 one-per-family samples , we used R scripts to combine observed and expected heterozygosities using the standard equation FIS = ( He-Ho ) /He [54] . We next exported 50 one-per-family data sets from R and imported them into GENEPOP to test for genotypic disequilibrium ( a proxy for LD ) . For each patient we used 1000 batches and 10000 iterations per batch to calculate the percentage of loci pairs that were significantly associated with one another ( p≤0 . 05 , averaged over the 50 one-per-family data sets ) . Because statistical tests can be affected by sample size , we repeated the above procedure on “downsampled” empirical data sets , where we reduced the sample size of miracidia from each patient to the equivalent sample size in the one-per-family data sets from that patient . Unlike in the one-per-family data sets , in the downsampled datasets we removed individuals randomly without respect to the predetermined family structure . This process allowed for equitable comparisons of LD between the one-per-family data sets and the uncorrected ( empirical ) data sets with the same sample size . To further validate the one-per-family approach to correcting for family structure , we also created simulated data sets where we could precisely control which individuals belonged to given families . We used a distribution of family sizes that matched the empirical distribution from the COLONY output to recreate the observed family structure . To create simulated adult schistosomes , we used the empirical allele frequencies to generate multilocus male and female genotypes in accordance with HWE . We next paired adults monogamously and randomly selected one allele from each parent to create offspring in accordance with Mendelian expectation . We generated 1000 offspring per adult pair and then randomly sampled a precise number of offspring per full-sib family in accordance with the COLONY output from the empirical data for family size . For example , if a patient had two families of sizes 5 and 30 , then 5 full-sib offspring would be sampled from one pair , 30 full-sib offspring would be sampled from the next pair , and the remaining offspring would be discarded . Each offspring was assigned a unique individual and family ID to validate the downstream analyses . All simulated data sets were constructed with a script written in R 2 . 15 . 1 [55] . A fully annotated version of this script is freely available at dryad . Using the exact approach as described for the empirical data , we tested the effect of correcting samples on 1000 simulated datasets for measuring FST , FIS , and LD . Notice that with these simulated data sets we knew the family structure with 100% certainty , whereas with the empirical data we assumed that COLONY accurately captured all of the family structure . Thus , these simulated data sets allowed us to accurately verify the utility of our approach to correcting for family structure . As even further proof of principle , we also conducted analyses with Kenyan schistosomes in a mouse model system [two mice infected with field-collected schistosomes; see 10 for laboratory methods] . Because we could sample both the adults and the offspring ( which is not logistically feasible with human patients ) the samples of offspring could be directly compared to the samples of adults for measurements of linkage disequilibrium , FST , HWE , and parentage analysis ( See Online Supporting Information , S1 , for details ) . If reproductive rates are constant , then samples from a single patient should be genetically homogenous over multiple days . However , if reproduction occurs in “bursts” , then sampling a single fecal sample could miss diversity within a patient . For three patients 2–3 fecal samples were collected 1–2 days apart . To determine if these samples differed in genetic composition , we first calculated pairwise FST between temporal samples from the same patient using the raw data and tested their significance via 10 , 000 permutations of genotypes among samples using FSTAT 2 . 9 . 3 [56] . Standardized FST ( F′ST ) was calculated and standardized GST ( G″ST ) was calculated and tested for significance with 1000 permutations using GenAlEx v . 6 . 5 [57] , [58] . Next , for the patients for which the raw data indicated significant temporal differences , we corrected them for family structure using our one-per-family approach . Families were resampled 1000 times and mean pairwise FST was calculated among sample days within each of the three patients . Because the reduced sample size should influence significance testing , we performed the same tests on downsampled samples to match the sample size of the corrected samples . Another application of the calculation of the number of full sibling families in a sample of miracidia is to estimate the minimum worm burden within a patient ( i . e . minimum number of breeding pairs ) . However , it is possible that with adequate sampling , this measurement , as well as other genetic parameters , may serve as relative or even absolute measurements of worm burden . The number of families detected , as well as other genetic parameters such as allelic diversity , should be strongly correlated with the number of worms that were reproductively active when the sample was taken . With this aim , we investigated the relationship of several parameters obtained from our genetic data and compared them to the World Health Organization gold-standard estimator of worm burden , the number of eggs per gram of feces as determined by the Kato Katz method [59] . The genetic parameters we measured included: the number of full sibling families ( FSF ) , a standardized measure of full sibling families ( FSF/n ) , allelic richness ( number of alleles rarefied to the smallest sample size ( AR ) , and the effective number of breeders [Nb; 43] . Nb was estimated using the sibling assignment method implemented in COLONY [44] and the linkage disequilibrium method implemented in LDNe [45] . Although the actual worm burdens in our patients cannot be determined , it is worth determining the extent to which the genetic estimators are intercorrelated and capture the same information as the Kato Katz method ( eggs per gram of feces ) . Pearson's correlations between egg count ( i . e . Kato Katz ) , and the genetic parameters: FSF/n , AR , and Nb were performed using Graphpad Prism v . 5 . 01 ( GraphPad Software , Inc . ) . To further determine similarity among the estimators , a multivariate principal components analysis ( PCA ) using egg output ( i . e . , Kato Katz ) , FSF/n , AR , and both measures of Nb as variables was performed using Systat 11 ( Systat Software , Inc . ) .
The samples of miracidia from humans varied substantially in the amount of family structure they contained ( Table 1 , Figure 1 ) . The percentage of miracidia that belonged to a family of four or more ( i . e . , the number of miracidia which belong to a family with robust support ) ranged from 18–91% among samples , and the variance to mean ratio ( VMR ) for the number of offspring per family ranged from 0 . 37 to 7 . 9 . The lower values of VMR likely reflect our inability to detect a skew in samples with very large populations rather than equal reproductive output among families . When infrapopulations are large , one needs very large sample sizes to accurately describe the family-size distribution ( else you wind up with mostly unrelated individuals ) . Indeed , several lines of evidence in our data indicate that samples with measured VMR less than one are likely those in which the sample size is much smaller than the actual number of families . First , the ratio of the estimated Nb to the sample size was high in samples with VMR<1 ( Table 1 , Figure 1 ) . Furthermore , the percentage of individuals belonging to a family of 4 or more was highly correlated with VMR ( Pearson's r = 0 . 807 , P = 0 . 0008 ) indicating that the samples with a low VMR had a large number of families rather than having a small number of families of equal size . Thus , the measured reproductive skew of samples will be highly dependent on the infrapopulation size and the sample size . In the analyses of samples of miracidia from humans in which the mating system was designated as “polygamous” , many half sibships were inferred in the COLONY partitions . However , most of these included only 2 or 3 members , or involved a large full sibling group with one or two half siblings . Such small groupings are likely to be spurious . Only three patients had half sibling families that consisted of greater than three individuals per FSF ( two such families per patient ) . Furthermore , analysis of our simulated samples revealed that the majority of half-sib assignments were incorrect ( see Online Supporting Information , S2 ) . Thus , we do not see strong evidence for a large number of half sibling groups in our samples from humans , which indicates that these patients were not getting infected with large numbers of genetic clones derived from a single snail . The global FST for the uncorrected data set was 27 . 8 times higher than the corrected , one-per-family data sets ( Corrected dataset global FST = 0 . 00026; Uncorrected dataset FST = 0 . 0074; Uncorrected dataset standardized global F′ST = 0 . 027 ) . G″ST indicated significant population subdivision of the uncorrected dataset ( G″ST = 0 . 031 , P = 0 . 001 ) , and was greatly reduced in the 10 corrected datasets and permutation tests indicated no significant population subdivision ( mean G″ST = −0 . 0013 , range −0 . 004 to 0 ) . Thus , correction of the dataset reduced standardized G″ST from 0 . 031 to a mean of −0 . 0013; and statistical significance was lost with this correction . Furthermore , pairwise FST between patients was greatly reduced in the one-per-family data sets ( Figure 2 ) . In fact , for the simulated datasets the mean pair-wise FST was reduced to 0 . As expected , this analysis revealed that Weir and Cockerham's FST estimator is unbiased , though some comparisons differed from 0 due to decreased precision . The analysis using the empirical data set yielded a nearly identical result , with most of the genetic differentiation being removed from the corrected data sets . Interestingly , the mean pairwise FST for the empirical data set was slightly greater than 0 ( approximately 0 . 001 ) , suggesting a very low level of residual real or artifactual differentiation ( see Discussion ) . The one-per-family correction worked well for both the empirical and simulated data and the confidence intervals surrounding the means from the 1000 iterations were narrow ( smaller than the points on the plots ) . Also , it is important to note that simply downsampling the data randomly ( without regard to family structure ) to match the sample sizes of the corrected data yielded a mean pairwise FST value ( from 5000 resampled datasets ) that was very similar to the value for the uncorrected data ( FST = 0 . 009 ) . Thus , the change in FST value was due to the removal of family structure . As predicted , samples with a large amount of family structure also yielded lower estimates of FIS as indicated by the negative correlation between FIS and VMR ( r = −0 . 638 , P = 0 . 0128 ) ( Figure 3A ) . This relationship was not detected in the corrected samples ( r = −0 . 3797 , P = 0 . 112 ) . The increase in FIS between raw and corrected samples was greatest in the samples having most family structure and these values were positively correlated ( r = 0 . 733 , P = 0 . 003 ) ( Figure 3B ) . All of the uncorrected samples of miracidia from humans showed significant genotypic disequilibrium , ranging from 38 to 91% of pairwise comparisons of loci ( Figure 4 ) . This percentage of loci was much lower in the corrected samples ranging from 0 . 015 to 0 . 065% , suggesting that most of the linkage disequilibrium was due to family structure . Correction also reduced the number of loci in LD when compared to the raw samples that were downsampled to match the same sample size as the corrected samples ( Fig . 4 ) . Correction also removed the relationship between VMR and the percentage of loci in disequilibrium . The same sampling artifacts were detected in the samples from mice ( See Online Supporting Information , S1 , for details and methods ) . FST between the adult schistosome infrapopulations in the two mice was low and not statistically significantly different from zero ( −0 . 026 [95% CI: −0 . 032–0 . 01]; P = 0 . 998 ) . In contrast , FST was substantially higher ( 0 . 021 [95% CI: 0 . 014–0 . 029] and statistically significant between the samples of miracidia collected from each of the mice ( P = 0 . 0001 ) . Although , neither adult nor offspring samples deviated significantly from HWE , point estimates of FIS were more negative in the samples of miracidia , consistent with predictions . These samples of miracidia also showed many pairs of loci in significant linkage disequilibrium , while the adult samples showed no loci in significant disequilibrium . When the temporal fecal samples from the same patient were analyzed in their raw form ( no correction for sibships ) , allele frequencies of miracidia differed significantly among fecal samples collected on different days for two of the three patients ( patient 2: F′ST = 0 . 021 , P = 0 . 0009 , G″ST = 0 . 021 , P = 0 . 003; patient 3: F′ST = 0 . 001 , P = 0 . 333 , G″ST = 0 , P = 0 . 513; patient 12: F′ST = 0 . 020 , P = 0 . 0001 , G″ST = 0 . 018 , P = 0 . 002 ) . We sampled patient 12 on three days . Pairwise tests indicated that the fecal sample collected from day 3 was significantly different from those on day 1 and day 2 , but samples from day 1 and 2 were not significantly different ( Day 1&2: G″ST = 0 . 006 , P = 0 . 173; Day 1&3: G″ST = 0 . 028 , P = 0 . 001; Day 2&3: G″ST = 0 . 019 , P = 0 . 003 ) . We corrected the samples using the one-per-family method only for those two patients for which the raw data showed significant differentiation from the pairwise FST tests . Correcting the samples reduced global FST in all the patient samples and the amount of correction was the highest in the patient with the highest VMR ( #12 ) ( patient 2: raw/corrected FST = 0 . 006/0 . 003; patient 12: 0 . 006/0 . 001 ) . After correction , there were no significant temporal differences between fecal samples . However , the differences between the uncorrected samples also became non-significant when they were downsampled to match the sample sizes of the corrected samples . Egg output ( i . e . , Kato Katz method ) was not correlated with either the number of full sibling families FSF/n ( r = −0 . 097 , P = 0 . 383 ) , with AR ( r = 0 . 153 , P = 0 . 318 ) , or either measure of Nb ( SA: r = −0 . 049 , P = 0 . 440; LDNE: r = −0 . 169 , P = 0 . 300 ) . However , PCA analysis indicated that all the genetic factors were strongly related and loaded heavily on one factor ( loading values = 0 . 90–0 . 96 ) , with egg counts loading heavily on a second factor ( 0 . 99 ) . Factor one ( genetic measures ) explained 69 . 1% of the variation and factor two ( egg counts ) explained 22 . 2% of the variation . Further examination of univariate correlations indicated that all the genetic measures were strongly correlated ( Pearson's r = 0 . 79–0 . 83 and P≤0 . 001 ) ( Fig . 5 ) . The strong intercorrelations among the genetic parameters suggests they are all capturing the same information about variation among patients in number of families contributing to each sample . However , that variation among patients is uncorrelated with variation in the traditional , Kato-Katz estimate of worm burden .
We observed a statistically significant FST>0 between the raw data from two sets of temporal samples that were each collected from a patient over multiple days . Tests using only the raw data would cause one to conclude that genetically different subsets of worms are producing eggs on different days . Family structure and differential representation of families in sequential fecal samples appear to be driving the statistical differences in our data . It should be noted that these biases can occur in samples with both large and small family structure ( i . e . patient 12 and patient 2 , see online supplement , S5 ) . The differentiation between samples may be due to random sampling error ( some families are missed by chance ) , or biological attributes that change a worm pair's contribution to a single fecal sample ( i . e . location in host , age , competition among pairs ) . To answer this question , much larger samples are necessary . In any case , fecal samples from multiple days may be more representative than a single fecal sample . However , it still remains unclear whether sampling artifacts can be best overcome by increasing the sample size on one day , or collecting several small samples over multiple days . The number of full sibling families detected in a sample of miracidia is a measure of the minimum number of worm pairs present in a patient . The strong covariation among genetic variables in our data suggests that genetic parameters could be used further: to depict relative worm burdens in patients . It is also possible that , given a large enough sample size , the true worm burden within a patient could be detected via the number of full sibling families . The challenge will be obtaining a large enough sample size to account for the true worm burden , a parameter that unfortunately may not be known without obtaining genetic data first . As shown in Table 1 , the sample size necessary to detect most sibling families varies widely among patients . For example , for Patient 6 , 75% of the miracidia collected were partitioned into robust full sibling families ( >3 ) with a sample size of 81 miracidia . In contrast , with 412 sampled miracidia , only 49% were partitioned into robust families for Patient 3 . However , it may be possible to obtain accurate worm burden estimates even without exhaustive sampling , by fitting the observed sibship sizes to an expected sibship size distribution to predict the number of unsampled sibships . The lack of correlation among fecal egg counts ( i . e . , Kato Katz ) and our genetic proxies for worm burden is interesting , particularly considering the strong correlations detected with genetic measures even with a low sample size of 12 patients . The Kato Katz methodology for egg enumeration is known to be highly variable between fecal samples collected from the same patient and has been deemed unreliable by some particularly when worm burdens are low [67]–[69] . However , others have found evidence of reliability of this method across a broad range of infection intensities [e . g . 70] , [71] . It may be that the infection levels in our study were not broad enough for the Kato Katz method to accurately detect relative differences in the worm burdens of our patients . Although we have no independent data to determine which approach gives the most accurate estimates of the true number of adult breeding worms , we suggest that genetic methods potentially give more reliable estimates of the relative number of adult breeding worms per host . These methods should be explored further because they could be valuable tools for epidemiological studies that measure the success of control programs . A previous study did not find a relationship between their calculated Ne from schistosome offspring populations and allelic richness [19] . However , it appears that their estimates of Ne may not have been very accurate because there was no correlation between Ne values from the same samples calculated by two techniques and they were estimated with large confidence intervals . This lack of accuracy could be due to the small number of markers used , relatively small sample sizes ( 10–30 miracidia per patient ) , and pooling of samples from several infrapopulations rather than using a subdivided breeders model [31] to calculate Ne . The lack of correlation may also be due to saturation of allelic richness since there is a limit to the number of alleles that will be found in a population and a correlation beyond this saturation point is not expected . Genetic epidemiology is a powerful tool for infectious disease research . However , in cases where offspring must be collected in lieu of adults , data analysis and interpretation should be carefully considered . We have shown that samples of parasite larvae collected from humans can contain significant family structure , which can lead to inflated estimates of linkage disequilibrium and FST , and underestimates of FIS . The amount of bias in each of these parameters is positively correlated with the skew in reproductive output of individuals . It should also be noted that sibling structure and skewed reproductive output among individuals ( small Nb ) could skew additional population genetic parameters and analyses not evaluated here such as observed heterozygosity ( Ho ) , gene diversity ( He ) [72] , genetic distance , and clustering algorithms , thus , care should be taken in their interpretation . Correcting samples by performing a sibship analysis and then excluding all but one member of each full sibling family is effective at removing or reducing this bias . The number of full sibling families detected by our analyses gives an estimate of the minimum number of worm pairs within a patient and may be a reliable estimator of the relative worm burdens within patients , an important epidemiological parameter . Microsatellite genotype data and annotated R scripts to perform the “leave-one-out” procedure will be made available at Dryad . | Genetic epidemiology uses genetic data to uncover patterns of disease processes . To acquire data for these analyses , individual pathogens are collected and scored at genetic markers , and the resultant data are analyzed to infer biological patterns about the pathogen populations . In lieu of invasive sampling of adult pathogens in humans , researchers have relied on non-invasive sampling of parasite offspring ( often shed in fecal samples ) . One potential problem with this approach is that analyses using the offspring data will be biased because many of the offspring are related and family sizes are likely to be unequal . We show that this sampling issue is relevant in a natural transmission zone in western Kenya and that it yields biases in three important parameters: genetic differentiation , inbreeding coefficients , and estimates of the amount of non-random association between loci ( linkage disequilibrium ) . We also develop a method to remove these biases by removing the sibling structure present in the dataset . Finally , we suggest that our measure of family number , as well as other genetic measures , may be useful measures of the worm burdens in patients . | [
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] | [] | 2013 | Non-Invasive Sampling of Schistosomes from Humans Requires Correcting for Family Structure |
A major obstacle to understanding neural coding and computation is the fact that experimental recordings typically sample only a small fraction of the neurons in a circuit . Measured neural properties are skewed by interactions between recorded neurons and the “hidden” portion of the network . To properly interpret neural data and determine how biological structure gives rise to neural circuit function , we thus need a better understanding of the relationships between measured effective neural properties and the true underlying physiological properties . Here , we focus on how the effective spatiotemporal dynamics of the synaptic interactions between neurons are reshaped by coupling to unobserved neurons . We find that the effective interactions from a pre-synaptic neuron r′ to a post-synaptic neuron r can be decomposed into a sum of the true interaction from r′ to r plus corrections from every directed path from r′ to r through unobserved neurons . Importantly , the resulting formula reveals when the hidden units have—or do not have—major effects on reshaping the interactions among observed neurons . As a particular example of interest , we derive a formula for the impact of hidden units in random networks with “strong” coupling—connection weights that scale with 1 / N , where N is the network size , precisely the scaling observed in recent experiments . With this quantitative relationship between measured and true interactions , we can study how network properties shape effective interactions , which properties are relevant for neural computations , and how to manipulate effective interactions .
Establishing relationships between a network’s architecture and its function is a fundamental problem in neuroscience and network science in general . Not only is the architecture of a neural circuit intimately related to its function , but pathologies in wiring between neurons are believed to contribute significantly to circuit dysfunction [1–15] . A major obstacle to uncovering structure-function relationships is the fact that most experiments can only directly observe small fractions of an active network . State-of-the-art methods for determining connections between neurons in living networks infer them by fitting statistical models to neural spiking data [16–25] . However , the fact that we cannot observe all neurons in a network means that the statistically inferred connections are “effective” connections , representing some dynamical relationship between the activity of nodes but not necessarily a true physical connection [24–33] . Intuitively , reverberations through the network must contribute to these effective interactions; our goal in this work is to formalize this intuition and establish a quantitative relationship between measured effective interactions and the true synaptic interactions between neurons . With such a relationship in hand we can study the effective interactions generated by different choices of synaptic properties and circuit architectures , allowing us to not only improve interpretation of experimental measurements but also probe how circuit structure is tied to function . The intuitive relationship between measured and effective interactions is demonstrated schematically in Fig 1 . Fig 1A demonstrates that in a fully-sampled network the directed interactions between neurons—here , the change in membrane potential of the post-synaptic neuron after it receives a spike from the pre-synaptic neuron—can be measured directly , as observation of the complete population means different inputs to a neuron are not conflated . However , as shown in Fig 1B , the vastly more realistic scenario is that the recorded neurons are part of a larger network in which many neurons are unobserved or “hidden . ” The response of the post-synaptic neuron 2 to a spike from pre-synaptic neuron 1 is a combination of both the direct response to neuron 1’s input as well as input from the hidden network driven by neuron 1’s spiking . Thus , the measured membrane response of neuron 2 due to a spike fired by neuron 1—which we term the “effective interaction” from neuron 1 to 2—may be quite different from the true interaction . It is well-known that circuit connections between recorded neurons , as drawn in Fig 1C , are at best effective circuits that encapsulate the effects of unobserved neurons , but are not necessarily indicative of the true circuit architecture . The formalized relationship we will establish in the Results is given in Fig 2 . Even once we establish a relationship between the effective and true connections , we will in general not be able to use measurements of effective interactions to extrapolate back to a unique set of true connections; at best , we may be able to characterize some of the statistical properties of the full network . The obstacle is that several different networks—different both in terms of architecture and intrinsic neural properties—may give rise to the same network behaviors , a theme of much focus in the neuroscience literature [34–39] . That is , inferring the connections and intrinsic neural properties in a full network from activity recordings from a subset of neurons is in general an ill-posed problem , possessing several degenerate solutions . Several statistical inference methods have been constructed to attempt to infer the presence of , and connections to , hidden neurons [28 , 40–42]; the subset of the degenerate solutions that each of these methods finds will depend on the particular assumptions of the inference method ( e . g . , the regularization penalties applied ) . As an example , we demonstrate two small circuit motifs that give rise to nearly identical effective interactions , despite crucial differences between the circuits ( Figs 3 and 4 ) . Understanding the effect of hidden neurons on small circuit motifs is only a piece of the hidden neuron puzzle , and a full understanding necessitates scaling up to large circuits containing many different motifs . Having an analytic relationship between true and effective interactions greatly facilitates such analyses by directly studying the structure of the relationship itself , rather than trying to extract insight indirectly through simulations . In particular , in going to large networks we focus on the degree to which hidden neurons skew measured interactions ( Fig 5 ) , and how we can predict the features of effective interactions we expect to measure when recording from only a subset of neurons in a network with hypothesized true interactions ( Fig 6 ) . Establishing a theoretical relationship between measured and “true” interactions will thus enable us to study how one can alter the network properties to reshape the effective interactions , and will be of immediate importance not only for interpreting experimental measurements of synaptic interactions , but for elucidating their role in neural coding . Moreover , understanding how to shape effective interactions between neurons may yield new avenues for altering , in a principled way , the computations performed by a network , which could have applications for treating neurological diseases caused in part by pathological synaptic interactions .
Our goal is to derive a relationship between the effective synaptic interactions between recorded neurons and the true synaptic interactions that would be obtained if the network were fully observed . This makes explicit how the synaptic interactions between neurons are modified by unobserved neurons in the network , and under what conditions these modifications are—or are not—significant . We derive this result first , using a probabilistic model of network activity in which all properties are known . We then build intuition by applying our result to two simple networks: a 3-neuron feedforward-inhibition circuit in which we are able to qualitatively reproduce measurements by Pouille and Scanziani [43] , and a 4-neuron circuit that demonstrates how degeneracies in hidden networks are handled within our framework . To extend our intuition to larger networks , we then study the effective interactions that would be observed in sparse random networks with N cells and strong synaptic weights that scale as 1 / N [44–47] , as has been recently observed experimentally [48] . We show how unobserved neurons significantly reshape the effective synaptic interactions away from the ground-truth interactions . This is not the case with “classical” synaptic scaling , in which synaptic strengths are inversely proportional to the number of inputs they receive ( assumed O ( N ) ) , as we will also show . ( The case of classical scaling has also been studied previously using a different approach in [49–52] ) . We model the full network of N neurons as a nonlinear Hawkes process [53] , often referred to as a “Generalized linear ( point process ) model” in neuroscience , and broadly used to fit neural activity data [16–23 , 54] . Here we use it as a generative model for network activity , as it approximates common spiking models such as leaky integrate and fire systems driven by noisy inputs [55 , 56] , and is equivalent to current-based leaky integrate-and-fire models with soft-threshold ( stochastic ) spiking dynamics ( see Methods ) . To derive an approximate model for an observed subset of the network , we partition the network into recorded neurons ( labeled by indices r ) and hidden neurons ( labeled by indices h ) . Each recorded neuron has an instantaneous firing rate λr ( t ) such that the probability that the neuron fires within a small time window [t , t + dt] is λr ( t ) dt , when conditioned on the inputs to the neuron . The instantaneous firing rate in our model is λ r ( t ) = λ 0 ϕ ( μ r + ∑ r ′ J r , r ′ * n ˙ r ′ ( t ) + ∑ h J r , h * n ˙ h ( t ) ) , ( 1 ) where λ0 is a characteristic firing rate , ϕ ( x ) is a non-negative , continuous function , μr is a tonic drive that sets the baseline firing rate of the neuron , and J i , j * n ˙ j ( t ) ≡ ∫ - ∞ ∞ d t ′ J i , j ( t - t ′ ) n ˙ j ( t ′ ) is the convolution of the synaptic interaction ( or “spike filter” ) Ji , j ( t ) with spike train n ˙ j ( t ) from pre-synaptic neuron j to post-synaptic neuron i , for neural indices i and j that may be either recorded or hidden . In this work we take the tonic drive to be constant in time , and focus on the steady-state network activity in response to this drive . We consider interactions of the form J i , j ( t ) ≡ J i , j g j ( t ) , where the temporal waveforms gj ( t ) are normalized such that ∫ 0 ∞ d t g j ( t ) = 1 for all neurons j . Because of this normalization , the weight J i , j carries units of time . We include self-couplings Ji , i ( t ) not to represent autapses , but to account for intrinsic neural properties such as refractory periods ( J i , i < 0 ) or burstiness ( J i , i > 0 ) . The firing rates for the hidden neurons follow the same expression with indices h and r interchanged . We seek to describe the dynamics of the recorded neurons entirely in terms of their own set of spiking histories , eliminating the dependence on the activity of the hidden neurons . This demands calculating the effective membrane response of the recorded neurons by averaging over the activity of the hidden neurons conditioned on the activity of the recorded neurons . In practice this is intractable to perform exactly [57–60] . Here , we use a mean field approximation to calculate the mean input from the hidden neurons ( again , conditioned on the activity of the recorded neurons ) . The value of deriving such a relationship analytically , as opposed to simply numerically determining the effective interactions , is that the resulting expression will give us insight into how the effective interactions decompose into contributions of different network features , how tuning particular features shapes the effective interactions , and conditions under which we expect hidden units to skew our measurements of connectivity in large partially observed networks . As shown in detail in the Methods , the instantaneous firing rates of the recorded neurons can then be approximated as λ r ( t ) ≈ λ 0 ϕ ( μ r eff + ∑ r ′ J r , r ′ eff * n ˙ r ′ ( t ) + ξ r ( t ) ) . The effective baselines μ r eff = μ r + ∑ h J r , h ν h , are simply modulated by the net tonic input to the neuron , so we do not focus on them here . The ξr ( t ) are effective noise sources arising from fluctuation input from the hidden network . At the level of our mean field approximation these fluctuations vanish; corrections to the mean field approximation are straightforward and yield non-zero noise correlations , but will not impact our calculation of the effective interactions ( see the Methods and SI ) , so as with the effective baselines we will not focus on the effective noise here . The effective coupling filters are given in the frequency domain by J ^ r , r ′ eff ( ω ) = J ^ r , r ′ ( ω ) + ∑ h , h ′ J ^ r , h ( ω ) Γ ^ h , h ′ ( ω ) J ^ h ′ , r ′ ( ω ) . ( 2 ) These results hold for any pair of recorded neurons r′ and r , and any choice of network parameters for which the mean field steady state of the hidden network exists . Here , the νh are the steady-state mean firing rates of the hidden neurons and Γ ^ h , h ′ ( ω ) is the linear response function of the hidden network to perturbations in the input . That is , Γh , h′ ( t − t′ ) is the linear response of hidden neuron h at time t due to a perturbation to the input of neuron h′ at time t′ , and incorporates the effects of h′ propagating its signal to h through other hidden neurons , as demonstrated graphically in Fig 2 . Both νh and Γ ^ h , h ′ ( ω ) are calculated in the absence of the recorded neurons . In deriving these results , we have neglected both fluctuations around the mean input from the hidden neurons , as well as higher order filtering of the recorded neuron spikes . For details on the derivations and justification of approximations , see the Methods and Supporting Information ( SI ) . The effective coupling filters are what we would—in principle—measure experimentally if we observe only a subset of a network , for example by pairwise recordings shown schematically in Fig 1 . For larger sets of recorded neurons , interactions between neurons are typically inferred using statistical methods , an extremely nontrivial task [16–23 , 28 , 40 , 41] , and details of the fitting procedure could potentially further skew the inferred interactions away from what would be measured by controlled pairwise recordings . We will put aside these complications here , and assume we have access to an inference procedure that allows us to measure J r , r ′ eff ( t ) without error , so that we may focus on their properties and relationship to the ground-truth coupling filters . The ground-truth coupling filters J ^ r , r ′ ( ω ) ( as defined in Eq ( 1 ) ) are modified by a correction term ∑ h , h ′ J ^ r , h ( ω ) Γ ^ h , h ′ ( ω ) J ^ h ′ , r ′ ( ω ) . The linear response function Γ ^ h , h ′ ( ω ) admits a series representation in terms of paths through the network through which neuron r′ is able to send a signal to neuron r via hidden neurons only . We may write down a set of “Feynmanesque” graphical rules for explicitly calculating terms in this series [53] . First , we define the input-output gain of a hidden neuron h , γ h ≡ λ 0 ϕ ′ ( μ h + ∑ h ′ J h , h ′ ν h ′ ) , calculated in the absence of recorded neurons . The contribution of each term can then be written down using the following rules , shown graphically in Fig 2: i ) for the edge connecting recorded neuron r′ to a hidden neuron hi , assign a factor J ^ h i , r ′ ( ω ) ; ii ) for each node corresponding to a hidden neuron hi , assign a factor γ h i / ( 1 - γ h i J ^ h i , h i ( ω ) ) ; iii ) for each edge connecting hidden neurons hi ≠ hj , assign a factor J ^ h j , h i ( ω ) ; and iv ) for the edge connecting hidden neuron hj to recorded neuron r , assign a factor J ^ r , h j ( ω ) . All factors for each path are multiplied together , and all paths are then summed over . The graphical expansion is reminiscent of recent works expanding correlation functions of linear models of network spiking in terms of network “motifs” [61–63] . Computationally , this expression is practical for calculating the effective interactions in small networks involving only a few hidden neurons ( as in the next section ) , but is generally unwieldy for large networks . In practice , for moderately large networks the linear response matrix Γ ^ h , h ′ ( ω ) can be calculated directly by numerical matrix inversion and an inverse Fourier transform back into the time domain . The utility of the path-length series is the intuitive understanding of the origin of contributions to the effective coupling filters and our ability to analytically analyze the strength of contributions from each path . For example , one immediate insight the path decomposition offers is that neurons only develop effective interactions between one another if there is a path by which one neuron can send a signal to the other . Constructing networks that produce particular effective interactions is tractable for small circuits , but much more difficult for larger circuits composed of many circuit motifs . Not only can combinations of different circuit motifs interact in unexpected ways , one must also take care to ensure the resulting network is both active and stable—i . e . , that firing will neither die out nor skyrocket to the maximum rate . Stability in networks is often implemented by either building networks with classical ( or “weak” ) synapses whose strength scales inversely with the number of inputs they receive , assumed here to be proportional to network size , and hence J i , j ∼ 1 / N , or by building balanced networks in which excitatory and inhibitory synaptic strengths balance out , on average , and scale as J i , j ∼ 1 / N [44 , 48] ( but note the distinction that we use a “soft threshold” firing model with nonlinearity that is fixed as N varies , whereas previous work has typically used hard threshold models ) . In both cases the synapses tend to be small in value in large networks , but are compensated for by large numbers of incoming connections . In the case of 1/N scaling , neurons are driven primarily by the mean of their inputs , while in “strong” balanced 1 / N networks neurons are driven primarily by fluctuations in their inputs . Our goal is to understand how the interplay between the presence of hidden neurons and different synaptic scaling or network architectures shapes effective interactions . Previous work has studied the hidden-neuron problem in the weak coupling limit [49–52] using a different approach to relate inferred synaptic parameters to true parameters; here we use our approach to study the 1 / N strong coupling limit , theoretically predicted to be an important feature that supports computations in networks in a balanced regime [44–47] . Moreover , experiments in cultured neural tissue have been found to be more consistent with the 1 / N scaling than 1/N [48] , indicating that it may have intrinsic physiological importance . We analytically determine how significantly effective interaction strengths are skewed away from the true interaction strengths as a function of both the number of observed neurons and typical synaptic strength . We consider several simple networks ubiquitous in neural modeling: first , an Erdős-Réyni ( ER ) network with “mixed synapses” ( i . e . , a neuron may have both positive and negative synaptic weights ) , a balanced ER network with Dale’s law imposed ( a neuron’s synapses are all the same sign ) , and a Watts-Strogatz ( WS ) small world network with mixed synapses . Each network has N neurons and connection sparsity p ( only 100p% of connections are non-zero ) . Connections in ER networks are chosen randomly and independently , while connections in the WS network are determined by randomly rewiring a fraction β of the connections in a ( pN ) th-nearest-neighbor ring network . such that the overall network has a backbone of local synaptic connections with a web of sparse long-range connections . In each network Nrec neurons are recorded randomly . For simplicity we take the baselines of all neurons to be equal , μi = μ0 ( such that in the absence of synaptic input the probability that a neuron fires in a short time window Δt is λ0Δt exp ( μ0 ) ) . We choose the rate nonlinearity to be exponential , ϕ ( x ) = ex; this is the “canonical” choice of nonlinearity often used when fitting this model to data [16–18 , 20 , 67] . We will further assume exp ( μ0 ) ≪ 1 , so that we may use this as a small control parameter . For i ≠ j , the non-zero synaptic weights between neurons J i , j are independently drawn from a normal distribution with zero mean and standard deviation J0/ ( pN ) a , where J0 controls the overall strength of the weights and a = 1 or 1/2 , corresponding to “weak” and “strong” coupling . For simplicity , we do not consider intrinsic self-coupling effects in this part of the analysis , i . e . , we take J i , i = 0 for all neurons i . For the Dale’s law network , the overall distribution of synaptic weights follows the same normal distribution as the mixed synapse networks , but the signs of the weights correspond to whether the pre-synaptic neuron is excitatory or inhibitory . Neurons are randomly chosen to be excitatory and inhibitory , the average number of each type being equal so that the network is balanced . Numerical values of all parameters are given in Table 1 . We seek to assess how the presence of hidden neurons can shape measured network interactions . We first focus on the typical strength of the effective interactions as a function of both the fraction of neurons recorded , f = Nrec/N , and the strength of the synaptic weights J0 . We quantify the strength of the effective interactions by defining the effective synaptic weights J r , r ′ eff ≡ ∫ 0 ∞ d τ J r , r ′ eff ( τ ) = J ^ r , r ′ eff ( ω = 0 ) ; c . f . J r , r ′ = ∫ 0 ∞ d τ J r , r ′ ( τ ) for the true synaptic weights . We then study the sample statistics of the difference , J r , r ′ eff - J r , r ′ , averaged across both subsets of recorded neurons and network instantiations , to estimate the typical contribution of hidden neurons to the measured interactions . The mean of the synaptic weights is near zero ( because the weights are normally distributed with zero mean in the mixed synapse networks and due to balance of excitatory and inhibitory neurons in the Dale’s law network ) , so we focus on the root-mean-square of J r , r ′ eff - J r , r ′ . This measure is a conservative estimate of changes in strength , as J r , r ′ eff ( τ ) may have both positive and negative components that partially cancel when integrated over time , unlike Jr , r′ ( τ ) . An alternative measure we could have chosen that avoids potential cancellations is ∫ 0 ∞ d τ | J r , r ′ eff ( τ ) - J r , r ′ ( τ ) | , i . e . , the integrated absolute difference between effective and true interactions . However , this will depend on our specific choices of waveform g ( τ ) in our definition J i , j ( τ ) = J i , j g ( τ ) , whereas J r , r ′ eff - J r , r ′ does not due to our normalization ∫ 0 ∞ d τ g ( τ ) = 1 . As |∫ dτ f ( τ ) | ≤ ∫ dτ |f ( τ ) | , for any f ( τ ) , we can consider our choice of J r , r ′ eff - J r , r ′ as a lower bound on the strength that would be quantified by ∫ 0 ∞ d τ | J r , r ′ eff ( τ ) - J r , r ′ ( τ ) | . Numerical evaluations of the population statistics for all three network types are shown as solid curves in Fig ( 5 ) , for both strong coupling and weak coupling . All three networks yield qualitatively similar results . The vertical axes measure the root-mean-square deviations between the statistically expected true synaptic J r , r ′ and the corresponding effective synaptic weight J r , r ′ eff , normalized by the true root mean square of J r , r ′ . Thus , a ratio of 0 . 5 corresponds to a 50% root-mean-square difference in effective versus true synaptic strength . We measure these ratios as a function of both the fraction of neurons recorded ( horizontal axis ) and the parameter J0 ( labeled curves ) . There are two striking effects . First , deviations are nearly negligible ( O ( 1 / p N ) ) for 1/N scaling of connections ( gray traces in Fig 5 ) . Thus , for large networks with synapses that scale with the system size , vast numbers of hidden neurons combine to have negligible effect on effective couplings . This is in marked contrast to the case when coupling is strong ( 1 / N scaling ) , when hidden neurons have a pronounced O ( 1 ) impact ( purple traces in Fig 5 ) . This is particularly the case when f ≪ 1—the typical experimental case in which the hidden neurons outnumber observed ones by orders of magnitude—or when J0 ≲ 1 . 0 , when typical deviations become half the magnitude of the true couplings themselves ( upper blue line ) . For J0 ≳ 1 . 0 , the network activity is unstable for an exponential nonlinearity . To gain analytical insight into these numerical results , we calculate the standard deviation σ [ J r , r ′ eff - J r , r ′ ] , normalized by σ [ J r , r ′ ] , for contributions from paths up to length-3 , focusing on the case of the ER network with mixed synapses ( the Dale’s law and WS networks are more complicated , as the moments of the synaptic weights depend on the identity of the neurons ) . For strong 1 / N coupling we find σ [ J r , r ′ eff - J r , r ′ ] σ [ J r , r ′ ] ≈ λ 0 J 0 e μ 0 1 - f × ( 1 + 3 2 ( λ 0 J 0 e μ 0 ) 2 ( 1 - f ) ) , ( 4 ) corresponding to the black dashed curves in Fig 5 left . Eq ( 4 ) is a truncation of a series in powers of λ 0 J 0 e μ 0 1 - f , where f = Nrec/N is the fraction of recorded neurons . The most important feature of this series is the fact that it only depends on the fraction of recorded neurons f , not the absolute number , N . Contributions from long paths remain finite , even as N → ∞ . In contrast , the corresponding expression for σ [ J r , r ′ eff - J r , r ′ ] / σ [ J r , r ′ ] in the case of weak 1/N coupling is a series in powers of λ 0 J 0 e μ 0 ( 1 - f ) / ( p N ) , so that contributions from long paths are negligible in large networks N ≫ 1 . ( See [67] for derivation and results for N = 100 . ) Deviations of Eq ( 4 ) from the numerical solutions in Fig 5 indicate that contributions from truncated terms are not negligible when f ≪ 1 . As these terms correspond to paths of length-4 or more , this shows that long chains through the network contribute significantly to shaping effective interactions . The above analysis demonstrates that the strength of the effective interactions can deviate from that of the true direct interactions by as much as 50% . However , changes in strength do not give us the full picture—we must also investigate how the temporal dynamics of the effective interactions change . To illustrate how hidden units can skew temporal dynamics , in Fig 6 we plot the effective vs . true interactions between Nrec = 3 neurons in an N = 1000 neuron network . Because the three network types considered in Fig 5 yield qualitatively similar results , we focus on the Erdős-Réyni network with mixed synapses . Four of the true interactions between neurons shown in Fig 6 are non-zero ( J 1 , 2 eff ( t ) , J 3 , 2 eff ( t ) , J 3 , 1 eff ( t ) , and J 2 , 3 eff ( t ) ) . Of these , three exhibit only slight differences between the true and effective interactions: J 1 , 2 eff ( t ) and J 3 , 1 eff ( t ) have slightly longer decay timescales than their true counterparts , while J 2 , 3 eff ( t ) has a slightly shorter timescale , indicating the contribution of the hidden network to these interactions was either small or cancelled out . However , the interaction J 3 , 2 eff ( t ) differs significantly from the true interaction , becoming initially excitatory but switching to inhibitory after a short time , as in our earlier example case of feedforward inhibition . This indicates that neuron 2 must drive a cascade of neurons that ultimately provide inhibitory input to neuron 3 . Contrasting the true and effective interactions shown in Fig 6 highlights many of the ways in which hidden neurons skew the temporal properties of measured interactions . An immediately obvious difference is that although the true synaptic connections in the network are sparse , the effective interactions are not . This is a generic feature of the effective interaction matrix , as in order for an effective interaction from a neuron r′ to r to be identically zero there cannot be any paths through the network by which r′ can send a signal to r . 1 In a random network the probability that there are no paths connecting two nodes tends to zero as the network size N grows large . Note that this includes paths by which each neuron can send a signal back to itself , hence the neurons developed effective self-interactions , even though the true self-interactions are zero in these particular simulations .
The issue of degeneracy in complex biological systems and networks has been discussed at length in the literature , in the context of both inherent degeneracies—multiple different network architectures can produce the same qualitative behaviors [34 , 37–39] , as well as degeneracies in our model descriptions—many models may reproduce experimental observations , demanding sometimes arbitrary criteria for selecting one model over another . All have implications for how successfully one can infer unobserved network properties . One kind of model degeneracy , “sloppiness” [35 , 65] , describes models in which the behavior of the model is sensitive to changes in only a relatively small number of directions in parameter space . Many models of biological systems have been shown to be sloppy [35]; this could account for experimentally observed networks that are quite different in composition but produce remarkably similar behaviors . Sloppiness suggests that rather than trying to infer all properties of a hidden network , there may be certain parameter combinations that are much more important to the overall network operation , and could potentially be inferred from subsampled observations . Another perspective on model degeneracy comes from the concepts of “universality” that occur in random matrix theory [68 , 69] and Renormalization Group methods of statistical physics [64] . Many bulk properties of matrices ( e . g . , the distribution of eigenvalues ) whose entrees are combinations random variables , such as our J r , r ′ eff , are universal in that they depend on only a few key details of the distribution that individual elements are drawn from [70] . Similarly , one of the central results of the Renormalization Group shows that models with drastically disparate features may yield the same coarse-grained model structure when many degrees of freedom are averaged out , as in our case of approximately averaging out hidden neurons . Different distributions ( in the case of random matrix theory ) or different models ( in the case of the Renormalization group ) that yield the same bulk properties or coarse-grained models are said to be in the same “universality class . ” Measuring particular quantities under a range of experimental conditions ( e . g . , different stimuli ) may be able to reveal which universality class an experimental system belongs to and eliminate models belonging to other universality classes as candidate generating models of the data , but these measurements cannot distinguish between models within a universality class . Purely feedforward networks and recurrent networks are simple examples of broad universality classes in this context . In any randomly sampled feedforward network , only the feedforward interactions are modified or generated; no lateral or feedback connections develop because there is no path through hidden neurons that a recorded neuron can send signals to recorded neurons in the same or previous layers . Thus , the feedforward structure—a topological property of the network—is preserved . However , adding even a single feedback connection can destroy this topological structure if it joins two neurons connected by a feedforward path—i . e . , such a link creates a cycle within the network , and it is no longer feedforward . If this network is heavily subsampled ( f ≪ 1 ) the resulting effective interactions J r , r ′ eff ( t ) can even be fully recurrent . The majority of the effective interactions may be very weak , but nonetheless from a topological perspective the network has been fundamentally altered . Accordingly , any interactions J r , r ′ eff ( t ) that represent a purely feedforward network could not have come from a network with recurrent interactions . In practice , we expect few , if any , cortical networks to be purely feedforward , so most networks will be recurrent if we consider only the network connections and not the connection strengths . Thus , a more interesting question is how the statistics and dynamics of synaptic weights further partition topologically-defined universality classes; for example , whether the distribution of synaptic weights can split the sets of J r , r ′ eff ( t ) that arise from recurrent networks and predominantly feedforward networks with sparse feedback and lateral interactions into different universality classes . A thorough investigation of such phenomena will be the focus of future work . Despite the many possible confounds network degeneracy produces , much of the work on inference of hidden network properties has focused on inferring the individual interactions between neurons , with varying degrees of success . Both Dunn and Roudi [40] as well as Tyrcha and Hertz [41] studied inference of hidden activity in kinetic Ising models , sometimes used as simple minimal models of neuronal network activity . They found that the synaptic weights between pairs of observed neurons and observed-hidden pairs could be recovered to within reasonably small mean-squared-error when the number of hidden neurons was less than the number of observed neurons . However , both methods also found it difficult to infer connections between pairs of hidden neurons , resorting to setting such connections to zero in order to stabilize their algorithms . Tyrcha and Hertz also note their method recovers only an equivalence class of connections due to degeneracy in the possible assignment of signs of the synaptic weights and hidden neuron labels . This suggests inferring hidden network structure will be nearly impossible in the realistic limit Nrec ≪ Nhid . A series of papers by Bravi and Sollich perform theoretical analyses of hidden dynamics inference in chemical reaction networks , modeled by a system of Langevin equations [57–60] . Although the applications the authors have in mind are signaling pathways such as epidermal growth factor reaction networks , one could imagine re-interpreting or adapting these equations to describe rate models . The authors develop a variety of approaches , including Plefka expansions [57–59] and variational Gaussian approximations [60] , to study how observations constrain the inferred hidden dynamics , assuming particlar properties of the network structure . Ref . [60] in particular takes an approach most similar to ours , deriving an effective system of Langevin equations for the subsampled dynamics of the chemical reaction network . The effective system of equations contains a memory kernel that plays a role analogous to the correction to the interactions between neurons in our work ( second term in our Eq ( 2 ) . However , the structure of the memory kernel in [60] has a rather different form , being exponentially dependent on the integral of the hidden-hidden interactions , in contrast to our Γh , h′ ( t − t′ ) , which depends on the inverse of δh , h′δ ( t − t′ ) − γhJh , h′ ( t − t′ ) ( see Methods ) . Though Bravi and Sollich do not expand their memory kernel in a series as we do , it would admit a similar series and interpretation in terms of paths through the hidden network , as in our Fig 2A . However , due to the exponential dependence on the hidden-hidden interactions , long paths of length ℓ through hidden networks are suppressed by factors ℓ ! , suggesting the hidden network may have less influence in such networks compared to the network dynamics we study here . Closest to our choice of model , Pillow and Latham [28] and Soudry et al[25] both use modifications of nonlinear Hawkes models to fit neural data with unobserved neurons . Pillow and Latham outline a statistical approach for inferring not just interactions with and between hidden neurons , but also the spike trains of hidden neurons , testing the method on a network of two neurons ( one hidden ) . To properly infer the spike train of the hidden neurons , the model must allow for acausal synaptic interactions . This is acceptable if the goal is inferring hidden spike trains: for example , if the hidden neuron were to make a strong excitatory synapse onto the observed neuron , then a spike from the observed neuron increases the probability that the hidden neuron fired a spike in the recent past . An acausal synaptic interaction captures this effect , but is of course an unphysical feature in a mechanistic model , precluding physiological interpretation of such an interaction . Soudry et al are concerned with the fact that common input from hidden neurons will skew estimates of network connectivity . To get around this issue , they present a different take on the hidden unit problem: rather than attempt to infer connectivity in a fixed subsample of a network , they propose a shotgun sampling method , in which a sequence of overlapping random subsets of the network are sampled over a long experiment . Under this procedure , a large fraction of the network can be sampled , just not contiguously in time , and reconstruction of the entire network could in principle be accomplished . Soudry et al show this strategy works in their simulated networks ( even when the generative model is a hard-threshold leaky-integrate-and-fire rather than the nonlinear Hawkes model , which can be interpreted as a soft-threshold leaky-integrate-and-fire model; see SI ) . However , sampling the entire network may only be feasible in vitro; sampling of neurons in vivo , such as in wide-field calcium imaging studies , will still necessarily miss neurons not in the field of view or too deep in the tissue; in such cases our work provides the means to properly interpret the inferred effective interactions obtained with such a method . Although a thorough treatment of statistical inference of hidden network properties is beyond the scope of our present work , we may make some general remarks on future work in these directions . The nonlinear Hawkes model we use here is commonly used to fit neural population activity data , and one could infer the effective baselines μ r eff , interactions J r , r ′ eff ( t ) , and noise ξr ( t ) using existing techniques . In particular , Vidne et al . [22] explicitly fit the noise , which is likely important for proper inference , as otherwise effects of the noise could be artificially inherited by the effective interactions . Once such estimates are obtained , one could then in principle infer certain hidden network properties by combining a statistical model for these properties with the relationships between effective and true interactions derived in this work , such as Eq ( 2 ) . ( Detailed physiological measurements of ground-truth synaptic interactions in small volumes of neural tissue can be used to refine estimates ) . As we have stressed throughout this paper , inferring the exact connections between hidden neurons may be impossible due to a large number of degenerate solutions consistent with observations . However , one may be able to infer bulk properties of the network , such as the parameters governing the distribution of hidden-network connections , or even more exotic properties such as the eigenvalue distribution of the hidden network connection weight matrix . We leave these ideas as interesting directions for future work . Given the challenges that hidden network inference poses , one might wonder if there are network properties that can be reliably measured even with subsampled neural activity . Collective , low-dimensional dynamics have emerged as a possible candidate: recent work has investigated the effect that subsampled measurements have on estimating collective low-dimensional dynamics of trial-averaged network activity ( using , e . g . , principal components analysis ) . During a task , the effective dimensionality of a network’s dynamics is constrained [71 , 72] , opening the possibility that the subsampled population may be sufficient to accurately represent these task-constrained low-dimensional dynamics . Indeed , under certain assumptions—in particular that the collective dynamical modes are approximately random superpositions of neural activity and that sampled neurons are statistically representative of the hidden population—Gao et al . [72] calculate a conservative upper bound on the number of sampled neurons necessary to reconstruct the collective dynamics , finding it is often less than the effective dimensionality of the network . The assumption that the collective dynamics are random superpositions of neural activity is crucial , because it means that each neuron’s trial-averaged dynamics are in turn a superposition of the collective modes . Hence , every neuron’s activity contains some information about the collective modes , and if only a few of these modes are important , then they can be extracted from any sufficiently large subset of neurons . While modes of collective activity alone may be sufficient for answering certain questions , such as decoding task parameters or elucidating circuit function , explaining the structure of these modes—and in particular how the dynamical patterns that emerge under different task conditions or sensory environments are related—will ultimately require an understanding of the distribution of possible underlying network properties , which remain difficult to estimate from subsampled populations . We may be able to establish such structure-function relationships using our theory of effective interactions presented in this work: if we can relate the collective dynamics extracted from subsampled neurons to the properties of the effective interactions J r r ′ eff ( t ) , then we can link them to the true interactions through our Eq ( 2 ) . With an understanding of how network properties shape such collective dynamics , we can begin to understand what network manipulations achieve desired patterns of activity , and therefore circuit function . The fact that many different hidden networks may yield the same set of effective interactions or low-dimensional dynamics suggests that the effective interactions themselves may yield direct insight into a circuit’s functions . For instance , many circuits consist of principal neurons that transmit the results of circuit computation to downstream circuitry , but often do not make direct connections with one another , instead interacting through ( predominantly inhibitory ) intermediaries called interneurons . From the point of view of a downstream circuit , the principal neurons are “recorded” and the interneurons are “hidden . ” A potential reason for this general arrangement is that direct synaptic interactions alone are insufficient to produce the membrane responses required to perform the circuit’s computations , and the network of interneurons reshapes the membrane responses of projection neurons into effective interactions that can perform the desired computations—it may thus be that the effective interactions should be of primary interest , not necessarily the ( possibly degenerate choices of ) physiological synaptic interactions . For example , in the feedforward inhibitory circuits of Figs 3 and 4 , the roles of the hidden inhibitory neurons may simply be to act as interneurons that reshape the interaction between the excitatory projection neurons 1 and 2 , and the choice of which particular circuit motif is implemented in a real network is determined by other physiological constraints , not only computational requirements . One of the greatest achievements in systems neuroscience would be the ability to perform targeted modifications to a large neural circuit and selectively alter its suite of computations . This would have powerful applications for both studying a circuit’s native computations , but also repurposing circuits or repairing damaged circuitry ( due to , e . g . , disease ) . If the computational roles of circuits are indeed most sensitive to the effective interactions between principal neurons , this suggests we can exploit potential degeneracies in the interneuron architecture and intrinsic properties to find some circuit that achieves a desired computation , even if it is not a physiologically natural circuit . Our main result relating effective and true interactions , Eq ( 2 ) , provides a foundation for future work investigating how to identify sets of circuits that perform a desired set of computations . We have shown in this work that it can be done for small circuits ( Figs 3 and 4 ) , and that the effective interactions in large random networks can be significantly skewed away from the true interactions when synaptic weights scale as 1 / N , as observed in experiments [48] . This holds promise for identifying principled approaches to tuning or controlling neural interactions , such as by using neuromodulators to adjust interneuron properties or inserting artificial or synthetic circuit implants into neural tissue to act as “hidden” neurons . If successful , this could contribute to the long term goal of using such interventions to aid in reshaping the effective synaptic interactions between diseased neurons , and thereby restore healthy circuit behaviors .
The firing rate of a neuron i in the full network is given by λ i ( t ) = λ 0 ϕ ( μ i + μ i ext ( t ) + ∑ j ∫ - ∞ ∞ d t ′ J i j ( t - t ′ ) n ˙ j ( t ′ ) ) , ( 5 ) where λ0 is a characteristic rate , ϕ ( x ) ≥ 0 is a nonlinear function , μi ( potentially a function of some external stimulus θ ) is a time-independent tonic drive that sets the baseline firing rate of the neuron in the absence of input from other neurons , μ i ext ( t ) is an external input current , and Jij ( t − t′ ) is a coupling filter that filters spikes n ˙ j ( t ′ ) fired by presynaptic neuron j at time t′ , incident on post-synaptic neuron i . We will take μ i ext ( t ) = 0 for simplicity in this work , focusing on the activity of the network due to the tonic drives μi ( which could be still be interpreted as external tonic inputs , so the activity of the network need not be interpreted as spontaneous activity ) . While we need not attach a mechanistic interpretation to these filters , a convenient interpretation is that the nonlinear Hawkes model approximates the stochastic dynamics of a leaky integrate-and-fire network model driven by noisy inputs [55 , 56] . In fact , the nonlinear Hawkes model is equivalent to a current-based integrate-and-fire model in which the deterministic spiking rule ( a spike fires when a neuron’s membrane potential reaches a threshold value Vth ) is replaced by a stochastic spiking rule ( the higher a neuron’s membrane potential , the higher the probability a neuron will fire a spike ) . ( It can also be mapped directly to a conductance-based in special cases [73] ) . For completeness , we present the mapping from a leaky integrate-and-fire model with stochastic spiking to Eq ( 5 ) in the Supporting Information ( SI ) . To study how hidden neurons affect the inferred properties of recorded neurons , we partition the network into “recorded” neurons , labeled by indices r ( with sub- or superscripts to differentiate different recorded neurons , e . g . , r and r′ or r1 and r2 ) and “hidden” neurons labeled by indices h ( with sub- or superscripts ) . The rates of these two groups are thus λ r ( t ) = λ 0 ϕ ( μ r + ∑ r ′ J r , r ′ * n ˙ r ′ + ∑ h J r , h * n ˙ h ) , λ h ( t ) = λ 0 ϕ ( μ h + ∑ r J h , r * n ˙ r + ∑ h ′ J h , h ′ * n ˙ h ′ ) . To simplify notation , we write J i , j * n ˙ j = ∫ - ∞ ∞ d t ′ J i , j ( t - t ′ ) n ˙ j ( t ′ ) . If we seek to describe the firing of the recorded neurons only in terms of their own spiking history , input from hidden neurons effectively acts like noise with some mean amount of input . We thus begin by splitting the hidden input to the recorded neurons up into two terms , the mean plus fluctuations around the mean: ∑ h J r , h * n ˙ h ( t ) = ∑ h J r , h * E [ n ˙ h ( t ) | { n ˙ r } ] + ξ r ( t ) , where E [ n ˙ h ( t ) | { n ˙ r } ] denotes the mean activity of the hidden neurons conditioned on the activity of the recorded units , and ξr ( t ) are the fluctuations , i . e . , ξ r ( t ) ≡ ∑ h J r , h * ( n ˙ h - E [ n ˙ h ( t ) | { n ˙ r } ] ) . Note that ξr ( t ) is also conditional on the activity of the recorded units . By construction , the mean of the fluctuations is identically zero , while the cross-correlations can be expressed as E [ ξ r ( t ) ξ r ′ ( t ′ ) ] = ∫ - ∞ ∞ d t 1 d t 2 ∑ h 1 , h 2 J r , h 1 ( t - t 1 ) J r ′ , h 2 ( t ′ - t 2 ) C h 1 , h 2 ( t 1 , t 2 ) , where C h 1 , h 2 ( t 1 , t 2 ) is the cross-covariance between hidden neurons h1 and h2 ( conditioned on the spiking of recorded neurons ) . If the autocorrelation of the fluctuations ( r = r′ ) is small compared to the mean input to the recorded neurons ( ∑ h J r , h * E [ n ˙ h ( t ) | { n ˙ r } ] ) , or if Jr , h is small , then we may neglect these fluctuations and focus only on the effects that the mean input has on the recorded subnetwork . At the level of the mean field theory approximation we make in this work , the spike-train correlations are zero . One can calculate corrections to mean field theory ( see SI ) to estimate the size of this noise . Even when this noise is not strictly negligible , it can simply be treated as a separate input to the recorded neurons , as shown in the main text , and hence will not alter the form of the effective couplings between neurons . Averaging out the effective noise , however , would generate new interactions between neurons; we leave investigation of this issue for future work . In order to calculate how hidden input shapes the activity of recorded neurons , we need to calculate the mean E [ n ˙ h | { n ˙ r } ] . This mean input is difficult to calculate in general , especially when conditioned on the activity of the recorded neurons . In principle , the mean can be calculated as E [ n ˙ h | { n ˙ r } ] = E [ λ 0 ϕ ( μ h + ∑ r J h , r * n ˙ r + ∑ h ′ J h , h ′ * n ˙ h ′ ) |{ n ˙ r } ] . This is not a tractable calculation . Taylor series expanding the nonlinearity ϕ ( x ) reveals that the mean will depend on all higher cumulants of the hidden unit spike trains , which cannot in general be calculated explicitly . Instead , we again appeal to the fact that in a large , sufficiently connected network , we expect fluctuations to be small , as long as the network is not near a critical point . In this case , we may make a mean field approximation , which amounts to solving the self-consistent equation E [ n ˙ h | { n ˙ r } ] = λ 0 ϕ ( μ h + ∑ r J h , r * n ˙ r + ∑ h ′ J h , h ′ * E [ n ˙ h ′ | { n ˙ r } ] ) . ( 6 ) In general , this equation must be solved numerically . Unfortunately , the conditional dependence on the activity of the recorded neurons presents a problem , as in principle we must solve this equation for all possible patterns of recorded unit activity . Instead , we note that the mean hidden neuron firing rate is a functional of the filtered recorded input I h ( t ) ≡ ∑ r J h , r * n ˙ r , so we can expand it as a functional Taylor series around some reference filtered activity I h 0 ( t ) = ∑ r J h , r * n ˙ r 0 , E [ n ˙ h ( t ) | { I h ( t ) } ] = E [ n ˙ h ( t ) | { I h 0 ( t ) } ] + ∫ d t 1 ∑ h 1 δ E [ n ˙ h ( t ) | { I h 0 ( t ) } ] δ I h 1 ( t 1 ) ( I h 1 ( t 1 ) - I h 1 0 ( t 1 ) ) + 1 2 ∫ d t 1 d t 2 ∑ h 1 , h 2 δ 2 E [ n ˙ h ( t ) | { I h 0 ( t ) } ] δ I h 2 ( t 2 ) δ I h 1 ( t 1 ) ( I h 2 ( t 2 ) - I h 2 0 ( t 2 ) ) ( I h 1 ( t 1 ) - I h 1 0 ( t 1 ) ) + … Within our mean field approximation , the Taylor coefficients are simply the response functions of the network—i . e . , the zeroth order coefficient is the mean firing rate of the neurons in the reference state I h 0 ( t ) , the first order coefficient is the linear response function of the network , the second order coefficient is a nonlinear response function , and so on . There are two natural choices for the reference state I h 0 ( t ) . The first is simply the state of zero recorded unit activity , while the second is the mean activity of the recorded neurons . The zero-activity case conforms to the choice of nonlinear Hawkes models used in practice . Choosing the mean activity as the reference state may be more appropriate if the recorded neurons have high firing rates , but requires adjusting the form of the nonlinear Hawkes model so that firing rates are modulated by filtering the deviations of spikes from the mean firing rate , rather than filtering the spikes themselves . Here , we focus on the zero-activity reference state . We present the formulation for the mean field reference state in the SI . For the zero-activity reference state I h 0 ( t ) = 0 , the conditional mean is E [ n ˙ h ( t ) | { I h ( t ) } ] = E [ n ˙ h | 0 ] + ∫ d t 1 ∑ h 1 δ E [ n ˙ h ( t ) | 0 ] δ I h 1 ( t 1 ) I h 1 ( t 1 ) + 1 2 ∫ d t 1 d t 2 ∑ h 1 , h 2 δ 2 E [ n ˙ h ( t ) | 0 ] δ I h 2 ( t 2 ) δ I h 1 ( t 1 ) I h 2 ( t 2 ) I h 1 ( t 1 ) + … . The mean inputs E [ n ˙ h | 0 ] are the mean field approximations to the firing rates of the hidden neurons in the absence of the recorded neurons . Defining ν h ≡ E [ n ˙ h | 0 ] , these firing rates are given by ν h = λ 0 ϕ ( μ h + ∑ h ′ J h , h ′ ν h ′ ) ; in writing this equation we have assumed that the steady-state mean field firing rates will be time-independent , and hence the convolution J h , h ′ * ν h ′ = J h , h ′ ν h ′ , where J h , h ′ = ∫ 0 ∞ d t J h , h ′ ( t ) . This assumption will generally be valid for at least some parameter regime of the network , but there can be cases where it breaks down , such as if the nonlinearity ϕ ( x ) is bounded , in which case a transition to chaotic firing rates νh ( t ) may exist ( c . f . [74] ) . The mean field equations for the νh are a system of transcendental equations that in general cannot be solved exactly . In practice we will solve the equations numerically , but we can develop a series expansion for the solutions ( see below ) . The next term in the series expansion is the linear response function of the hidden unit network , Γ h , h ′ ( t - t ′ ) ≡ δ E [ n ˙ h ( t ) | 0 ] δ I h ′ ( t ′ ) , given by the solution to the integral equation Γ h , h ′ ( t - t ′ ) = γ h ( δ h , h ′ δ ( t - t ′ ) + ∑ h ′ ′ ∫ 0 ∞ d t ′ ′ J h , h ′ ′ ( t - t ′ ′ ) Γ h ′ ′ , h ′ ( t ′ ′ - t ′ ) ) . The “gain” γh is defined by γ h ≡ λ 0 ϕ ′ ( μ h + ∑ h ′ J h , h ′ ν h ′ ) , where ϕ′ ( x ) is the derivative of the nonlinearity with respect to its argument . For time-independent drives μr and steady states νh ( and hence γh ) , we may solve for Γh , h′ ( t − t′ ) by first converting to the frequency domain and then performing a matrix inverse: Γ ^ h , h ′ ( ω ) = [ I - V ^ ( ω ) ] h , h ′ - 1 γ h ′ , where V ^ h , h ′ ( ω ) = γ h J h , h ′ ( ω ) . If the zero and first order Taylor series coefficients in our expansion of E [ n ˙ h ( t ) | { n ˙ r } ] are the dominant terms—i . e . , if we may neglect higher order terms in this expansion—then we may approximate the instantaneous firing rates of the recorded neurons by λ r ( t ) ≈ λ 0 ϕ ( μ r eff + ∑ r ′ J r , r ′ eff * n ˙ r ′ ( t ) ) , where μ r eff = μ r + ∑ h J r , h ν h are the effective baselines of the recorded neurons and J ^ r , r ′ eff ( ω ) = J ^ r , r ′ ( ω ) + ∑ h , h ′ J ^ r , h ( ω ) Γ ^ h , h ′ ( ω ) J ^ h ′ , r ′ ( ω ) are the effective coupling filters in the frequency domain , as given in the main text . In addition to neglecting the higher order spike filtering terms here , we have also neglected fluctuations around the mean input from the hidden network . These fluctuations are zero within our mean field approximation , but we could in principle calculate corrections to the mean field predictions using the techniques of [53]; we do so to estimate the size of the effective noise correlations in the SI . In the main text , we decompose our expression for J ^ r , r ′ eff ( ω ) into contributions from all paths that a signal can travel from neuron r′ to r . To arrive at this interpretation , we note that we can expand Γ ^ h , h ′ ( ω ) in a series over paths through the hidden network . To start , we note that if | | V ^ ( ω ) | | < 1 for some matrix norm ||⋅|| , then the matrix [ I - V ( ω ) ] - 1 admits a convergent series expansion [75] [ I - V ^ ( ω ) ] - 1 = ∑ ℓ = 0 ∞ V ^ ( ω ) ℓ , where V ^ ( ω ) ℓ is a matrix product and V ^ ( ω ) 0 ≡ I . We can write an element of the matrix product out as [ V ^ ( ω ) ℓ ] h , h ′ = ∑ h 1 , … , h ℓ V ^ h , h 1 ( ω ) V ^ h 1 , h 2 ( ω ) … V ^ h ℓ - 1 , h ℓ ( ω ) V ^ h ℓ , h ′ ( ω ) ; inserting V ^ h i , h j ( ω ) = γ h i J ^ h i , h j ( ω ) yields [ V ^ ( ω ) ℓ ] h , h ′ = ∑ h 1 , … , h ℓ γ h J ^ h , h 1 ( ω ) γ h 1 J ^ h 1 , h 2 ( ω ) … γ h ℓ - 1 J ^ h ℓ - 1 , h ℓ ( ω ) γ h ℓ J ^ h ℓ , h ′ ( ω ) . This expression can be interpreted in terms of summing over paths through network of hidden neurons that join two observed neurons: the J ^ h i , h j ( ω ) are represented by edges from neuron hj to hi , and the γ h i are represented by the nodes . In this expansion , we allow edges from one neuron back to itself , meaning we include paths in which signals loop back around to the same neuron arbitrarily many times before the signal is propagated further . However , such loops can be easily factored , contributing a factor ∑ m = 0 ∞ ( γ h J ^ h , h ( ω ) ) m = 1 / ( 1 - γ h J ^ h , h ( ω ) ) . We thus remove the need to consider self-loops in our rules for calculating effective coupling filters by assigning a factor γh/ ( 1 − γh Jh , h ( ω ) ) to each node , as discussed in the main text and depicted in Fig 2 . ( The contribution of the self-feedback loops can be derived rigorously; see the SI for the full derivation ) . Although we have worked here in the frequency domain , our formalism does adapt straightforwardly to handle time-dependent inputs; however , among the consequences of this explicit time-dependence are that the mean field rates νh ( t ) are not only time-dependent , but solutions of a system of nonlinear integral equations , and hence more challenging to solve . Furthermore , quantities like the linear response of the hidden network , Γh , h′ ( t , t′ ) , will depend on both absolute times t and t′ , rather than just their difference , t − t′ , and hence we must also ( numerically ) solve for Γh , h′ ( t , t′ ) directly in the time domain . We leave these challenges for future work . Our main result , Eq ( 2 ) , is valid for general network architectures with arbitrary weighted synaptic connections , so long as the hidden subset of the network has stable dynamics when the recorded neurons are removed . An example for which our method must be modified would be a network in which all or the majority of the hidden neurons are excitatory , as the hidden network is unlikely to be stable when the inhibitory recorded neurons are disconnected . Similarly , we find that synaptic weight distributions with undefined moments will generally cause the network activity to be unstable . For example , J i , j drawn from a Cauchy distribution generally yield unstable network dynamics unless the weights are scaled inversely with a large power of the network size N . The nonlinear function ϕ ( x ) sets the instantaneous firing rate for the neurons in our model . Our main analytical results ( e . g . , Eq ( 2 ) hold for arbitrary choice of ϕ ( x ) . Where specific choices are required in order to perform simulations , we used ϕ ( x ) = max ( x , 0 ) for the results presented in Figs 3 and 4 and ϕ ( x ) = exp ( x ) otherwise . The rectified linear choice is convenient for small networks , as high-order derivatives are zero , which eliminates corresponding high-order “loop corrections” to mean field theory [53] . The exponential function is the “canonical” choice of nonlinearity for the nonlinear Hawkes process [16–18 , 20] . The exponential has particularly nice theoretical properties , but is also convenient for fitting the nonlinear Hawkes model to data , as the log-likelihood function of the model simplifies considerably and is convex ( though some similar families of nonlinearities also yield convex log-likelihood functions ) . An important property that both choices of nonlinearity possess is that they are unbounded . This property is necessary to guarantee that a neuron spikes given enough input . A bounded nonlinearity imposes a maximum firing rate , and neurons cannot be forced to spike reliably by providing a large bolus of input . The downside of an unbounded nonlinearity is that it is possible for the average firing rates to diverge , and the network never reaches a steady state . For example , in a purely excitatory network ( all J i , j ≥ 0 ) with an exponential nonlinearity , neural firing will run away without a sufficiently strong self-refractory coupling to suppress the firing rate . This will not occur with a bounded nonlinearity , as excitation can only drive neurons to fire at some maximum but finite rate . This can be a problem in simulations of networks obeying Dale’s law . For unbounded nonlinearities , the mean field theory for the hidden network occasionally does not exist due to an imbalance of excitatory and inhibitory neurons caused by our random selection of recorded of neurons . However , the Dale’s law network is stable if the nonlinearity is bounded . We demonstrate this below in Figs 7 and 8 , comparing simulations of the effective interaction statistics in Erdős-Réyni networks with and without Dale’s law for a sigmoidal nonlinearity ϕ ( x ) = 2/ ( 1 + e−x ) . Another consequence of unbounded nonlinearities is that the mean firing rates are either finite or they diverge . Bounded nonlinearities , on the other hand , may allow for the possibility of a transition to chaotic dynamics in the mean-field firing rate dynamics ( cf . the results of the [74] ) . The mean field firing rates for the hidden neurons are given by ν h = λ 0 exp ( μ h + ∑ h ′ J h , h ′ ν h ′ ) , where we focus specifically on the case of exponential nonlinearity ϕ ( x ) = exp ( x ) . For this choice of nonlinearity , γh = νh , so we do not need to calculate a separate series for the gains . This system of transcendental equations generally cannot be solved analytically . However , for small exp ( μh ) ≪ 1 we can derive , recursively , a series expansion for the firing rates . We first consider the case of μh = μ0 for all hidden neurons h . Let ϵ = exp ( μ0 ) . We may then write ν h = λ 0 ϵ ∑ ℓ = 0 ∞ a h ( ℓ ) ( λ 0 ϵ ) ℓ . Plugging this into the mean field equation , ∑ ℓ = 0 ∞ a h ( ℓ ) ( λ 0 ϵ ) ℓ = exp ( ∑ h ′ J h , h ′ ∑ ℓ = 0 ∞ a h ′ ( ℓ ) ( λ 0 ϵ ) ℓ + 1 ) = 1 + ∑ m = 1 ∞ 1 m ! ( ∑ h ′ J h , h ′ ∑ ℓ = 0 ∞ a h ′ ( ℓ ) ( λ 0 ϵ ) ℓ + 1 ) m = 1 + ∑ m = 1 ∞ 1 m ! ∑ ℓ 1 , … , ℓ m , h 1 ′ , … , h m ′ J h , h 1 ′ a h 1 ′ ( ℓ 1 ) … J h , h m ′ a h m ′ ( ℓ m ) ( λ 0 ϵ ) ℓ 1 + ⋯ + ℓ m + m = 1 + ∑ ℓ = 1 ∞ { ∑ m = 1 ∞ 1 m ! ∑ ℓ 1 , … , ℓ m , h 1 ′ , … , h m ′ J h , h 1 ′ a h 1 ′ ( ℓ 1 ) … J h , h m ′ a h m ′ ( ℓ m ) δ ℓ , ℓ 1 + ⋯ + ℓ m + m } ( λ 0 ϵ ) ℓ . Thus , matching powers of λ0ϵ on the left and right hand sides , we find a h ( 0 ) = 1 and a h ( ℓ ) = ∑ m = 1 ∞ 1 m ! ∑ ℓ 1 , … , ℓ m , h 1 ′ , … , h m ′ J h , h 1 ′ a h 1 ′ ( ℓ 1 ) … J h , h m ′ a h m ′ ( ℓ m ) δ ℓ , ℓ 1 + ⋯ + ℓ m + m for ℓ > 0 . For ℓ = 1 , the sum in m truncates at m = 1 ( as δ ℓ , ℓ 1 + ⋯ + ℓ m + m is zero for m > ℓ , as all indices are positive ) . Thus , a h ( 1 ) = ∑ h 1 ′ J h , h 1 ′ , a h ( 2 ) = ∑ h 1 ′ , h 2 ′ { J h , h 1 ′ J h 1 ′ , h 2 ′ + 1 2 J h , h 1 ′ J h , h 2 ′ } , a h ( 3 ) = ∑ h 1 ′ , h 2 ′ , h 3 ′ { J h , h 1 ′ J h 1 ′ , h 2 ′ J h 2 ′ , h 3 ′ + 1 2 J h , h 1 ′ J h 1 ′ , h 2 ′ J h 1 ′ , h 3 ′ + J h , h 1 ′ J h , h 2 ′ J h 2 ′ , h 3 ′ + 1 3 ! J h , h 1 ′ J h , h 2 ′ J h , h 3 ′ } . With this we have calculated the firing rates to O ( ϵ 4 ) . The analysis can be straightforwardly extended to the case of heterogeneous μh , though it becomes more tedious to compute terms in the ( now multivariate ) series . Assuming ϵh ≡ exp ( μh ) ≪ 1 for all h , to O ( ϵ 3 ) we find ν h = λ 0 ϵ h ( 1 + ∑ h ′ J h , h ′ λ 0 ϵ h ′ + ∑ h 1 ′ , h 2 ′ { J h , h 1 ′ J h 1 ′ , h 2 ′ + 1 2 J h , h 1 ′ J h , h 2 ′ } λ 0 ϵ h 1 ′ λ 0 ϵ h 2 ′ + … ) . To estimate the strength of the hidden paths , we would like to calculate the variance of the effective coupling J r , r ′ eff and compare its strength to the variance of the direct couplings J r , r ′ , where J r , r ′ eff ≡ ∫ 0 ∞ d t J r , r ′ eff ( t ) and J r , r ′ ≡ ∫ 0 ∞ d t J r , r ′ ( t ) , as in the main text . We assume that the synaptic weights J i , j are independently and identically distributed with zero mean and variance var ( J ) = p J 0 2 ( p N ) 2 a for i ≠ j , where a = 1 corresponds to weak coupling and a = 1/2 corresponds to strong coupling . We assume no self-couplings , J i , i = 0 for all neurons i . The overall factor of p in var [ J ] comes from the sparsity of the network . For example , for normally distributed non-zero weights with variance J 0 2 / N 2 a , the total probability for every connection in the network is ρ E R × J ( J ) = ( 1 - p ) δ ( J ) + p exp ( - N 2 a 2 J 2 J 0 2 ) 2 π J 0 2 / N 2 a . Because the J i , j are i . i . d . , the mean of J r , r ′ eff: J r , r ′ eff ¯ = J r , r ′ ¯ + ∑ h , h ′ J r , h Γ ^ h , h ′ J h ′ , r ′ ¯ = 0 + ∑ h , h ′ J r , h ¯ Γ ^ h , h ′ ¯ J h ′ , r ′ ¯ = 0 , where we used the fact that Γ ^ h , h ′ ≡ Γ ^ h , h ′ ( 0 ) depends only on the hidden neuron couplings J h , h ′ , which are independent of the couplings to the recorded neurons , J r , h and J h ′ , r ′ . This holds for any pair of neurons ( r , r′ ) , including r = r′ because of the assumption of no self-coupling . The variance of J r , r ′ eff is thus equal to the mean of its square , for r ≠ r′ , var [ J r , r ′ eff ] = ( J r , r ′ eff ) 2 ¯ = ( J r , r ′ ) 2 ¯ + ( ∑ h , h ′ J r , h Γ ^ h , h ′ J h ′ , r ′ ) 2 ¯ = var [ J ] + ∑ h 1 , h 1 ′ , h 2 , h 2 ′ J r , h 1 Γ ^ h 1 , h 1 ′ J h 1 ′ , r ′ J r , h 2 Γ h 2 , h 2 ′ J h 2 ′ , r ′ ¯ = var [ J ] + ∑ h , h ′ J r , h 2 ¯ Γ ^ h , h ′ 2 ¯ J h ′ , r ′ 2 ¯ = var [ J ] + var [ J ] 2 ∑ h , h ′ Γ ^ h , h ′ 2 ¯ In this derivation , we used the fact that J r , h 1 J r , h 2 ¯ = J r , h 1 2 ¯ δ h 1 , h 2 due to the fact that the synaptic weights are uncorrelated . We now need to compute Γ ^ h , h ′ 2 ¯ . This is intractable in general , so we will resort to calculating this in a series expansion in powers of ϵ ≡ exp ( μ0 ) for the exponential nonlinearity model . Our result will also turn out to be an expansion in powers of J0 and 1 − f ≡ Nhid/N . The lowest order approximation is obtained by the approximation νh ≈ λ0ϵ and Γh , h′ ≈ νhδh , h′ , yielding var [ J r , r ′ eff ] var [ J ] = 1 + ( λ 0 ϵ ) 2 N hid var [ J ] = 1 + ( λ 0 J 0 ϵ ) 2 ( 1 - f ) 1 ( p N ) 2 a - 1 . ( 7 ) This result varies linearly with f , while numerical evaluation of the variance shows obvious curvature for f ≪ 1 and J0 ≲ 1 , so we need to go to higher order . This becomes tedious very quickly , so we will only work to O ( ϵ 4 ) ( it turns out O ( ϵ 3 ) corrections vanish ) . We calculate Γ ^ h , h ′ 2 ¯ using a recursive strategy , though we could also use the path-length series expression for Γ ^ h , h ′ ( ω ) , keeping terms up to fourth order in ϵ . We begin with the expression Γ ^ h , h ′ = ν h δ h , h ′ + ∑ h ′ ′ ν h J h , h ′ ′ Γ ^ h ′ ′ , h ′ and plug it into itself until we obtain an expression to a desired order in ϵ . In doing so , we note that ν h ∼ O ( ϵ ) , so we will first work to fourth order in νh , and then plug in the series for νh in powers of ϵ . We begin with Γ ^ h , h ′ 2 = ν h 2 δ h , h ′ + 2 δ h , h ′ ∑ h ′ ′ ν h 2 J h , h ′ ′ Γ ^ h ′ ′ , h ′ + ( ∑ h ′ ′ ν h J h , h ′ ′ Γ ^ h ′ ′ , h ′ ) 2 = ν h 2 δ h , h ′ + 2 δ h , h ′ ∑ h ′ ′ ν h 2 J h , h ′ ′ Γ ^ h ′ ′ , h ′ + ∑ h 1 , h 2 ν h 2 J h , h 1 J h , h 2 Γ ^ h 1 , h ′ Γ ^ h 2 , h ′ ≈ ν h 2 δ h , h ′ + 2 δ h , h ′ ∑ h ′ ′ ν h 2 J h , h ′ ′ { ν h ′ ′ δ h ′ ′ , h ′ + ∑ h 2 ν h ′ ′ J h ′ ′ , h 2 ν h 2 δ h 2 , h ′ } + ∑ h 1 , h 2 ν h 2 ν h ′ 2 J h , h 1 J h , h 2 δ h 1 , h ′ δ h 2 , h ′ = ν h 2 δ h , h ′ + 2 δ h , h ′ { ν h 2 ν h ′ J h , h ′ + ∑ h ′ ′ ν h 2 ν h ′ ′ J h , h ′ ′ J h ′ ′ , h ′ ν h ′ } + ν h 2 ν h ′ 2 J h , h ′ 2 = { ν h 2 + 2 ν h 2 ν h ′ J h , h ′ + 2 ∑ h ′ ′ ν h 2 ν h ′ ′ J h , h ′ ′ J h ′ ′ , h ′ ν h ′ } δ h , h ′ + ν h 2 ν h ′ 2 J h , h ′ 2 = { ν h 2 + 2 ∑ h ′ ′ ν h 3 ν h ′ ′ J h , h ′ ′ J h ′ ′ , h } δ h , h ′ + ν h 2 ν h ′ 2 J h , h ′ 2 The third order term ν h 3 J h , h ′ δ h , h ′ vanished because we assume no self-couplings . We have obtained Γ ^ h , h ′ 2 to fourth order in νh; now we need to plug in the series expression for νh to obtain the series in powers of λ0ϵ . We will do this order by order in νh . The easiest terms are the fourth order terms , as ν h 2 ν h ′ 2 ≈ ( λ 0 ϵ ) 4 and ν h 3 ν h ′ ′ ≈ ( λ 0 ϵ ) 4 . The second order term is ν h 2 ≈ ( λ 0 ϵ ) 2 ( 1 + ∑ h 1 J h , h 1 λ 0 ϵ + ∑ h 1 , h 2 a h , h 1 , h 2 ( 2 ) ( λ 0 ϵ ) 2 ) × ( 1 + ∑ h 1 ′ J h , h 1 ′ λ 0 ϵ + ∑ h 1 ′ , h 2 ′ a h , h 1 ′ , h 2 ′ ( 2 ) ( λ 0 ϵ ) 2 ) ≈ ( λ 0 ϵ ) 2 ( 1 + 2 ( ∑ h 1 J h , h 1 λ 0 ϵ + ∑ h 1 , h 2 a h , h 1 , h 2 ( 2 ) ( λ 0 ϵ ) 2 ) + ( ∑ h 1 J h , h 1 λ 0 ϵ ) 2 ) = ( λ 0 ϵ ) 2 ( 1 + 2 ∑ h 1 J h , h 1 λ 0 ϵ + ∑ h 1 , h 2 { 2 a h , h 1 , h 2 ( 2 ) + J h , h 1 J h , h 2 } ( λ 0 ϵ ) 2 ) , where a h , h 1 , h 2 ( 2 ) = J h , h 1 J h 1 , h 2 + 1 2 J h , h 1 J h , h 2 . We need the average ν h 2 ¯ . The third-order term will vanish upon averaging , and 2 a h , h 1 , h 2 ( 2 ) + J h , h 1 J h , h 2 ¯ = 2 J h , h 1 J h 1 , h 2 + 2 J h , h 1 J h , h 2 ¯ = 2 var [ J ] δ h 1 , h 2 ( 1 - δ h , h 1 ) , using the fact that synaptic weights are independent ( giving the δ h 1 , h 2 factor ) and self-couplings are zero ( giving the 1 - δ h , h 1 factor ) . We thus obtain ν h 2 ¯ = ( λ 0 ϵ ) 2 + 2 ( λ 0 ϵ ) 4 ( N hid - 1 ) var [ J ] . The first fourth order term in Γ ^ h , h ′ 2 , 2 ∑ h ″ ν h 3 ν h ″ J h , h ″ J h ″ , h δ h , h ′ , will vanish upon averaging because matching indices requires h′′ = h = h′ and we assume no self-couplings . The second fourth order term is J h , h ′ 2 , which averages to var [ J ] ( 1 - δ h , h ′ ) , where the factor of ( 1 − δh , h′ ) again accounts for the fact that this term does not contribute when h = h′ due to no self-couplings . We thus arrive at Γ ^ h , h ′ 2 ¯ = ( ( λ 0 ϵ ) 2 + 2 ( λ 0 ϵ ) 4 ( N hid - 1 ) var [ J ] ) δ h , h ′ + ( λ 0 ϵ ) 4 var [ J ] ( 1 - δ h , h ′ ) = ( ( λ 0 ϵ ) 2 + ( λ 0 ϵ ) 4 ( 2 N hid - 3 ) var [ J ] ) δ h , h ′ + ( λ 0 ϵ ) 4 var [ J ] ; Putting everything together , var [ J r , r ′ eff ] var [ J ] = 1 + var [ J ] ∑ h , h ′ Γ ^ h , h ′ 2 ¯ = 1 + var [ J ] [ ∑ h { ( λ 0 ϵ ) 2 + ( λ 0 ϵ ) 4 ( 2 N hid - 3 ) var [ J ] } + ∑ h , h ′ ( λ 0 ϵ ) 4 var [ J ] ] = 1 + var [ J ] [ N hid { ( λ 0 ϵ ) 2 + ( λ 0 ϵ ) 4 ( 2 N hid - 3 ) var [ J ] } + N hid 2 ( λ 0 ϵ ) 4 var [ J ] ] = 1 + N hid var [ J ] [ ( λ 0 ϵ ) 2 + ( λ 0 ϵ ) 4 ( 2 N hid - 3 ) var [ J ] + N hid ( λ 0 ϵ ) 4 var [ J ] ] = 1 + N hid var [ J ] [ ( λ 0 ϵ ) 2 + ( λ 0 ϵ ) 4 ( 3 - 3 N hid ) N hid var [ J ] ] For weak coupling , this tends to 1 in the N ≫ 1 limit , as N hid var [ J ] = ( 1 - f ) J 0 2 / N → 0 , for fixed fraction of observed neurons f = Nrec/N . For strong coupling , N hid var [ J ] = ( 1 - f ) J 0 2 , which is constant as N → ∞ , and hence var [ J r , r ′ eff ] var [ J ] = 1 + ( λ 0 J 0 ϵ ) 2 ( 1 - f ) + 3 ( λ 0 J 0 ϵ ) 4 ( 1 - f ) 2 + o ( ( λ 0 J 0 ϵ ) 4 ( 1 - f ) 2 ) , ( 8 ) where we have used little-o notation to denote that there are higher order terms dominated by ( λ0 J0ϵ ) 4 ( 1 − f ) 2 . With this expression , we have improved on our approximation of the relative variance of the effective coupling to the true coupling; however , the neglected higher order terms still become significant as f → 0 and J0 → 1 , indicating that hidden paths have a significant impact when synaptic strengths are moderately strong and only a small fraction of the neurons have been observed . Because the synaptic weights J i , j are independent , we may rewrite Eq ( 8 ) as var [ J r , r ′ eff - J r , r ′ ] var [ J ] ≈ ( λ 0 J 0 ϵ ) 2 ( 1 - f ) + 3 ( λ 0 J 0 ϵ ) 4 ( 1 - f ) 2 ; or , in terms of the ratio of standard deviations , σ [ J r , r ′ eff - J r , r ′ ] σ [ J ] ≈ ( λ 0 J 0 ϵ ) 1 - f ( 1 + 3 2 ( λ 0 J 0 ϵ ) 2 ( 1 - f ) ) , where we used the approximation 1 + x ≈ 1 + x / 2 for x small . In the main text , we plotted results for N = 1000 total neurons ( Fig 5A ) . For strongly coupled networks , the results should only depend on the fraction of observed neurons , f = Nrec/N , while for weak coupling the results do depend on the absolute number N . To demonstrate this , in Fig 9 we remake Fig 5 for N = 100 neurons . We see that the strongly coupled results have not been significantly altered , whereas the weakly coupled results yield stronger deviations ( as the deviations are O ( 1 / N ) ) . | No experiment in neuroscience can record from more than a tiny fraction of the total number of neurons present in a circuit . This severely complicates measurement of a network’s true properties , as unobserved neurons skew measurements away from what would be measured if all neurons were observed . For example , the measured post-synaptic response of a neuron to a spike from a particular pre-synaptic neuron incorporates direct connections between the two neurons as well as the effect of any number of indirect connections , including through unobserved neurons . To understand how measured quantities are distorted by unobserved neurons , we calculate a general relationship between measured “effective” synaptic interactions and the ground-truth interactions in the network . This allows us to identify conditions under which hidden neurons substantially alter measured interactions . Moreover , it provides a foundation for future work on manipulating effective interactions between neurons to better understand and potentially alter circuit function—or dysfunction . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] | [
"medicine",
"and",
"health",
"sciences",
"action",
"potentials",
"neural",
"networks",
"decision",
"making",
"nervous",
"system",
"membrane",
"potential",
"social",
"sciences",
"electrophysiology",
"neuroscience",
"cognitive",
"psychology",
"cognition",
"network",
"analysis",
"computer",
"and",
"information",
"sciences",
"animal",
"cells",
"neural",
"pathways",
"cellular",
"neuroscience",
"psychology",
"neuroanatomy",
"cell",
"biology",
"anatomy",
"synapses",
"physiology",
"neurons",
"biology",
"and",
"life",
"sciences",
"cellular",
"types",
"cognitive",
"science",
"neurophysiology"
] | 2018 | Predicting how and when hidden neurons skew measured synaptic interactions |
During the 2014 Ebola Virus Disease ( EVD ) epidemic , the Ebola-Tx trial evaluated the use of convalescent plasma ( CP ) in Guinea . The effectiveness of plasmapheresis trials depends on the recruitment of plasma donors . This paper describes what motivated or deterred EVD survivors to donate CP , providing insights for future plasmapheresis trials and epidemic preparedness . This qualitative study , part of Ebola-Tx , researched and addressed emergent trial difficulties through interviewing , participant observation and focus group discussions . Sampling was theoretical and retroductive analysis was done in NVivo 10 . Willingness or hesitance to participate in plasma donation depended on factors at the interface of pre-existing social dynamics; the impact of the disease and the consequent emergency response including the trial set-up . For volunteers , motivation to donate was mainly related to the feeling of social responsibility inspired by having survived EVD and to positive perceptions of plasmapheresis technology despite still unknown trial outcomes . Conversely , confidentiality concerns when volunteering due to stigmatization of survivors and perceived decrease in vital strength and in antibodies when donating , leading to fears of loss in protection against EVD , were main deterrents . The dynamic ( dis ) trust in Ebola Response Actors and in other survivors further determined willingness to participate and lead to the emergence/decline of rumours related to blood stealing and treatment effectiveness . Historic inter-ethnic relations in the health care setting further defined volunteering along socio-economic and ethnic lines . Finally , lack of follow-up and of dedicated care further impacted on motivation to volunteer . Ebola-Tx was the first trial to solicit and evaluate blood-product donation as an experimental treatment on a large scale in Sub-Saharan Africa . An effective donation system requires directly engaging with emergent social barriers and providing an effective ethical response , including improved and transparent communication , effective follow-up after donation , assuring confidentiality and determining ethical incentives .
The urgency to contain the 2013–2016 West-African Ebola virus disease ( EVD ) epidemic led to the World Health Organization’s prioritization of the clinical evaluation of convalescent plasma ( CP ) or whole blood in September 2014 [1–4] as a potential cure , resulting in fast-tracked non-randomized clinical trials in Sierra Leone , Guinea and Liberia [5] . The Ebola-Tx trial in Guinea aimed to determine whether the administration of EVD antibodies from the blood plasma of EVD survivors could increase EVD patients’ survival rate compared to a historical control group receiving only supportive care at the same site . Although the treatment was found to be safe , receiving two consecutive units of 200-250ml CP did not significantly improve the overall survival rate [6] . The effective implementation of such trials depends on the recruitment of EVD survivors as volunteers for plasma donation . Throughout sub-Saharan Africa , even the sampling of small amounts of blood has been associated with considerable fears of ‘blood stealing’ or ‘blood selling’ , documented since colonial times [7 , 8] , leading to reticence for participation in clinical research [8–13] . Reports from Guinea , Sierra Leone and Liberia suggested similar concerns as blood-related rumors were frequently reported during the EVD outbreak [14–17] , an example of which being the local suspicion that burial teams of EVD victims “filled their tanks” with the blood of the deceased in exchange for international aid [18] . There were further concerns that the stigmatization of EVD survivors [19 , 20] could hinder efforts to recruit survivors to donate plasma [21 , 22] . This stigmatization was often related to suspicions of Ebola Treatment Centre workers and survivors having killed patients through mystical means to ensure their own survival in addition to fears of survivors still being infectious . Due to these complexities , scientists were concerned that collection of CP from EVD survivors during the outbreak might not be feasible . Nevertheless , the Ebola-Tx trial recruited 98 survivors , sufficient to provide experimental treatment to all trial patients [23] . This anthropological study was part of the Ebola-Tx trial “Emergency evaluation of convalescent plasma for Ebola Virus Disease ( EVD ) in Guinea” ( NCT02342171 ) and aimed to provide insights into what motivated or deterred EVD survivors to donate convalescent plasma , in order to contribute to the effectiveness of future plasmapheresis trials and outbreak preparedness .
The qualitative research was carried out in both Ebola-Tx trial sites in Guinea– ( i ) the National Blood Transfusion Centre ( NBTC ) , and ( ii ) the Médecins Sans Frontières ( MSF ) run Donka Ebola Treatment Centre–as well as in surrounding areas in Conakry and the three peripheral prefectures of Coyah , Kindia and Dubréka . Sampling for the study was theoretical ( i . e . based on emerging results ) [28] ( Table 1 ) . Participants were selected based on characteristics such as involvement in the trial , social ties with survivors , professional background , socio-economic status , gender , age , and urban or rural residence . Data collection was intermittently carried out between January and September 2015 by two European anthropologists from ITM , in some cases accompanied by an EVD survivor/gatekeeper . This period coincided with phases E and F of the Guinean epidemic [29 , 30] , right after the highest peak of October-December 2014 , characterized by a decrease in cases and a movement of the epidemic to the Lower Guinean region ( including the capital ) . The position of the researchers and gatekeepers , especially within the context of this epidemic , initially proved challenging in gaining trust of different types of informants . This was gradually overcome through regular visits and constant rapport building . The anthropological team often reflected on and discussed their role in the trial and thus navigated the tension between their critical independence with regards to the trial on the one hand , and , on the other hand , their direct involvement in strategies and activities to improve the effectiveness of the trial . Semi-structured in-depth interviews were carried out using continuously evolving question guides addressing personal experiences with EVD , power relations and social dynamics within the SA , perception and experience of the trial , recommendations for improvement , plasma versus blood perceptions , internal work dynamics , recommendations for further steps of the trial , suggestions for improvement , and rumors pertaining to the trial itself as well as to the Ebola outbreak more generally . Participant observation was used at different stages and places linked to trial preparation and implementation , allowing insights into the trial dynamics . Focus Group Discussions aimed at eliciting contrasting opinions . Separate group discussions were organized for survivors , donors , trial staff ( including doctors ) and patients . Analysis was retroductive and concurrent to data collection . Preliminary data were intermittently analyzed in the field . Temporary results were further confirmed or refuted through constant validity checks until saturation was reached and the data was theoretically supported . Thematic analysis was carried out with NVivo Qualitative Data Analysis software ( QSR International Pty Ltd . Cardigan UK ) . Oral informed consent was obtained from all adult participants . Considering the sensitive nature of this qualitative research , oral rather than written consent was preferred for a number of reasons: ( i ) to avoid the high illiteracy in the study population challenging the practical obtaining of written consent; ( ii ) written consent might seem a breach of confidentiality by the participants; ( iii ) written consent , requiring a signature and therefore introducing formality in the procedure , affects trust between respondents and researchers and consequently willingness to participate in the study , as conversations usually do not require formal procedures . In addition , this formality introduces bias and hence reduces data quality . Informed consent was documented in written by the person who obtained the consent and , if present , by a witness . The study , including the use of oral consent , was reviewed and approved by the Institutional Review Board of the ITM ( IRB/AB/ac/092 Ref:1098/16 ) and the Guinean National Ethics Committee for Research in Health–CNERS ( 131/CNERS/16 ) . Ethical reviews and approval for the Ebola-Tx trial are described in [6] .
Blood was considered sacred and a vital element associated with health . A decrease in volume of blood was perceived to weaken the body and cause ( potentially irreversible ) sickness , limiting people’s willingness to participate in the trial . Despite the use of plasma instead of blood , and the trial team’s explicit explanation of the difference ( e . g . in color , time for recovery , quantities required ) , participants still perceived the procedure to be “blood” donation . The absence of a word for plasma in the Susu language , leading to the use of the word ‘blood’ ( ‘wouly’ ) , contributed to this ambiguity . Donating plasma was thus associated with blood donation , leading to common fears such as the fear of the “unknown” for first-time donors , fear of knowing one’s serological status , fear of the pain , needles , and physical weakening , especially when survivors felt they had not completely recovered from EVD or when they felt they were affected by poor nutrition . Some donors reported needing considerable time to recover . For donors who , conversely , were positively surprised by their smooth recovery , this was mostly linked to their perceptions of the technology being cutting-edge . In addition , the Plasma Mobile and its high-quality equipment was welcomed by survivors as an acknowledgement of previous hardship . Survivors , especially those with a medical background , perceived the loss of immunity and the increased risk of reinfection and relapse due to the transference of antibodies when donating to be a health risk , as illustrated by the following quote: “You know it’s the antibodies that helped us recover . If you want to weaken our antibodies even more , what am I going to do ? Won’t I become sick again ? ” ( EVD survivor , IDI ) One of the main factors influencing decision-making was trust , or the absence of it , in the organizations and people involved . Due to the stigmatization of EVD survivors by neighbors , relatives and more broadly in social and professional networks , the difficulty of dealing with anonymity was decisive for non-participation as it could expose donors as survivors . While some survivors required to be contacted by peers in a confidential way , others demanded openness from peer educators and requested them to sensitize their families , prior to taking the decision to participate . The perception of contributing to patients’ survival and indirect motives such as a sense of patriotism , the feeling of being part of a historic fight against a lethal disease and the feeling of being ‘in debt’ because they survived while others did not , leading to a moral obligation to do something in return , motivated potential donors to participate in the trial .
Time limitations , security measures , initial people’s distrust and financial and political incentives for respondents as part of the response to the epidemic did not always allow for the researchers’ immersion in the study setting as is usual with ethnographic research methods . This limited access to sensitive information and informants , which limited the scope and depth of the collected data .
Actively engaging with , responding to and following up participants , especially when concerning vulnerable populations such as EVD survivors enrolled as plasma donors , should be integral to the trial design . Effective and ethical donation systems , emergency trials , and even trials in low-resource settings more broadly , require directly engaging with emergent factors that occur at the interface between pre-existing social dynamics , the impact of the disease , and the consequent ( emergency ) response including the trial set-up . We advocate for the integration of methodologies to face emergent factors into trial designs from the start . | During the 2014 West-African Ebola Virus Disease epidemic , the Ebola-Tx clinical trial in Guinea aimed to determine whether the administration of Ebola antibodies from the blood plasma of Ebola survivors could increase Ebola patients’ survival rate . Ebola-Tx was the first trial to solicit and evaluate blood-product donation as an experimental treatment on a large scale in Sub-Saharan Africa . In this qualitative study , part of the Ebola-Tx study , we report on factors of motivation and demotivation influencing Ebola survivors to donate their plasma . Understanding these factors is essential as the successful recruitment of this specific subgroup , providing the therapy itself , directly impacts the effectiveness of such a trial . We show that organizing an effective and ethical donation system requires directly engaging with emerging social barriers at the interface between pre-existing social dynamics , the impact of the disease , and the consequent emergency response including the trial set-up . These results provide insights that can be useful for future plasma trials but also for emergency clinical trials as part of general epidemic preparedness . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion",
"Conclusion"
] | [
"guinea",
"medicine",
"and",
"health",
"sciences",
"clinical",
"laboratory",
"sciences",
"body",
"fluids",
"pathology",
"and",
"laboratory",
"medicine",
"education",
"pathogens",
"sociology",
"geographical",
"locations",
"tropical",
"diseases",
"social",
"sciences",
"ebola",
"hemorrhagic",
"fever",
"microbiology",
"viruses",
"research",
"design",
"filoviruses",
"rna",
"viruses",
"neglected",
"tropical",
"diseases",
"transfusion",
"medicine",
"africa",
"research",
"and",
"analysis",
"methods",
"medical",
"education",
"viral",
"hemorrhagic",
"fevers",
"infectious",
"diseases",
"medical",
"microbiology",
"microbial",
"pathogens",
"qualitative",
"studies",
"blood",
"plasma",
"hematology",
"people",
"and",
"places",
"diagnostic",
"medicine",
"blood",
"anatomy",
"medical",
"humanities",
"ebola",
"virus",
"physiology",
"viral",
"pathogens",
"biology",
"and",
"life",
"sciences",
"viral",
"diseases",
"blood",
"transfusion",
"hemorrhagic",
"fever",
"viruses",
"organisms"
] | 2018 | What motivates Ebola survivors to donate plasma during an emergency clinical trial? The case of Ebola-Tx in Guinea |
The human blood fluke Schistosoma mansoni causes intestinal schistosomiasis , a widespread neglected tropical disease . Infection of freshwater snails Biomphalaria spp . is an essential step in the transmission of S . mansoni to humans , although the physiological interactions between the parasite and its obligate snail host that determine success or failure are still poorly understood . In the present study , the B . glabrata embryonic ( Bge ) cell line , a widely used in vitro model for hemocyte-like activity , was used to investigate membrane properties , and assess the impact of larval transformation proteins ( LTP ) on identified ion channels . Whole-cell patch clamp recordings from Bge cells demonstrated that a Zn2+-sensitive H+ channel serves as the dominant plasma membrane conductance . Moreover , treatment of Bge cells with Zn2+ significantly inhibited an otherwise robust production of reactive oxygen species ( ROS ) , thus implicating H+ channels in the regulation of this immune function . A heat-sensitive component of LTP appears to target H+ channels , enhancing Bge cell H+ current over 2-fold . Both Bge cells and B . glabrata hemocytes express mRNA encoding a hydrogen voltage-gated channel 1 ( HVCN1 ) -like protein , although its function in hemocytes remains to be determined . This study is the first to identify and characterize an H+ channel in non-neuronal cells of freshwater molluscs . Importantly , the involvement of these channels in ROS production and their modulation by LTP suggest that these channels may function in immune defense responses against larval S . mansoni .
Schistosomiasis , a neglected tropical disease afflicting over 250 million people worldwide [1] , is caused by parasitic flatworms of the genus Schistosoma . Schistosoma spp . have a two-host life cycle involving sexual reproduction within a mammalian host and asexual reproduction within a snail intermediate host . The pathology associated with the intestinal form of human schistosomiasis arises in chronic infections when eggs released by female worms occupying mesenteric veins become trapped in the liver ( and other organs ) and elicit an intense inflammatory response leading to the formation of granulomas that damage tissues and block circulation [2 , 3] . Eggs from ruptured intestinal capillaries exit the host by fecal excretion , and upon exposure to freshwater , hatch to release the free-swimming snail-infective miracidia . Upon infection of snails , miracidia transform through two sporocyst stages , ultimately completing their life cycle by the production and release of free-swimming cercariae , the human-infective stage [4] . Because of the absolute dependency of human schistosome transmission on the snail host , one of the keys to sustained control of schistosomiasis is to block or eliminate the snail’s participation in the life cycle . The freshwater snail Biomphalaria glabrata serves as the most common invertebrate host of S . mansoni , the most widely distributed species of Schistosoma [5] . Hemocytes ( phagocytic immune cells ) of B . glabrata , genetically-selected for susceptibility or resistance to infection by larval S . mansoni , have been shown to react differentially to invading miracidia . Circulating hemocytes of susceptible strains do not recognize and kill invading larvae , whereas in resistant snails developing larvae are rapidly encapsulated by hemocytes and killed within 24–48 hours of infection [6–8] . Hemocyte larvicidal activity has been linked to the production and release of reactive oxygen species ( ROS ) , mainly hydrogen peroxide ( H2O2 ) , and the reactive nitrogen species , nitric oxide [9 , 10] . Although hemocytes of both resistant and susceptible B . glabrata strains produce H2O2 , resistant hemocytes generate and release higher levels than susceptible cells [11] , and this production appears to depend on the extracellular signal regulated protein kinase ( Erk ) [12] . However , a critical question arising from these observations is what are the signaling mechanisms that regulate ROS responses ? A critical period of larval development in the snail host is 24–48 hours post-infection , when the newly invading miracidium completes its transformation to the primary sporocyst stage . Larval killing depends on the ability of circulating hemocytes to recognize and encapsulate the newly formed sporocyst [4 , 13–15] . Among various sporocyst factors that may be contributing to hemocyte reactivity are glycoproteins that are released during the miracidium-to-sporocyst transition . In vitro studies have shown that these larval transformation proteins ( LTPs ) [16] modulate phagocytic activity , motility , and ROS production in B . glabrata hemocytes [17–21] , and disrupt hemocyte immune signaling [22–24] . However , questions regarding specific mechanisms by which LTPs modulate hemocyte immune responses remain unanswered . For over four decades a cell line derived from embryos of a schistosome-susceptible strain of B . glabrata , the B . glabrata embryonic ( Bge ) cell line [25] , has served as an in vitro model for the study of larval schistosome-snail host interactions in schistosomiasis . Bge cells share many characteristics with B . glabrata hemocytes including their morphology , adhesive properties , phagocytic activity , and larval encapsulation response [26] . In fact , co-culture of Bge cells with S . mansoni larvae results in the development of the parasite from the miracidium to the final cercarial stage , similar to the development that occurs with susceptible B . glabrata strains [27–30] . We have therefore adopted Bge cells as an in vitro model system to study the molecular interactions between snail cells and S . mansoni LTP . Because ion channels in the plasma membrane of human immune cells , including eosinophils , macrophages , neutrophils and lymphocytes , play important roles in immune responses , often by regulating the production and release of ROS [31] , we explored the role ion channels may play in signaling and ROS production in Bge cells . Using the whole cell patch clamp technique , we discovered an LTP-sensitive H+ channel that serves as the dominant ion conductance of Bge cell membranes . In addition , using a fluorescent probe to measure intracellular ROS , we also found that this channel mediates the production of ROS , thus suggesting a possible function for H+ channels in snail immune responses .
The Bge cell line was originally obtained from American Type Culture Collection ( ATCC CRL 1494 ) and is currently available through the BEI Resources ( https://www . beiresources . org ) . Cells were maintained at 26°C under normoxic conditions in complete Bge ( c-Bge ) medium consisting of 22% Schneider’s Drosophila Medium , 0 . 45% lactalbumin enzymatic hydrolysate , and 7 . 2 mM galactose supplemented with 10% heat-inactivated fetal bovine serum and 1% penicillin/streptomycin [25 , 28] . Bge cells were passaged at 80% confluency . S . mansoni eggs were isolated , hatched , and miracidia cultured in vitro as previously described [28] . Approximately ~ 5000 miracidia/mL in Chernin’s balanced salt solution ( CBSS; 47 . 9 mM NaCl , 2 . 0 mM KCl , 0 . 5 mM Na2HPO4 , 0 . 6 mM NaHCO3 1 . 8 mM MgSO4 , 3 . 6 mM CaCl2 and pH 7 . 2 ) [32] supplemented with glucose ( 1 mg/mL ) , trehalose ( 1 mg/mL ) , penicillin G ( 100 units/mL ) and streptomycin sulfate ( 0 . 05 mg/mL ) adjusted to pH 7 . 2 ( CBSS+ ) were then plated in a 24-well tissue culture plate and incubated at 26°C under normal atmospheric conditions to allow in vitro transformation of miracidia to primary sporocysts . The LTP-containing culture medium was collected after 48 hr , and the newly transformed primary sporocysts were washed once with CBSS+ . The LTP and CBSS+ wash were combined , filtered with a 0 . 45 μm Nalgene syringe filter ( Thermo Scientific , Waltham , MA ) , and concentrated using 3 kDa molecular weight cut-off ultrafiltration tubes ( Amicon Ultra Centricon , Billerica , MA ) . A NanoDrop ND-1000 spectrophotometer ( NanoDrop Technologies , Wilmington , DE ) was used to determine the protein concentration , after which a protease inhibitor cocktail ( Calbiochem , Billerica , MA ) was added . Multiple collections of LTP were pooled and stored in aliquots at -20°C . To denature LTP , pools were boiled at 100°C for 5 min . Bge cells ( ~4 x 106 ) were plated in 60x15 mm petri dishes in c-Bge medium , and allowed to attach overnight . In order to make recordings under defined ionic conditions , cells were washed 3X with CBSS before recording and kept in this buffer during subsequent manipulations . In experiments involving the treatment of Bge cells with ZnCl2 , 10 mM HEPES replaced NaH2PO4 in CBSS due to the insolubility of Zn3 ( PO4 ) 2 . Adherent cells were viewed with an Axioskop microscope equipped with a 63X water-immersion objective ( Carl Zeiss , Thornwood , NY ) . Bge cells were imaged with a CCD camera and viewed on a monitor . Patch electrodes fabricated from borosilicate glass capillaries had resistances of 3–7 MΩ when filled with a solution containing ( in mM ) 60 K-gluconate , 1 CaCl2 , 1 MgCl2 , 1 Mg-ATP , 10 HEPES , and 5 EGTA . The bathing solution for recordings was a slightly modified version of CBSS consisting of ( in mM ) : 47 NaCl , 2 KCl , 0 . 5 NaH2PO4 , 0 . 6 NaHCO3 , 1 . 8 MgSO4 , 3 . 6 CaCl2 . The pH of the pipette solution and external CBSS was adjusted to 5 or 7 with KOH or HCl . Modified versions of the internal and external solutions are stated in the Results section where they are used . Pressure-ejection pipettes were modified patch electrodes with tip diameters of ~2 μm . A Picospritzer II ( General Valve Corp . ) was used to apply 5–10 PSI of pressure to ejection pipettes . Patch clamp recordings were made with an Axopatch 200B amplifier ( Molecular Devices , Sunnyvale , CA ) , with data read into a PC through a Digidata 1440 A interface . The computer program pClamp 10 ( Molecular Devices ) controlled data acquisition , voltage steps , and pressure application by the Picospritzer . Data were filtered with a low-pass Bessel filter at 2 kHz before digitization at 10 kHz . The fluorescent probe 2’7’-dichlorofluorescein-diacetate ( DCFH-DA; Sigma-Aldrich , St . Louis , MO ) was used to measure ROS production in Bge cells following a method described previously with hemocytes [33] . Bge cells ( ~1 . 5 x 105 ) in suspension were washed 3X with CBSS before incubation in CBSS ( control ) , CBSS containing either 30 μg/mL LTP , 1 mM ZnCl2 or 30 μg/mL LTP + 1 mM ZnCl2 for 1 hr at 26°C . After treatment , cells were washed 3X with CBSS and centrifuged at 1000 rpm for 10 min . The final cell pellets were then re-suspended in 150 μL of CBSS containing 10 μM DCFH-DA , and distributed in three wells of a 96-well black-walled plate ( BD Falcon ) . The oxidation of DCFH-DA to fluorescent 2’7’-dichlorofluorescein ( DCF ) was measured in triplicate at 10 min intervals for up to 60 min using a Bio-Tek Synergy fluorescence plate reader ( Winooski , VT ) with excitation and emission wavelengths of 485 ± 20 and 528 ± 20 , respectively . Data analysis was conducted with Origin software ( Microcal , Northhampton , MA , USA ) . Five independent replicates of each experiment were conducted , with the raw data presented as mean ± SEM , and ratios of means of treated groups to controls presented separately . For molecular analysis of H+ channels , the hydrogen voltage gated channel 1 ( HVCN1 ) gene was identified in the nonredundant NCBI database , and sequence comparisons were conducted with PCR products from Bge cells and B . glabrata hemocytes . Schistosome-susceptible ( NMRI ) and resistant ( BS-90 ) B . glabrata strains were maintained in laboratory colonies in 10-gallon aquaria at 26°C under 12:12 hr light/dark cycling . Hemolymph , containing hemocytes , was collected by headfoot retraction [34] and immediately transferred to Eppendorf tubes containing an equal volume of CBSS on ice . Hemocytes were then pelleted by centrifugation at 1000 RPM for 10 min and washed 3 times in CBSS . Bge cells , grown in a flask to ~80% confluency , were detached mechanically using a cell scrapper , transferred to a 15 mL conical tube and pelleted by centrifugation as described for hemocytes . Total RNA was extracted from Bge cells and hemocytes of both B . glabrata strains using TRIzol reagent . Normalized concentrations of isolated total RNA samples were subjected to cDNA synthesis reactions using the GoScriptTM Reverse Transcription System ( Promega Corp . , Madison , WI ) . The cDNA was then used as the template for PCR using primers for the B . glabrata voltage-gated H+ channel 1-like gene ( BgHVCN1-like; Forward 5’-TGCTATGGGCTTAGCTTACTTC-3’; Reverse 5’-ATGTAGGGTCTTCAAACCATTCT-3’ ) that were designed using the predicted mRNA sequence for the gene with the National Center for Biotechnology Information ( NCBI ) database ( Accession number XM_013231505 ) . The expected amplicon size is ~362 bp , ~65% of the coding DNA sequence . As a positive control , primers for B . glabrata α–tubulin ( Forward 5’ -GTGAGACTGGCTGTGGTAAA-3’; Reverse 5’ -GGGAAGTGAATCCTGGGATATG-3’ ) with Accession number XP_013094834 . 1 were used to amplify an expected product of ~643 bp . Gel electrophoresis of the PCR products was performed followed by Big Dye sequencing at the University of Wisconsin Biotechnology Center DNA Sequencing Facility ( Madison , WI ) . The resulting nucleotide sequences were used in a search using BLASTn search against the non-redundant nucleotide NCBI database to confirm that the PCR amplified product encoded an HVCN1-like protein . Patch clamp data were analyzed with Clampfit ( Molecular Devices , Sunnyvale , CA ) and Origin Pro ( Microcal , Northhampton , MA ) . One-way RM-ANOVA and post-hoc statistical analyses were conducted in Origin Pro to assess significance . Results are presented as means ± SEM . The asterisks ( * ) represent p < 0 . 05 in all figures .
Whole cell patch clamp recordings were made from Bge cells to explore their membrane properties . Voltage steps from -75 to 25 mV for 500 msec induced an outward current that activated rapidly and then weakly inactivated in ~10–20 msec before stabilizing ( Fig 1A , control trace , top ) . To identify the ions responsible for this current , we manipulated the composition of the recording solutions . When Cl- was replaced by gluconate in the internal and bathing solutions , voltage steps induced currents similar to those seen with control solutions ( Fig 1A , second trace from top ) . Further substitution of Cs+ for K+ in the internal solution reduced the current to roughly 68% of control currents ( Fig 1A , third trace from top ) . The mean peak and plateau current amplitudes for these solutions are shown in Fig 1B . For gluconate and Cs+ substitution , current amplitudes were not significantly different from the control . Thus , Cl- and K+ replacement experiments indicated that these are not major permeating ions . In addition , comparisons of the Nernst potentials ( equilibrium potential for each ion based on internal and external concentrations ) with reversal potentials in current-voltage relationships did not support channels selective for Na+ or Ca2+ ( Supplemental S1 Fig ) . These results suggested that the major ions in our recording solutions do not permeate the membranes of Bge cells . H+ channels play important roles in many types of immune cells [35] , so we explored the possibility that H+ channels reside in the membranes of Bge cells . Subjecting Bge cells to pH gradients ( by adjusting the pH of the pipette and bathing solutions–see Methods ) [36] altered the current elicited by voltage steps and shifted the relationship between current and voltage ( Fig 2 ) . A gradient of two pH units ( pH 5in/pH 7out ) reduced the current amplitude at all voltages and shifted the reversal potential in the plot of peak current versus voltage in the negative direction by 17 . 5 mV ( Fig 2B , dashed line ) . Reversing the pH gradient ( pH 5out/pH 7in ) shifted the peak current-voltage plot in the opposite direction with a positive shift in the reversal potential of 27 . 5 mV ( Fig 2B , dotted line ) . Plots of plateau current versus voltage showed similar shifts ( Supplemental S2 Fig ) . Table 1 presents the reversal potentials along with the Nernst potentials for H+ . The shifts are in the direction of the H+ Nernst potential but smaller in magnitude because the H+ concentration is much lower relative to the concentrations of other ions in the solutions . Channels permeable to other ions generally result in H+ current reversal potential shifts that are less than the change in the H+ Nernst potential [37] . The effects of pH gradients on membrane currents are consistent with the presence of an H+ channel in Bge cell membranes . As an additional test for the presence of H+ channels we applied the H+ channel blocker Zn2+ [36 , 38] . Pressure application of 1 mM ZnCl2 from a glass pipette onto a Bge cell significantly reduced both peak and plateau currents elicited by voltage steps from -50 to 20 mV ( Fig 3A ) . This blockade was reversible , as demonstrated by current recovery after ZnCl2 removal ( Fig 3A , wash trace ) . Time course plots in which ZnCl2 was perfused onto cells through the bathing medium showed a 3 . 5-fold reduction in current amplitude ( Fig 3B and 3C ) , from 621 ± 4 pA to 177 ± 1 pA ( N = 4 ) , supporting the presence of H+ channels in Bge cell membranes . Although other actions of Zn2+ cannot be ruled out , the block of membrane current is consistent with the presence of H+ channels in Bge cells . As larval schistosome proteins have been shown to modulate a variety of snail hemocyte immune functions [14 , 15] , we tested the effects of S . mansoni LTP on Bge cell membrane current . Pressure application of LTP onto Bge cells dramatically increased the peak and plateau currents evoked by steps from -50 mV to 20 mV ( Fig 4A ) . LTP increased the current significantly by over 2-fold ( 478 ± 6 pA ) compared to control ( 212 ± 4 pA ) , and this increase only partially reversed with a 17% decrease ( 397 ± 7 pA ) following removal of LTP . Recovery was slow , and 5 min after LTP removal the current had decreased only slightly ( Fig 4B and 4C ) . Plotting current versus time also illustrated the opposite effects of LTP and ZnCl2 on Bge cells ( Fig 4C ) . This plot showed a >2-fold increase in current amplitude in the presence of LTP ( Fig 4C blue circles ) and a >2-fold reduction in the presence of ZnCl2 ( Fig 4C , red triangles ) compared to control ( Fig 4C black squares ) . The reversal of block by ZnCl2 was rapid and essentially complete , while the reversal of enhancement by LTP was slow . Moreover , when heat-denatured LTP was pressure-applied onto Bge cells , we observed no significant change compared to control current amplitudes ( Fig 5C and 5D ) , indicating that the action of LTP on H+ channels depends on heat-labile factors . To determine whether LTP increased Bge cell membrane current by opening H+ channels , we applied LTP and ZnCl2 simultaneously , and observed no statistically significant change ( Fig 5 ) , indicating that ZnCl2 counters the effect of LTP . Finally , we noted that current-voltage curves shifted in the presence of LTP and ZnCl2; LTP caused a 9 mV right-shift from control , toward the H+ Nernst potential , while ZnCl2 caused a 23 mV left-shift , away from the H+ Nernst potential ( Fig 6 ) . These results are consistent with the blockade of H+ channels by ZnCl2 and enhancement of H+ channels by LTP . Because H+ channels contribute to ROS production in mammalian immune cells [35 , 39] , we measured the generation of ROS in Bge cells with the fluorescent probe 2’7’-dichlorofluorescein-diacetate ( DCFH-DA ) . We observed a rapid and robust fluorescence increase that reflects constitutive ROS production . ZnCl2 and LTP + ZnCl2 inhibited this activity by ~50% compared to the untreated control ( F3 , 16 = 24 . 26 , p < 0 . 05 ) . These results demonstrate a linkage between H+ channels and the production of ROS in Bge cells . LTP alone produced a small apparent increase in ROS production , but this increase was not statistically significant . This suggests that at control level of H+ current in Bge cells , other factors limit ROS production ( Fig 7A and 7B ) . To identify putative H+ channel proteins expressed by Bge cells and B . glabrata hemocytes we searched the B . glabrata genome ( https://www . vectorbase . org/organisms/biomphalaria-glabrata ) using Blastp for homologues of human HVCN1 protein . The closest match was an HVCN1-like protein ( BgHVCN1-like , Accession number . XM_013231505 ) with 31% identity to human HVCN1 . This sequence contained the motif RLWRVTR , which is consistent with the H+ channel consensus sequence RxWRxxR [36] . A segment of the predicted sequence was then used to design primers for polymerase chain reactions ( PCR ) . Using cDNA from Bge cells and B . glabrata hemocytes ( NMRI and BS-90 strains ) as templates , PCR using the primers stated in the Methods section yielded amplicons of similar size with 99% sequence identity ( E = 0 . 0 ) to the predicted B . glabrata HVCN1-like sequence ( Supplemental S3 Fig ) . The amplified products encode 120 amino acid stretch of the 186 residues predicted for molluscan HVCN1-like protein . These results indicate that mRNA with the predicted sequence for a BgHVCN1-like gene is present in both Bge cells and hemocytes .
This investigation revealed the presence of functional ion channels in Bge cell membranes . pH manipulations altered the voltage dependence of membrane currents in a manner consistent with a dominant H+ permeability . Since the H+ concentration was several orders of magnitude lower than the other ions in our solutions , even low permeabilities to other ions can make large contributions to the observed reversal potentials and move them away from the H+ Nernst potential . Thus , although currents did not reverse at the H+ Nernst potential , the shifts were in the appropriate direction and supported the hypothesis that H+ channels are the predominant ion permeability in the plasma membrane of Bge cells . We also found that the H+ channel blocker Zn2+ significantly reduced the current through Bge cell membranes , providing additional support for the presence of H+ channels . Finally , we identified and sequenced an HVCN1-like transcript expressed in both this snail cell line and B . glabrata hemocytes , suggesting a functional linkage between these cell types . Thus , three independent lines of evidence support the conclusion that Bge cells express functional H+ channels . With few exceptions [40] , previous studies focusing on ion channels in molluscs almost exclusively have involved neuronal cell systems and/or emphasized Na+ , K+ , Ca2+ or Cl- channel activities [41–43] . To our knowledge , this is the first report of a functional H+ channel in non-neuronal cells of freshwater gastropods . Similar to the well documented association between H+ channels and ROS production in mammalian immunocytes [39] , we also found that blockade of the H+ channel with Zn2+ significantly abrogated Bge cell ROS production , indicating a functional association between channel-mediated H+ flux across the membrane and the oxidative response . This finding is significant since the formation and release of several ROS , especially H2O2 , and RNS are known to be involved in the killing of larval S . mansoni by B . glabrata hemocytes [9 , 10] . It is possible that , as in mammalian immune cells [39 , 44] , changes in membrane potential associated with ROS production also require a compensatory activation of H+ channels to maintain pH balance in immunocyte-like molluscan cells . It is important to note that hemocytes from both resistant and susceptible strains of B . glabrata snails are capable of generating ROS [11 , 32] , but differ both qualitatively and quantitatively in their responses [11] . Since Bge cells were originally derived from a S . mansoni-susceptible Puerto Rican strain of B . glabrata [25] , it is likely that hemocytes from a related susceptible strain ( NMRI ) also share both molecular and functional similarities to Bge cells . These shared characteristic have been well-documented in previous studies [26 , 45 , 46] , supporting the use of this cell line as a hemocyte-like model , as well as a general model for Biomphalaria-schistosome interactions [29 , 47] . Based on the presence and expression of the HVCN1-like gene in B . glabrata hemocytes , it is quite possible that voltage-gated H+ channels are also involved in regulating cellular ROS production as demonstrated in Bge cells . Proteins released during the S . mansoni miracidium-to-sporocyst transformation ( LTP ) have been shown to modulate a variety of functions in both hemocytes and Bge cells [14 , 24 , 45] . Such a role is supported by our finding of an LTP-induced potentiation of H+ channel activity . Exposure to LTP elicited a rapid and sustained enhancement of Bge cell membrane current . Because the reversal potential moved toward the H+ Nernst potential , it is likely that LTP increased the current through H+ channels . This activity was heat-labile , suggesting that the channel-active LTP component ( s ) may be a protein ( s ) with irreversible or slowly reversing action . However , it remains unclear whether the regulation of Bge cell H+ channels by schistosome LTPs results from factors thought to play a role in host-parasite compatibility [48–50] or other , yet unidentified , larval factors . The H+ channel may play a role in co-evolutionary mechanisms , known to affect oxidant-antioxidant levels during parasite-host interaction [51] . Identifying the active components of LTP and determining whether this response reflects the action of a single or multiple species will require further investigation . Despite the channel stimulating action of LTP , LTP treatment of Bge cells resulted in no statistically significant increase in ROS production . These results are consistent with previous findings that exposure of B . glabrata hemocytes to excretory-secretory products of larval S . mansoni exerted little effect on the production of ROS [52] . However , the question remains as to why LTP-stimulated H+ channel activation failed to enhance ROS production . Based on the H+ current data , it might be speculated that LTP binding to Bge cells is linked to the opening of H+ channels through receptor-mediated activation of a channel-associated signaling pathway , possibly through interactions with pathogen recognition receptors such as fibrinogen-related proteins , Toll-like receptors , or bacterial binding proteins that have been implicated in B . glabrata immunity [50 , 53–55] . Mitogen-activated and extracellular-signal regulated protein kinases shown to function in molluscan immunity [12 , 22] could also play a role in signaling to the H+ channel . A final possibility is that LTP may be acting directly on the channel protein itself to induce opening . The consequence of H+ channel modulation would be an alteration or disruption of H+ ion balance and intracellular pH , but without stimulating ROS production . This may , in turn , serve as a potent anti-immune mechanism used by sporocysts for countering host ROS-mediated effector responses . Thus , H+ channels , while serving an important role in maintaining pH balance within Bge cells and hemocytes , may also be manipulated by schistosome larvae to reduce their immune efficacy . Since Bge cells were originally derived from a S . mansoni-susceptible PR albino strain of B . glabrata [25] , it is likely that hemocytes from a related susceptible strain ( NMRI ) also share sensitivity to H+ channel–reactive anti-immune proteins , thereby supporting a compatible snail-schistosome interaction . In conclusion , Bge cells possess a functional H+ channel that is responsible for a dominant conductance of their plasma membrane . ROS production is dependent on H+ channels . Exposure of cells to heat-labile LTP stimulates channel opening and H+ flux , but has little if any effect on the generation of ROS . Although H+ channels have not been tested directly in B . glabrata hemocytes , PCR amplification and amplicon sequencing demonstrated the presence of HVCN1-like transcripts in both susceptible and resistant B . glabrata strains . Thus , the association of the Bge cell H+ channel activity with cellular ROS production and the channel’s response to schistosome LTP suggest a role in regulating larval schistosome-snail interactions . Future identification of the specific mechanism ( s ) tying together these activities should provide important insights into host-parasite compatibility in this system . | Schistosoma mansoni is one of four major species of human blood flukes that , together , infect over 250 million people worldwide . Transmission of S . mansoni to humans requires infection of freshwater intermediate host snails , Biomphalaria spp . , in order to complete its life cycle . The B . glabrata embryonic ( Bge ) cell line , derived from a Puerto Rican strain of snail host shares characteristics with circulating hemocytes , the molluscan immune cells , and serves as an in vitro model for snail immune function . Electrical recordings from Bge cells demonstrated the presence of H+ channels that allow hydrogen ions ( H+ ) to cross the membrane . Furthermore , blocking these channels inhibited the production of reactive oxygen species ( ROS ) , an immune defense mechanism shared by Bge cells and hemocytes . Interestingly , Bge cell exposure to proteins produced by S . mansoni larvae exerted the opposite effect , enhancing H+ movement across the cell membrane . An H+ channel-encoding gene was expressed in both Bge cells and hemocytes suggesting that hemocytes may share similar functions with Bge cells . | [
"Abstract",
"Introduction",
"Materials",
"and",
"methods",
"Results",
"Discussion"
] | [
"blood",
"cells",
"schistosoma",
"invertebrates",
"schistosoma",
"mansoni",
"medicine",
"and",
"health",
"sciences",
"reactive",
"oxygen",
"species",
"immune",
"cells",
"pathology",
"and",
"laboratory",
"medicine",
"helminths",
"immunology",
"membrane",
"potential",
"electrophysiology",
"animals",
"molecular",
"biology",
"techniques",
"cellular",
"structures",
"and",
"organelles",
"research",
"and",
"analysis",
"methods",
"white",
"blood",
"cells",
"artificial",
"gene",
"amplification",
"and",
"extension",
"animal",
"cells",
"pathogenesis",
"molecular",
"biology",
"cell",
"membranes",
"immune",
"response",
"hemocytes",
"biochemistry",
"cell",
"biology",
"host-pathogen",
"interactions",
"polymerase",
"chain",
"reaction",
"physiology",
"biology",
"and",
"life",
"sciences",
"cellular",
"types",
"oxidative",
"damage",
"organisms"
] | 2017 | H+ channels in embryonic Biomphalaria glabrata cell membranes: Putative roles in snail host-schistosome interactions |
Endogenous small RNAs ( sRNAs ) and Argonaute proteins are ubiquitous regulators of gene expression in germline and somatic tissues . sRNA-Argonaute complexes are often expressed in gametes and are consequently inherited by the next generation upon fertilization . In Caenorhabditis elegans , 26G-RNAs are primary endogenous sRNAs that trigger the expression of downstream secondary sRNAs . Two subpopulations of 26G-RNAs exist , each of which displaying strongly compartmentalized expression: one is expressed in the spermatogenic gonad and associates with the Argonautes ALG-3/4; plus another expressed in oocytes and in embryos , which associates with the Argonaute ERGO-1 . The determinants and dynamics of gene silencing elicited by 26G-RNAs are largely unknown . Here , we provide diverse new insights into these endogenous sRNA pathways of C . elegans . Using genetics and deep sequencing , we dissect a maternal effect of the ERGO-1 branch of the 26G-RNA pathway . We find that maternal primary sRNAs can trigger the production of zygotic secondary sRNAs that are able to silence targets , even in the absence of zygotic primary triggers . Thus , the interaction of maternal and zygotic sRNA populations , assures target gene silencing throughout animal development . Furthermore , we explore other facets of 26G-RNA biology related to the ALG-3/4 branch . We find that sRNA abundance , sRNA pattern of origin and the 3’ UTR length of target transcripts are predictors of the regulatory outcome by the Argonautes ALG-3/4 . Lastly , we provide evidence suggesting that ALG-3 and ALG-4 regulate their own mRNAs in a negative feedback loop . Altogether , we provide several new regulatory insights on the dynamics , target regulation and self-regulation of the endogenous RNAi pathways of C . elegans .
A plethora of pathways based on non-coding small RNAs ( sRNAs ) regulates gene expression in every domain of life . These are collectively known as RNA interference ( RNAi ) or RNAi-like pathways . In invertebrates , which lack adaptive immune systems and interferon response , RNAi-like pathways fulfill an immune role at the nucleic acid level , by controlling viruses and transposable elements ( TEs ) . MicroRNA ( miRNA ) , Piwi-interacting RNA ( piRNA ) and endogenous small interfering RNA ( endo-siRNA ) pathways are the better described RNAi-like pathways , which differ in their biogenesis and specialized cofactors . MicroRNAs are commonly found in many , if not all , tissues and broadly regulate gene expression throughout development [1] . piRNAs are typically , but not exclusively , expressed in the metazoan germline , where they assume a central function in TE control [2–5] . Endo-siRNA pathways comprise varied classes of sRNAs expressed in the soma and germline that can regulate the expression of TEs and protein-coding genes [6–8] . A key commonality of RNAi-like pathways is the participation of Argonaute proteins . These proteins directly associate with sRNAs and Argonaute-sRNA complexes engage transcripts with sequence complementarity , typically resulting in target silencing . sRNA-directed gene silencing can occur both on the post-transcriptional level , by target RNA cleavage and degradation , and/or on the transcriptional level , via nuclear Argonautes that direct heterochromatin formation at target loci . sRNAs can be viewed as genome guardians against “foreign” nucleic acids [9] . In this light , the germline is an important tissue for sRNA production and function to control the transmission of “non-self” genetic elements to progeny . In multiple animals , Piwi-piRNA complexes have been shown to be maternally deposited into zygotes , where they may initiate TE silencing [10–19] . Endo-siRNAs are abundantly expressed in gametes , being often required to successfully complete gametogenesis . These may also be deposited into embryos and have roles in setting up gene expression in the next generation . For example in plants , TE-derived endo-siRNAs are abundant in male and female gametes [20] . Moreover , endo-siRNAs are expressed in Drosophila ovaries [21] and in mouse oocytes [22 , 23] to regulate protein-coding genes and TEs . Overall , gamete expression and maternal inheritance of Argonaute-sRNA complexes seem to be a widespread phenomenon in plants and animals , presumably important to tune gene expression during early development . RNAi was first identified in the nematode Caenorhabditis elegans [24] . Ever since , C . elegans has continuously been an important and fascinating model for studies on RNAi . C . elegans has an unprecedented 27 genomically encoded Argonaute genes , including a whole worm-specific clade of the Argonaute protein family [25] . Several sRNA species have been identified in worms: miRNAs , 21U-RNAs , 22G- and 26G-RNAs [26 , 27] . 21U-RNAs associate with PRG-1 , a Piwi class Argonaute , in the germline and are therefore considered the piRNAs of C . elegans [28–30] . 26G-RNAs can be considered primary endo-siRNAs , in that they elicit production of the overall more abundant secondary endo-siRNA pool , termed 22G-RNAs [31–33] . 26G-RNAs are produced by the RNA-dependent RNA Polymerase ( RdRP ) RRF-3 [31–35] . The ERI complex ( ERIC ) is an accessory complex that assists RRF-3 in producing 26G-RNAs [36–39] . The conserved CHHC zinc finger protein GTSF-1 and the Tudor domain protein ERI-5 form a pre-complex with RRF-3 that is responsible for tethering the RdRP to the ERIC [36 , 39] . Two distinct subpopulations of 26G-RNAs are synthesized in the germline and in embryos . One subpopulation is produced in the spermatogenic gonad in L4 hermaphrodites and in the male gonad , where they associate with the redundantly acting paralog Argonautes ALG-3 and ALG-4 ( henceforth referred to as ALG-3/4 ) [31 , 34 , 35 , 38] . These 26G-RNAs trigger the biogenesis of secondary 22G-RNAs that have been shown to either promote gene expression through the Argonaute CSR-1 or to inhibit gene expression through unidentified WAGO proteins [31 , 40] . Hence , the effects of ALG-3/4-dependent sRNAs on their targets is complex: while some targets appear to be silenced , the expression of others seems to be positively affected . The regulatory effects resulting of the combined action of ALG-3/4 and CSR-1 seem to be more physiologically relevant at elevated temperatures [40] . The conditions determining regulatory outcome , either silencing or licensing , are still unclear . In the oogenic hermaphrodite gonad and in embryos another subpopulation of 26G-RNAs is produced . These are 3’ 2’-O-methylated by the conserved RNA methyltransferase HENN-1 [41–43] and bind to the Argonaute ERGO-1 [33] . ERGO-1 targets pseudogenes , recently duplicated genes and long non-coding RNAs ( lncRNAs ) [33 , 36 , 44] . It has recently been shown that these targets generally have a small number of introns that lack optimal splicing signals [45] . ERGO-1 may thus serve as a surveillance platform to silence these inefficient transcripts , preventing detrimental accumulation of stalled spliceosomes . Effective silencing of these genes is achieved by secondary 22G-RNAs produced after ERGO-1 target recognition [32 , 33] . In turn , these secondary 22G-RNAs may associate with cytoplasmic Argonautes that mediate post-transcriptional gene silencing [33] , or to the Argonaute NRDE-3 , which is shuttled into the nucleus and further silences its targets on the transcriptional level [46 , 47] . Depletion of spermatogenic 26G-RNAs , for example in rrf-3 , gtsf-1 and alg-3/4 mutants , results in a range of sperm-derived fertility defects including complete sterility at higher temperatures [31 , 34–38] . The elimination of oogenic/embryonic 26G-RNAs , for example by impairment of rrf-3 , gtsf-1 and ergo-1 , gives rise to an Enhanced RNAi ( Eri ) phenotype , characterized by a response to exogenous dsRNA that is stronger than in wild-type [25 , 36–38] . This phenotype is thought to reflect competition for common factors between exogenous and endogenous RNAi pathways [37 , 48] . However , the Eri phenotype lacks characterization on the molecular level . Furthermore , a strong maternal rescue was reported for Eri factors [49] , suggesting that maternally deposited Eri factors or their dependent sRNAs have an important role in maintaining gene silencing . The basis for this maternal rescue was not further characterized . In this work , we address a number of gene regulatory aspects of the 26G-RNA pathways in C . elegans . First , we genetically dissect a maternal effect displayed by the ERGO-1 branch of the 26G-RNA pathway . Our findings suggest that both maternal and zygotic sRNAs drive gene silencing throughout embryogenesis and larval development until adulthood . Moreover , we interrogate a number of aspects on gene regulation in the ALG-3/4 branch of the 26G-RNA pathway . We report that sRNA abundance , origin of the sRNAs and 3’ UTR length of target transcripts are predictors of the regulatory outcome of ALG-3/4 targets . Lastly , we find that the 26G-RNA-binding Argonautes ALG-3 and ALG-4 may regulate their own expression in a negative feedback mechanism .
rrf-3 and gtsf-1 mutants lack the two subpopulations of 26G-RNAs and display the phenotypes associated with depletion of both subpopulations: the enhanced RNAi ( Eri ) phenotype , shared with ergo-1 mutants [25 , 36–38] , and sperm-derived fertility defects , shared with alg-3/4 double mutants [31 , 34–38 , 40] . S1A Fig offers a simplified scheme of these pathways . For clarity , the two subpopulations of 26G-RNAs and downstream 22G-RNAs , dependent on ERGO-1 or ALG-3/4 will be referred to as ERGO-1 branch sRNAs and ALG-3/4 branch sRNAs , respectively . We have previously shown that germline-specific GTSF-1 transgenes could rescue the enhanced RNAi ( Eri ) phenotype of gtsf-1 mutants [36] . This was an intriguing result , since the Eri phenotype arises after targeting somatically expressed genes with RNAi , indicating that germline-expressed GTSF-1 is able to affect RNAi in the soma , possibly through maternal deposition of GTSF-1 or GTSF-1-dependent sRNAs . We reasoned that if maternal GTSF-1 activity can prime gene silencing in embryos then the transmission of the Eri phenotype should show a maternal rescue . To address this experimentally , we linked gtsf-1 ( xf43 ) to dpy-4 ( e1166 ) and crossed the resulting double mutants with wild-type males ( Fig 1A ) . We then allowed for two generations of heterozygosity and assayed for RNAi sensitivity in homozygous gtsf-1 mutant F1 and F2 generations , scoring for larval arrest triggered by lir-1 RNAi . Indeed , the Eri phenotype showed a strong maternal effect , arising only in the F2 generation of gtsf-1 mutants ( Fig 1A ) . This is consistent with a maternal effect reported for other Eri factors [49] . We have previously shown that GTSF-1 is required to silence a GFP transgene reporting on ERGO-1 branch 22G-RNA activity , referred to as 22G sensor [36 , 43] . Therefore , we also looked at the dynamics of derepression of this transgene upon introduction of gtsf-1 mutation . We noticed that strong GFP expression appeared only in the second generation of homozygosity of the gtsf-1 allele ( Fig 1B and 1C ) . An identical maternal effect on the expression status of this transgene is observed after crossing in rrf-3 , ergo-1 and other gtsf-1 mutant alleles ( S1B Fig ) . Combined with our previously described rescue of the Eri phenotype using a germline promoter , these results strongly suggest that maternally provided ERGO-1 branch pathway components are sufficient to establish normal RNAi sensitivity in the soma of C . elegans . Although the silencing of the 22G sensor used in our experiments is dependent on ERGO-1 , ERGO-1 is not the Argonaute protein binding to the effector 22G-RNA [33 , 43] . This has been shown to be driven by the somatically expressed , nuclear Argonaute protein NRDE-3 [44 , 46] and maybe additional cytoplasmic WAGOs [33] ( S1A Fig ) . In absence of ERGO-1 and other 26G-RNA pathway factors , NRDE-3 is no longer nuclear , and in nrde-3 mutants the 22G sensor is activated , indicating that NRDE-3 requires sRNA input from ERGO-1 branch sRNAs [36 , 43 , 44 , 46] . Strikingly , loss of NRDE-3 derepressed the 22G sensor transgene in the first homozygous generation ( Fig 1D ) , showing that in contrast to 26G-RNAs , the downstream 22G-RNA pathway is not maternally provided . MUT-16 is a factor required for the nucleation of mutator foci and 22G-RNA biogenesis [50] . Confirming the requirement for zygotically produced 22G-RNAs , absence of MUT-16 derepresses the 22G sensor in the first homozygous mutant generation ( Fig 1E ) . These results suggest a scenario in which 1 ) NRDE-3 is loaded with zygotically produced 22G-RNAs that are primed by maternally provided 26G-RNAs and 2 ) NRDE-3 activity is maintained in somatic tissues until the adult stage , in absence of a zygotic 26G-RNA pathway . The results presented above show that maternal 26G-RNAs are sufficient for 22G sensor silencing . We also tested whether maternal 26G-RNAs are necessary for 22G sensor silencing by crossing rrf-3 mutant males with gtsf-1; 22G sensor hermaphrodites ( Fig 1F ) . Both of these strains lack 26G-RNAs and their downstream 22G-RNAs , therefore , their progeny will not receive a maternal and/or paternal complement of these sRNAs . The 22G sensor was silenced in all cross progeny , showing that in the absence of maternal 26G-RNAs , zygotic 26G-RNAs can induce production of silencing-competent 22G-RNAs . Thus , maternal 26G-RNAs appear to be sufficient but not necessary for target silencing . The maternal effects described above for the Eri phenotype and for 22G sensor silencing are related to the ERGO-1 branch of the pathway . Next , we wanted to determine if the ALG-3/4 branch also displays such a parental effect . To test this , we assessed the influence of maternal GTSF-1 activity on the temperature-sensitive sterility phenotype . Using the same setup as we used for the Eri experiment ( in Fig 1A ) , we observe that the temperature-sensitive sperm defect of gtsf-1 mutants was not rescued maternally ( S1C Fig ) . Given that the ALG-3/4 branch of the 26G-RNA pathway is mostly active during spermatogenesis , next we asked whether a paternal effect is observed for the temperature-sensitive sperm defect . As shown in S1D Fig , we did not detect any evidence supporting a paternal effect . Overall , these results indicate that 26G-RNA-derived parental effects are likely restricted to the ERGO-1 branch . The 22G sensor reports on the silencing activity of a single 22G-RNA that maps to the so-called X-cluster , a known set of targets of ERGO-1 [33 , 43] . Therefore , the experiments above using this 22G sensor have a limited resolution and our observations may not reflect the silencing status of most ERGO-1 targets . To characterize this maternal effect in more detail and in a broader set of ERGO-1 targets , we decided to analyze sRNA populations in young adult animals . Concretely , we outcrossed dpy-4; gtsf-1 and sequenced sRNAs from wild-type and two consecutive generations of Dpy young adult animals ( Fig 2A ) . First generation gtsf-1 homozygous mutants will henceforth be addressed as “mutant F1” and second generation gtsf-1 homozygous mutants as “mutant F2” ( Fig 2A ) . We sequenced young adult animals because they lack embryos , therefore avoiding confounding effects with zygotic sRNAs of the next generation . sRNAs were cloned and sequenced from four biological replicates . The cloning of sRNAs was done either directly ( henceforth referred to as untreated samples ) or after treatment with the pyrophosphatase RppH [51] before library preparation . The latter enriches for 22G-RNA species that bear a 5’ triphosphate group . Sequenced sRNAs were normalized to all mapped reads excluding structural reads ( S1 Table ) . In our analysis we strictly looked at 26G- and 22G-RNAs that map in antisense orientation to protein-coding and non-coding genes ( see Methods ) . Total 26G-RNA levels are depleted in young adults lacking GTSF-1 ( Fig 2B ) . Mutant F1s have significantly less 26G-RNAs than wild-type worms , while mutant F2s have 26G-RNA levels very close to zero ( Fig 2B ) . For a finer analysis we looked specifically at 26G-RNAs derived from ERGO-1 and ALG-3/4 targets ( as defined in reference 36 , see Methods ) . 26G-RNAs mapping to these two sets of targets recapitulate the pattern observed for global 26G-RNAs ( Fig 2C and 2D ) . The difference between the F1 and F2 mutants might reflect a maternal 26G-RNA pool that is still detectable in the young adult F1 , but no longer in the F2 . However , we point out that amongst the selected F1 Dpy animals , approximately 5 . 2% will in fact be gtsf-1 heterozygous , due to meiotic recombination between gtsf-1 and dpy-4 ( estimated genetic distance between these two genes is 2 . 6 map units ) . Hence , another explanation for the mutant F1 pool of 26G-RNAs may be a contamination of the gtsf-1 homozygous pool with heterozygous animals . The mutant F2 was isolated from genotyped F1 animals , excluding this confounding effect . We conclude that in young adult mutant F1 animals , maternally provided 26G-RNAs ( or 26G-RNAs produced zygotically by maternal proteins ) are no longer detectable at significant levels . Total levels of 22G-RNAs are slightly reduced in mutant F1 and F2 animals ( Fig 2E ) . However , total 22G-RNA levels encompass several distinct subpopulations of 22G-RNAs , including those that do not depend on 26G-RNAs . To have a closer look on 22G-RNAs that are dependent on 26G-RNAs , we focused on 22G-RNAs that map to ERGO-1 and ALG-3/4 targets . Strikingly , compared to wild-type , the 22G-RNA population from ERGO-1 targets is moderately higher in mutant F1 animals and is subsequently depleted in the mutant F2 generation ( Figs 2F and S2A ) . These effects are not only clear in overall analysis , but also on a well-established set of ERGO-1 branch targets , such as the X-cluster ( Fig 2G ) . Consistent with a role of NRDE-3 downstream of ERGO-1 , 22G-RNAs mapping to annotated NRDE-3 targets [47] show the same pattern of depletion as ERGO-1-dependent 22G-RNAs ( S2A Fig ) . These results are consistent with the idea that the Eri phenotype and 22G sensor derepression are caused by the absence of NRDE-3-bound , secondary 22G-RNAs downstream of 26G-RNAs . 22G-RNAs mapping to ALG-3/4 targets behave differently in this experiment ( Fig 2H ) . Upon disruption of gtsf-1 , these 22G-RNAs are only slightly affected in both the mutant F1 and F2 ( Figs 2H and S2A ) , despite the fact that their upstream 26G-RNAs are absent . This is illustrated in S2B Fig with genome browser tracks of ssp-16 , a known ALG-3/4 target . We conclude that 26G-RNA-independent mechanisms are in place to drive 22G-RNA production from these genes . Finally , 21U-RNAs and 22G-RNAs mapping to other known RNAi targets are not affected in this inheritance setup , supporting the notion that gtsf-1 is not affecting these sRNA species ( S2A and S2C Fig ) . One exception are the 22G-RNAs from CSR-1 targets , which seem to be slightly depleted in both the mutant F1 and F2 generations ( S2A Fig ) . It is not possible to dissect whether this is a direct effect or not , but we note that mRNA levels of CSR-1 targets are slightly downregulated in the analyzed mutants ( S2D Fig ) . Given that CSR-1 22G-RNAs tend to correlate positively with gene expression [52] , it is conceivable that the reduction of CSR-1 target 22G-RNAs is the result of decreased target gene expression . The very same samples used for generating sRNA sequencing data were also used for mRNA sequencing ( Fig 2A and S1 Table ) . First , we checked gtsf-1 expression . As expected , gtsf-1 is strongly depleted in the mutant samples ( S3A Fig ) . In the mutant F1 we still observe a low level of gtsf-1 derived transcripts ( about 9 . 5% of wild-type ) that is absent from the mutant F2 . These transcripts cover the region deleted in the gtsf-1 ( xf43 ) mutant allele , indicating they cannot represent zygotically transcribed gtsf-1 mutant mRNA . Rather , these transcripts likely originate from the above-described contamination of the homozygous F1 population with heterozygous animals . We hypothesized that ERGO-1 branch 22G-RNAs observed in the mutant F1 generation might be competent to maintain target silencing . If this is true , we should observe strong upregulation of ERGO-1 target mRNAs only the mutant F2 generation . Indeed , the X-cluster is upregulated only in the second mutant generation ( Fig 3A ) . When ERGO-1 targets are analyzed in bulk , we observe the same trend , with stronger upregulation only in the mutant F2 , consistent with the maternal effect ( Fig 3B ) . Regarding the mutant F1 , we note that the very slight , not statistically significant , upregulation of ERGO-1 target mRNAs may account for the slight increase of 22G-RNAs observed in the mutant F1 ( in Figs 2F and S2A ) , because of an increased number of molecules available to template RdRP activity . ALG-3/4 targets , as for instance ssp-16 , were found to be upregulated already in the F1 generation ( Figs 3B and S3B ) , supporting the notion that the maternal rescue of the 26G-RNA pathways is restricted to the ERGO-1 branch . ERGO-1 targets comprise a very diverse set of targets consisting of pseudogenes , fast evolving small genes , paralog genes and lncRNAs [33 , 44 , 45] . Considering the maternal effect described above for ERGO-1-dependent sRNA and corresponding target , we postulated that this maternal effect may exist to counteract embryonic expression of ERGO-1 targets . To address this we sequenced mRNA of synchronized populations of all developmental stages ( L1 , L2 , L3 , L4 , young adult and embryos ) of both wild-type ( N2 ) and rrf-3 ( pk1426 ) mutants ( S1 Table ) . In wild-type worms , ERGO-1 targets are most abundant in embryos ( Fig 3C , lower panel , in blue ) . Moreover , the effect of rrf-3 mutation on ERGO-1 target expression is stronger in embryos ( Fig 3C , lower panel ) . These results indicate that the maternal effect reported above can reflect deposition of factors which are required to initiate silencing of targets early in development . Differential gene expression data and normalized read counts calculated from the sequencing datasets described in Figs 2 and 3 , can be found in S2 and S3 Tables . The young adult sequencing datasets we obtained in this study ( Fig 2A ) , as well as previous datasets of gravid adults [36] , are not well suited to address ALG-3/4 biology , considering that in these developmental stages ALG-3/4 are not expressed , at least not abundantly . Therefore , in order to further our understanding of the dependency of ALG-3/4 branch sRNAs on GTSF-1 , as well as to explore the regulation of ALG-3/4 targets , we generated additional sRNA and mRNA datasets from wild-type and gtsf-1 male animals grown at 20°C ( S1 Table ) . As expected , global 26G-RNA levels are severely affected in gtsf-1 mutant males , reflecting downregulation of 26G-RNAs from both branches of the pathway ( Fig 4A–4C ) . Consistent with the absence of ERGO-1 in adult males , ERGO-1 branch 26G-RNAs are detected in extremely low numbers in wild-type animals ( Fig 4B ) . Global levels of 21U-RNAs seem to be moderately increased ( Fig 4D ) , possibly resulting from the lack of 26G-RNAs in the libraries . Global levels of 22G-RNAs are not affected ( Fig 4E ) , but consistent with a global depletion of 26G-RNAs , 22G-RNAs specifically mapping to ALG-3/4 and ERGO-1 targets are reduced in gtsf-1 mutant males ( Fig 4F ) . Next , we probed the effects of gtsf-1 mutation on male gene expression using mRNA sequencing . ALG-3/4 and ERGO-1 targets are both upregulated in gtsf-1 mutant males ( Fig 4G ) . These changes are illustrated for the X-cluster and ssp-16 in the genome browser tracks of S4 Fig . S4 Table includes differential gene expression data and normalized read counts calculated from the sequencing datasets described in Fig 4 . As a final note on the developmental aspects of ALG-3/4 branch , consistent with enrichment in the spermatogenic gonad [31 , 34–36 , 38 , 40] , ALG-3/4 targets are more highly expressed and more responsive to rrf-3 mutation in the L4 and young adult stages of hermaphrodite animals ( Fig 3C , upper panel ) . Given that the overall ALG-3/4 target mRNA levels go up upon depletion of gtsf-1 or rrf-3 ( Figs 3C and 4G ) , bulk 26G-RNA activity during spermatogenesis seems to be repressive at 20°C . We conclude that the activity of GTSF-1 is required in adult males for silencing of 26G-RNA targets by participating in 26G- and 22G-RNA biogenesis . ALG-3/4 were shown to have distinct effects on gene expression , either silencing or licensing [31 , 40] . However , how these different effects arise is currently unknown . Even though our analysis in males did not reveal a licensing effect of 26G-RNAs , the bulk analysis of targets in Figs 3B and 4G may mask the behavior of distinct target subpopulations . Of note , our sequencing datasets were obtained from animals grown at 20°C and are therefore blind to the strong positive regulatory effect of ALG-3/4 in gene expression at higher temperatures [40] . We reasoned that sRNA abundance may be correlated with different regulatory outcomes . Therefore , we defined ALG-3/4 targets that are upregulated , downregulated , and unaltered upon gtsf-1 mutation and plotted their 26G-RNA abundance . This reveals a tendency for genes that are upregulated upon loss of GTSF-1 to be more heavily targeted by 26G-RNAs in adult males ( Fig 5A , left panel ) . The same trend is observed for 22G-RNAs: upregulated genes are more heavily covered by 22G-RNAs ( Fig 5A , right panel , and 5B ) . In contrast , ALG-3/4 targets that are downregulated in gtsf-1 mutant males display a relatively low-level targeting by 22G-RNAs ( Fig 5A and 5B ) . We conclude that , in adult males , stronger 26G-RNA targeting promotes stronger 22G-RNA biogenesis and repression of targets , whereas low-level targeting by 26G- and 22G-RNAs does not . Transcripts that are downregulated in absence of GTSF-1 might be licensed for gene expression , but may also respond in a secondary manner to a disturbed 26G-RNA pathway . It was previously noticed that ALG-3/4-dependent 26G-RNAs mostly map to both the 5’ and 3’ ends of their targets , and that this may correlate with gene expression changes [31] . We followed up on this observation by performing metagene analysis of 26G-RNA binding using our broader set of targets . Indeed , ALG-3/4 branch 26G-RNAs display a distinctive pattern with two sharp peaks near the transcription start site ( TSS ) and transcription end site ( TES ) ( Figs 6A and S5A , left panels ) . In contrast , ERGO-1 branch 26G-RNAs map throughout the transcript , with a slight enrichment in the 3’ half ( Fig 6B , left panel ) . Contrary to 26G-RNAs , 22G-RNAs from both branches map throughout the transcript ( Figs 6A and 6B and S5A , right panels ) . These patterns are consistent with recruitment of RdRPs and production of antisense sRNAs along the full length of the transcript . These findings suggest substantially different regulation modes by ERGO-1- and ALG-3/4-branch 26G-RNAs . Conine and colleagues reported a correlation between 26G-RNA 5’ targeting and negative regulation [31] . We wanted to address whether our datasets show concrete correlations between the patterns of origin of ALG-3/4-dependent 26G-RNAs and distinct regulatory outcomes . To address this , we ranked genes by 5’ and 3’ abundance of 26G-RNAs , selected genes predominantly targeted at the 5’ or at the 3' ends and plotted their fold change upon gtsf-1 mutation . In adult males , dominant 5’ targeting by 26G-RNAs seems to be correlated with gene silencing ( fold change >0 in the mutant , Fig 6C ) , whereas dominant 3’ targeting is accompanied with only weak upregulation and in some cases very mild downregulation ( Fig 6C and 6D ) . In further support for a non-gene silencing , and potentially licensing role for ALG-3/4 targeting at the 3’ end , genes with predominant 3’ 26G-RNAs display an overall higher expression than genes predominantly targeted at the 5’ region ( Fig 6E ) . The same signatures are found in young adults , with an even stronger signature of the 3’ in promoting gene expression ( S5B , S5C and S5D Fig ) . Finally , we interrogated if the length of 5’ and 3’ UTRs may be a predictor of regulatory outcome in ALG-3/4 targets . 5’ UTR length was not significantly different between unchanged and upregulated genes ( Fig 6F , left panel ) . Downregulated genes have a statistically significant shorter 5’ UTR length , but these results should be interpreted with caution due to the low number of transcript isoforms analyzed . In contrast , 3’ UTR length is significantly smaller in targets that respond to loss of GTSF-1 in males ( Fig 6F , right panel ) . Interestingly , we find the same and possibly even stronger relation between 3’UTR length and responsiveness to GTSF-1 status in young adult animals ( S5E Fig ) . Upregulated ERGO-1 targets do not display significantly shorter 3’ UTRs ( S5F Fig ) , indicating that the trend is specific to ALG-3/4 targets . Altogether , our results suggest that , both in males and young adult hermaphrodites , 3’ vs 5’ targeting and 3’ UTR length are predictors of whether ALG-3/4 targets are silenced or not . While navigating the lists of GTSF-1 targets defined by differential gene expression analysis , we noticed that alg-3 and alg-4 are targets of 26G-RNAs ( this study and in reference 36 ) . These 26G-RNAs are sensitive to oxidation ( not enriched in oxidized libraries , see reference 36 ) and map predominantly to the extremities of the transcript ( Figs 7 , upper panels ) , indicating that these 26G-RNAs share features with ALG-3/4 branch 26G-RNAs . In addition to these 26G-RNAs , significant amounts of 22G-RNAs are found on alg-3/4 ( Fig 7 , middle panels ) . These sRNAs seem to silence gene expression , since mRNA-seq shows that alg-3 and alg-4 transcripts are 2–3 fold upregulated in gtsf-1 mutants ( Fig 7 , lower panels ) . These results strongly suggest that alg-3/4 are regulating their own expression in a negative feedback loop . Of note , the upregulation of alg-3 and alg-4 is in agreement with the results presented above , because these genes are more heavily targeted by 26G-RNAs at their 5’ ( although alg-4 also has a sharp 3’ 26G-RNA peak , upper panels ) . Furthermore , these same signatures of negative feedback loop are observed in young adults ( S6 Fig ) .
Animal male and female gametes are rich in RNA . Upon fertilization , several RNA species are thus provided to the zygote . Multiple lines of evidence from several distinct organisms indicate that sRNAs are included in the parental repertoire of inherited RNA . For example , piRNAs have been reported to be maternally deposited in embryos in arthropods , fish and C . elegans [10–19 , 53 , 54] . In C . elegans other endogenous sRNA populations have also been shown to be contributed by the gametes: 1 ) 26G-RNAs have been shown to be weakly provided by the male , while 22G-RNAs are more abundantly provided [55]; 2 ) 26G-RNAs and the Argonaute ERGO-1 are co-expressed during oogenesis and in embryos [33 , 35 , 56]; and 3 ) 22G-RNAs are deposited in embryos via the mother and participate in transgenerational gene silencing [53 , 57–61] . We describe a maternal effect in the transmission of the Eri phenotype and 22G sensor derepression and characterize the subjacent dynamics of sRNAs and mRNA targets ( Figs 1–3 and S1–S3 ) . We show that both maternal and zygotic 26G-RNAs are sufficient for silencing . Absence of either the maternal or the zygotic pools can thus be compensated , enhancing the robustness of this system . We note , however , that sufficiency has only been tested with the described 22G sensor . It may be that the silencing of other targets has differential dependencies on maternal and zygotic 26G-RNA populations . The maternal effect was observed in mutants of a variety of Eri genes , including gtsf-1 , rrf-3 and ergo-1 , but not alg-3/4 . Therefore , these defects are related to impairment of sRNA populations directly associated with and downstream of ERGO-1 . These results do not exclude a parental effect for ALG-3/4 . In fact , a paternal effect on embryogenesis has been described for rrf-3 mutants [34] . Such phenotype most likely arises due to ALG-3/4 branch sRNAs . Maternal rescue of Eri genes was previously reported [49] , although the genetic basis for this phenomenon was not characterized further . We demonstrate that in the first Eri mutant generation , primary 26G-RNAs are downregulated , while their downstream 22G-RNAs are still present ( Fig 2 ) . These ERGO-1-dependent 22G-RNAs , maintained in the absence of their primary triggers , seem to be competent to sustain silencing of ERGO-1 targets throughout life of the animal ( Fig 3 ) . Given that 1 ) ERGO-1 targets display higher expression during embryogenesis; and 2 ) upon disruption of endogenous RNAi by rrf-3 mutation , targets become upregulated in all developmental stages ( Fig 3C ) ; maternally deposited ERGO-1-dependent factors may be especially required to initiate target silencing during embryogenesis , and to prevent spurious expression throughout development . The ERGO-1-independent maintenance of this silencing response may be mechanistically similar to RNA-induced epigenetic silencing ( RNAe ) , involving a self-perpetuating population of 22G-RNAs [53 , 62 , 63] . Indeed , both processes depend on a nuclear Argonaute protein: HRDE-1 in RNAe [53 , 62 , 63] and NRDE-3 for ERGO-1-driven silencing [43 , 46 , 47] . Self-perpetuating 22G-RNA signals may be also in place in the male germline ( see below ) . Our genetic experiments and sequencing data are fully consistent with maternal inheritance of 26G-RNAs . However , these may not be the only inherited agent . A non-mutually exclusive idea is that GTSF-1 , as well as other ERIC proteins may be deposited in embryos to initiate production of zygotic sRNAs . In accordance with the latter , we have previously demonstrated that formation of the 26G-RNA generating ERIC is developmentally regulated [36] . While in young adults there is a comparable amount of pre- and mature ERIC , in embryos there is proportionally more mature ERIC . These observations suggest that pre-ERIC might be deposited in the embryo to swiftly jumpstart zygotic 26G-RNA expression after fertilization . We show that GTSF-1 is required in adult males , potentially in the male germline , to produce 26G- and downstream 22G-RNAs ( Fig 4 ) analogous to its role in the hermaphrodite germline and in embryos [36] . In addition , the bulk of targets from both 26G-RNA pathway branches seem to be deregulated . Interestingly , we note that although ERGO-1 and its cognate 26G-RNAs are not abundantly expressed in spermatogenic tissues ( Fig 4B ) , gtsf-1-dependent , secondary 22G-RNAs mapping to these genes maintain gene silencing in the adult males ( Fig 4F and 4G ) . In an analogous manner , we find that ALG-3/4 targets maintain 22G-RNAs in gravid adults [36] , even though ALG-3/4 are not expressed at that stage . Mechanistically this may be closely related to how maternal 26G-RNAs can trigger 22G-RNA-driven silencing ( see above ) . NRDE-3 is downstream of ERGO-1 and is likely to silence ERGO-1 targets throughout development . However , the Argonautes associated with 22G-RNAs mapping to 1 ) ERGO-1 targets in the male , and 2 ) to ALG-3/4 targets in gravid adults have not yet been identified . ALG-3/4-branch 26G-RNAs map very sharply to the 5’ and 3’ extremities of the targets , very close to the transcription start and end sites . We find that stronger targeting at the 3’ end does not drive robust gene silencing , and may even license expression , while targeting at the 5’ end is associated with stronger gene silencing . Targeting at the 3’ is consistent with RdRP recruitment to synthesize antisense secondary 22G-RNAs throughout the transcript . These may associate with CSR-1 and could have a positive effect on gene expression . The sharp 5’ peak in the metagene analysis could hint at additional regulatory modes , other than 22G-RNA targeting . 5’-end-bound ALG-3/4 could recruit other effector factors , which promote RNA decay or translation inhibition , e . g . by inhibiting the assembly of ribosomes . Of note , when single targets are considered individually , 26G-RNA peaks at 5’ and 3’ can be simultaneously detected ( Figs 7 , S2B and S5A , left panels and S6 ) . Hence , the resolution of a balance between Argonaute-sRNA complexes binding at 5’ and 3’ could determine regulatory outcome . Notably , we find shorter 3’ UTRs to be correlated with gene silencing ( Fig 6F ) . In a model where predominant 3’ UTR targeting by Argonaute-sRNA complexes promotes gene expression , shorter 3’ UTRs and therefore less chance of sRNA binding may shift the balance towards gene silencing . Another possibility may be that longer 3’ UTRs contain binding sites for additional RNA binding proteins that may help to restrict RdRP activity on the transcript in question . Further work will be needed to test such ideas . In C . elegans , primary sRNAs trigger the production of abundant secondary sRNAs . If left uncontrolled , such amplification mechanisms can be detrimental to biological systems . Endogenous and exogenous RNAi pathways in C . elegans compete for limiting shared factors and the Eri phenotype is a result of such competition [37 , 48] . Competition for shared factors is in itself a mechanism to limit accumulation of sRNAs . In support of this , exogenous RNAi was shown to affect endogenous sRNA populations , thus restricting the generations over which RNAi effects can be inherited [64] . We find that 26G-RNAs , likely ALG-3/4-bound , as well as 22G-RNAs map to alg-3 and alg-4 mRNAs ( Figs 7 and S6 ) . In the absence of GTSF-1 , a loss of these sRNAs is accompanied by a 2–3 fold upregulation of alg-3 and alg-4 on the mRNA level . This means that ALG-3 and ALG-4 may regulate their own expression . In the future , the retrieval of alg-3 and alg-4 mRNAs , as well as of 26G-RNAs complementary to their sequence , in immunoprecipitations of ALG-3 or ALG-4 will strongly support this regulatory loop . Such regulation is not unprecedented . Complementary endo-siRNAs to ago2 have been described in Drosophila S2 cells [65] . Since AGO2 is required for the biogenesis and silencing function of endo-siRNAs , it is likely that Ago2 regulates itself in S2 cells . In addition , other studies in C . elegans have described cases where sRNAs are regulating the expression of RNAi factors [64 , 66 , 67] . Such direct self-regulation of Argonaute genes may constitute an important mechanism to limit RNAi-related responses , but the biological relevance of this regulation will need to be addressed experimentally . These observations do suggest that the Eri phenotype is but one manifestation of intricate cross-regulation governing the RNAi pathways of C . elegans .
C . elegans was cultured on OP50 bacteria according to standard laboratory conditions [68] . Unless otherwise noted , worms were grown at 20°C . The Bristol strain N2 was used as the standard wild-type strain . All strains used and created in this study are listed in S5 Table . Wide-field photomicrographs were acquired using a Leica M165FC microscope with a Leica DFC450 C camera , and were processed using Leica LAS software and ImageJ . Cross outline . We first linked gtsf-1 ( xf43 ) and dpy-4 ( e1166 ) . These genes are 2 . 62 cM apart , which does not comprise extremely tight linkage . Therefore , throughout the outcrossing scheme , worms were consistently genotyped for gtsf-1 and phenotyped for dpy-4 . We started by outcrossing dpy-4;gtsf-1 hermaphrodites with N2 males ( in a 1:2 ratio ) . dpy-4 ( e1166 ) is reported as being weakly semi-dominant ( https://cgc . umn . edu/strain/CB1166 ) . Indeed , heterozygote worms look only very slightly Dpy , therefore for simplicity , we refer to the heterozygote phenotype as “wild-type” throughout this work . Wild-type looking worms were selected in the F1 and F2 generations . The F2s were allowed to lay embryos for 1–2 days and then were genotyped for gtsf-1 ( xf43 ) using PCR . Progenies of non-recombined gtsf-1 heterozygote worms were kept for follow up . F3 progenies that did not segregate dpy worms were discarded . F3 dpys were isolated , allowed to lay embryos , and genotyped for gtsf-1 ( xf43 ) . Progenies of non-homozygote mutant gtsf-1 ( xf43 ) worms were discarded . RNAi . dsRNA against lir-1 was supplemented to worms by feeding as described [69] . L1 worms were transferred to RNAi plates and larval arrest was scored 2–3 days later . L1 F3 and F4 worms were transferred to RNAi plates blinded to genotype/phenotype ( the dpy phenotype only shows clearly from L3 onwards ) . him-5 ( e1467 ) and him-5 ( e1467 ) ; gtsf-1 ( xf43 ) worm populations were synchronized by bleaching , overnight hatching in M9 and plated on OP50 plates the next day . Worms were grown until adulthood for approximately 73 hours and 400–500 male animals were hand-picked for each sample , in biological triplicates , and used to isolate RNA ( see below for RNA isolation protocol ) . Each sample was used to prepare small RNA and mRNA libraries ( see below details on library preparation ) . Plates with the hand-picked worms were rinsed and washed 4–6 times with M9 supplemented with 0 . 01% Tween . 50 μL of M9 plus worms were subsequently frozen in dry ice . N2 and rrf-3 ( pk1426 ) animal populations were synchronized by bleaching , overnight hatching in M9 and plated on OP50 plates the next day . L1 animals were allowed to recover from starvation for 5 hours , and then were collected . L2 worms were collected 11 hours after plating . L3 animals were collected 28 hours after plating . L4 animals were collected 50 hours after plating , and young adults were collected 56 hours after plating . Animals were rinsed off plates and washed 4–6 times with M9 supplemented with 0 . 01% Tween . 50 μL of M9 plus worms were subsequently frozen in dry ice . Embryo samples were collected from bleached gravid adult animals , followed by thorough washes with M9 . Samples were collected in triplicate and RNA isolation proceeded as described below . Worm aliquots were thawed and 500 μL of Trizol LS ( Life Technologies , 10296–028 ) was added and mixed vigorously . Next , we employed six freeze-thaw cycles to dissolve the worms: tubes were frozen in liquid nitrogen for 30 seconds , thawed in a 37°C water bath for 2 minutes , and mixed vigorously . Following the sixth freeze-thaw cycle , 1 volume of 100% ethanol was added to the samples and mixed vigorously . Then , we added these mixtures onto Direct-zol columns ( Zymo Research , R2070 ) and manufacturer’s instructions were followed ( in-column DNase I treatment was included ) . NGS library prep was performed with Illumina's TruSeq stranded mRNA LT Sample Prep Kit following Illumina’s standard protocol ( Part # 15031047 Rev . E ) . Starting amounts of RNA used for library preparation , as well as the number of PCR cycles used in amplification , are indicated in S6 Table . Libraries were profiled in a High Sensitivity DNA on a 2100 Bioanalyzer ( Agilent technologies ) and quantified using the Qubit dsDNA HS Assay Kit , in a Qubit 2 . 0 Fluorometer ( Life technologies ) . Number of pooled samples , Flowcell , type of run and number of cycles used in the different experiments are all indicated in S6 Table . For maternal effect sequencing , RNA was directly used for library preparation , or treated with RppH prior to library preparation . RppH treatment was performed as described in reference 51 with slight modifications . In short , 500 ng of RNA were incubated with 5 units of RppH and 10x NEB Buffer 2 for 1 hour at 37°C . Reaction was stopped by incubating the samples with 500 mM EDTA for 5 minutes at 65°C . RNA was reprecipitated in 100% Isopropanol and ressuspended in nuclease-free water . NGS library prep was performed with NEXTflex Small RNA-Seq Kit V3 following Step A to Step G of Bioo Scientific`s standard protocol ( V16 . 06 ) . Both directly cloned and RppH-treated libraries were prepared with a starting amount of 200ng and amplified in 16 PCR cycles . Amplified libraries were purified by running an 8% TBE gel and size-selected for 18–40 nts . Libraries were profiled in a High Sensitivity DNA on a 2100 Bioanalyzer ( Agilent technologies ) and quantified using the Qubit dsDNA HS Assay Kit , in a Qubit 2 . 0 Fluorometer ( Life technologies ) . All 24 samples were pooled in equimolar ratio and sequenced on 1 NextSeq 500/550 High-Output Flowcell , SR for 1x 75 cycles plus 6 cycles for the index read . RNA from adult males was RppH-treated as described above with the difference that 800 ng of RNA were used for RppH treatment . Library preparation of these samples was performed exactly as described above with the following modifications: starting amount of 460 ng; and amplification in 15 PCR cycles . A summary of the sequencing output can be found in S1 Table . | Small RNAs ( sRNAs ) and their partner Argonaute proteins regulate the expression of target RNAs . When sperm and egg meet upon fertilization , a diverse set of proteins and RNA , including sRNA-Argonaute complexes , is passed on to the developing progeny . Thus , these two players are important to initiate specific gene expression programs in the next generation . The nematode Caenorhabditis elegans expresses several classes of sRNAs . 26G-RNAs are a particular class of sRNAs that are divided into two subpopulations: one expressed in the spermatogenic gonad and another expressed in oocytes and in embryos . In this work , we describe the dynamics whereby oogenic 26G-RNAs setup gene silencing in the next generation . In addition , we show several ways that spermatogenic 26G-RNAs and their partner Argonautes , ALG-3 and ALG-4 , use to regulate their targets . Finally , we show that ALG-3 and ALG-4 are fine-tuning their own expression , a rare role of Argonaute proteins . Overall , we provide new insights into how sRNAs and Argonautes are regulating gene expression . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] | [
"invertebrates",
"rna",
"interference",
"caenorhabditis",
"gene",
"regulation",
"animals",
"invertebrate",
"genomics",
"animal",
"models",
"age",
"groups",
"developmental",
"biology",
"caenorhabditis",
"elegans",
"model",
"organisms",
"genome",
"analysis",
"experimental",
"organism",
"systems",
"epigenetics",
"embryos",
"research",
"and",
"analysis",
"methods",
"gene",
"silencing",
"genomic",
"libraries",
"embryology",
"genetic",
"interference",
"animal",
"studies",
"gene",
"expression",
"animal",
"genomics",
"people",
"and",
"places",
"biochemistry",
"rna",
"eukaryota",
"young",
"adults",
"nucleic",
"acids",
"genetics",
"nematoda",
"biology",
"and",
"life",
"sciences",
"population",
"groupings",
"genomics",
"computational",
"biology",
"organisms"
] | 2019 | Maternal and zygotic gene regulatory effects of endogenous RNAi pathways |
Functional magnetic resonance imaging ( fMRI ) , with blood oxygenation level-dependent ( BOLD ) contrast , is a widely used technique for studying the human brain . However , it is an indirect measure of underlying neuronal activity and the processes that link this activity to BOLD signals are still a topic of much debate . In order to relate findings from fMRI research to other measures of neuronal activity it is vital to understand the underlying neurovascular coupling mechanism . Currently , there is no consensus on the relative roles of synaptic and spiking activity in the generation of the BOLD response . Here we designed a modelling framework to investigate different neurovascular coupling mechanisms . We use Electroencephalographic ( EEG ) and fMRI data from a visual stimulation task together with biophysically informed mathematical models describing how neuronal activity generates the BOLD signals . These models allow us to non-invasively infer the degree of local synaptic and spiking activity in the healthy human brain . In addition , we use Bayesian model comparison to decide between neurovascular coupling mechanisms . We show that the BOLD signal is dependent upon both the synaptic and spiking activity but that the relative contributions of these two inputs are dependent upon the underlying neuronal firing rate . When the underlying neuronal firing is low then the BOLD response is best explained by synaptic activity . However , when the neuronal firing rate is high then both synaptic and spiking activity are required to explain the BOLD signal .
Functional magnetic resonance imaging ( fMRI ) is an extensively employed neuroimaging technique that allows the non-invasive recordings from human brain of neuronal activity with relatively high spatial resolution . However , the blood oxygenation level-dependent ( BOLD ) contrast on which fMRI is based is only an indirect measure of this activity . The processes that link the underlying neuronal activity to the BOLD signals are still a topic of much debate . In particular , there is no consensus on the relative roles of synaptic and spiking activity in the generation of BOLD signals . In order to relate findings from fMRI research to other measures of neuronal activity it is important to understand the underlying neurovascular coupling mechanism [1] . Most of our present knowledge about neurovascular coupling comes from animal experiments . These studies have combined hemodynamic measures such as cerebral blood flow ( CBF ) , with electrical measurements such as local field potentials ( LFPs ) and single/multi-unit activity ( S/MUA ) . LFPs correspond primarily to weighted averages of synchronised dendro-somatic components of synaptic signals in a neuronal population , whilst S/MUA measures the action potentials of a single cell or population of cells , respectively [2] . In a pioneering study [3] , found , in monkey visual cortex , that although both LFPs and MUA correlate with the BOLD response , this response could be predicted more accurately from LFPs . This result has been confirmed in awake animals [4] . On the other hand , [5] , [6] and [7] found strong positive correlations between blood flow and spiking activity . More recently , [8] , [9] and [10] have shown that when synaptic and spiking activity is uncoupled ( by drug injection in [8] , [10] and using a stimulus that elicits only synaptic activity in [9] ) , changes in CBF do not reflect underlying spiking activity and relate closer to LFPs . This growing body of evidence ( Table 1 ) therefore supports the hypothesis that BOLD signals are more closely coupled to synaptic input and processing activity than to the output spikes of a population of neurons . In addition , this work ( Table 1 ) provides support to a growing consensus in which the BOLD signal is thought to result from pre-synaptic activity and the release of neurotransmitters , in particular glutamate [11] , as well as vasodilatory substances , such as nitric oxide [12] , [13] and [14] . An increase in pre-synaptic activity and concomitant release of glutamate induces fluctuations in transmembrane potential at the post-synaptic neuron , and these fluctuations are measured with LFPs . This activity is also thought to be responsible for triggering the release of vasodilatory agents to the extracellular medium , which induce changes in blood flow and consequently the BOLD response [11] . However , when it comes to the human brain the number of studies directly addressing the question of how BOLD relates to synaptic versus spiking activity is relatively smaller ( Table 1 and 2 ) , and the data in these studies comes exclusively from neurosurgical patients , whose physiology may be compromised ( Table 2 ) . Of the few such studies [15] , observe significant correlations between BOLD signals and both synaptic and spiking signals in auditory cortex , whilst [16] found no correlation between BOLD signals and neuronal firing in the hippocampal area . The link between neuronal activity and the BOLD response has not only been investigated at a microscopic level , using invasive co-localised recordings , but also at a macroscopic scale using fMRI and Electroencephalography ( EEG ) . EEG ( and Magnetoencephalography ( MEG ) ) , are well established non-invasive techniques that are well suited to studying neuronal activity since they provide direct ( not confounded by the hemodynamic response ) measurement of post-synaptic potentials ( magnetic fields ) in cortical pyramidal cell populations with high temporal resolution [17] . Studies using both EEG and fMRI in humans have focused on correlations between BOLD signals and oscillatory EEG power measured in different frequency bands . For example , [18] , [19] and [20] have shown that reductions in ongoing-scalp EEG alpha power ( 8–13 Hz ) correlate with increases in BOLD activity in human occipital cortex . Using intra-cranial recordings in epileptic patients [21] , have found a close spatial correspondence between regions of fMRI activation and sites showing EEG energy variation in the gamma band ( ) . The main conclusion of this body of work is that increases in EEG frequency are associated with increases in BOLD signal . Even though these studies do not address our question ( input versus output ) directly , they seem to point in the direction of the biological hypothesis constructed from animal evidence ( see above ) : increases in pre-synaptic activity , decrease effective membrane time-constants and result in faster oscillatory dynamics; at the same time more neurotransmitters are released ( e . g . glutamate ) , which lead to increases in BOLD signal [14] . Here we design a powerful and efficient modelling framework to explicitly investigate competing hypotheses for the relationship between neuronal activity and the BOLD response in the healthy human brain . We use this framework to explore the relative contribution of synaptic and spiking activity to the generation of fMRI signals in visual cortex . The participation of healthy subjects prohibits the use of invasive electrophysiological measures . Therefore we use a mathematical modelling framework that allows us to non-invasively infer the degree of local synaptic and spiking activity , together with EEG-fMRI data , in which subjects were exposed to a reversing checkerboard of varying frequencies . This is similar in spirit to the use of ‘virtual electrodes’ in EEG analysis [22] , but provides more specific biophysical information . This framework consists of a biophysically informed forward model from neuronal activity to the observed EEG and fMRI signals . Models linking neuronal activity to EEG/MEG signals have been proposed by [23] , [24] and [25] , to mention a few . These models usually use one or two state variables to represent the mean electrical activity of neuronal populations at the macro-column level , and are referred to as neural mass models [26] . Models linking ‘neuronal activity’ to BOLD signals include the metabolic models proposed by [27] , [28] and the Balloon model , proposed by [29] . The Balloon model describes how evoked changes in blood flow are transformed into the BOLD response and has been extended by [30] , who introduced a blood flow-inducing signal relating ‘neuronal activity’ and CBF and by [31] , where different metabolic pathways have been proposed for supporting excitatory and inhibitory synaptic activity . In the above work ‘neuronal activity’ is usually not explicitly modelled and often corresponds to the stimulus input functions . Models linking a common underlying neuronal substratum to both EEG and fMRI signals have also been developed [32] . Some models are phenomenologically motivated , such as the ‘Heuristic’ proposed by [33] . This model aims to explain empirical results which relate frequency-specific power changes in EEG with fMRI signals and predicts that increases in the BOLD contrast reflect increases in the Root Mean Squared ( RMS ) frequency of EEG . We have validated these predictions in previous work [34] using simultaneous EEG-fMRI data in humans with a visual flicker stimulation task . As predicted by [33] , the RMS frequency significantly explained more BOLD activity than the total time-varying spectral power or any linear combination of frequency-band amplitude modulations ( e . g . alpha or gamma power ) . Biophysically motivated models include [35]–[37] . Most of these theoretical frameworks combine the neural mass model approach for EEG with the Balloon model for fMRI , but the coupling between neuronal activity and blood flow differs from model to model . For instance [35] , propose that the squared post-synaptic membrane potential from both excitatory and inhibitory cells from a cortical area drives increases in cerebral blood flow , whilst [37] consider all the incoming action potentials from populations within and outside the voxel to be the input to the BOLD response . In [36] this input is proportional to the total concentration of nitric oxide ( NO ) synthesised by neurons in the cortical unit . The parameters of this model have been estimated using EEG-fMRI data from the visual cortex of one subject exposed to a reversing checkerboard with varying frequency [38] . Despite these theoretical efforts , the existing modelling frameworks have not yet been used in conjunction with real electrophysiological and hemodynamic data to compare different neurovascular coupling mechanisms , although important steps in this direction have been taken by [36] , [39] . In [39] , the authors have compared different models to investigate the role of excitatory and inhibitory activity in the generation of BOLD signals , using fMRI data from one subject . They found BOLD signals to be best explained by excitatory activity alone . Here we use the forward model proposed by [36] and embed it within a Bayesian framework . Using EEG and fMRI data in combination with Bayesian inference allows us to estimate the underlying synaptic and spiking activity , along with other biophysical model parameters . These quantities are computed using the variational Laplace method described in [40] . This optimisation scheme has been successfully applied to other input-state-output systems , such as [41] , [42] . However , inverting generative models using multi-modality datasets , can be a technically demanding task , if the temporal characteristics of the datasets are very different , which is the case for EEG-fMRI data . Here we develop a computationally efficient scheme for model inversion . Instead of inverting the model in a single ( computationally demanding ) step we adopt a ‘multi-step inversion’ approach . This approach is based on partitioning model inversion into multiple , independent and computationally efficient steps that are motivated by the time-scales of data involved . This is a general procedure that can be used with other datasets and in other multimodal studies , such as with MEG-fMRI or LFP-fMRI data . Finally , once equiped with this mathematical and computational framework we posit models embodying different hypotheses about neurovascular coupling and adjudicate between them using Bayesian model evidence [43] . We compare three models . The first assumes that blood flow depends on the amount of vasodilatory substances ( e . g . nitric oxide ) released as a result of synaptic activity ( synaptic input model ) , as proposed by [36] . The second assumes blood flow is driven by the firing rate of pyramidal cells from the same unit ( spiking output model ) . These hypotheses are then compared against a third model where both these quantities contribute to the BOLD response ( mixture model ) . In the long term , we anticipate that this modelling framework will be used to test neurovascular coupling hypotheses in a variety of experimental contexts with a range of subject cohorts .
We use a realistic biophysical model , proposed by [36] , of how electrical and vascular dynamics are generated within a cortical unit . The unit comprises three subpopulations of cells: two layer IV GABAergic interneuron populations ( the transmission and feedback interneurons ( INs ) ) and a layer V pyramidal cell ( PC ) population ( Figure 1 ( a ) ) . Interneurons are modelled as single compartment neurons , whilst the pyramidal cell has three compartments ( soma , basal and apical tuft dendrites ) . Here we briefly describe the forward model . A summary of all the equations and parameters of the model can be found in Text S1 . For a more detailed description please consult the original work [36] . A neural mass model ( NMM ) characterises the population dynamics of electrical states such as the membrane potentials in the somas of the neurons and electric currents flowing in the neuropil . This modelling framework is appropriate for data that reflect the behaviour of neuronal populations , such as EEG and fMRI data . The neural mass model can be viewed as a special case of ensemble density models , where the ensemble density is summarised with a single number representing mean activity [44] . Assuming that the equilibrium density of the neuronal states has a point mass ( i . e . , a delta function ) , we can reduce the density dynamics to the location of that mass . What we are left with is a set of non-linear differential equations describing the evolution of this mode . The time variations of membrane potential in the individual compartments of the pyramidal cell and single compartment interneurons , , are determined by the differential equation for a simple voltage source circuit: ( 1 ) where is the effective membrane resistance of the compartment , and is cell-type and compartment specific . is the membrane time constant ( same for all cells and compartments ) . The current , , that flows through the membrane of the cell depends on the connections between different elements of the cortical unit and its external inputs ( Figure 1 ( a ) ) . The cortical unit receives external excitatory input in different subpopulations , whilst its sole output is the firing rate of the pyramidal cells , . The excitatory inputs to the transmission interneuron , , and basal dendrites of the pyramidal cell , , correspond to thalamo-cortical afferent projections . The input to the apical tuft dendrites , , mediates cortico-cortical interactions . These currents can be found in Figure 1 ( a ) . In terms of synaptic connections within the cortical unit , the total inhibitory synaptic effect on the pyramidal cell is given by: , where is the transmission inhibitory current and the feedback inhibitory current . The inhibitory synaptic currents depend nonlinearly on the membrane potential of the GABAergic cells through a threshold function: . The excitatory synaptic current generated by the pyramidal cell has the same form: : ( 2 ) The parameters are set to and to ensure that the output stays between and . The and parameters determine the voltage sensitivity by setting the membrane potential maximum growth and growth rate , respectively . These parameters are estimated from the data . determines the membrane potential near the asymptote where maximum growth occurs . The threshold function , , is also used to construct the firing rate coupling model ( see below ) . The equations for the membrane potential at the soma of the three-compartment pyramidal cell , as well as the extracellular potential along its apical dendrites can be determined from the potentials and currents at the individual compartments ( given by Eq . 1 ) . These equations can be found in Text S1 . The apical dendrites of the layer V pyramidal cells are arranged in parallel to each other and perpendicularly oriented to the surface of the cortex . This geometry facilitates the summation of electric currents in the neuropil . The mesoscopic effect resulting from the spatial average of these extracellular currents corresponds to the electrical signal measured with EEG . The state variables , , and parameters , , of the neural mass model described above are summarised in Table 3 of the main text and Tables 1 and 2 in Text S1 . The coupling between local neuronal activity , described by the neural mass model , and subsequent changes in vascular dynamics is our question of interest . These changes are expressed in the BOLD signal and have previously been modelled in an extended Balloon approach [30] , in which a set of four ordinary differential equations comprise the hemodynamic forward model from ‘neuronal activity’ to hemodynamic responses . The full derivation of these equations can be found in [29] and [30] . In brief , for a particular region , neuronal activity , , causes an increase in a vasodilatory signal , , that is subject to auto-regulatory feedback . Inflow , responds in proportion to this signal with concomitant changes in blood volume and deoxyhemoglobin content . These equations are summarised in Text S1 . The hemodynamic parameters , , comprise the rate constant of the vasodilatory signal decay , the rate constant for autoregulatory feedback by blood flow , transit time , Grubb's vessel stiffness exponent , and the resting oxygen extraction fraction , respectively . The whole dynamic system is driven by the input . Different inputs , , correspond to different aspects of neuronal activity and consequently different coupling hypotheses between neuronal activity and the BOLD response . A summary of the hemodynamic model's state variables , , and parameters , , can be found in Table 3 of the main text and Tables 1 and 2 in Text S1 . In the next section we specify the neurovascular coupling mechanisms we are interested in comparing . The original electro-vascular model proposed by [36] is represented by a set of stochastic differential equations describing the dynamics of the neuronal and vascular states , . In [36] the stochastic aspect of the model is instantiated by incorporating an additive multidimensional Wiener process to model physiological noise . In this paper , however , we use a deterministic version of the model . This means that the dynamics are completely determined by the state of the system and stochastic effects enter only at the observation level ( Eq . 3 ) . This deterministic approach resulted in very similar frequency-response curves to those in [36] ( see Results: synthetic data below ) and allows us to use standard Bayesian estimation routines , widely used with deterministic forward models for EEG ( e . g . [42] ) and fMRI ( e . g . [40] ) . The observation equations for EEG , , and fMRI , , data are then given by: ( 3 ) where the errors are assumed to be i . i . d . , . The temporal variations of the EEG signal are well approximated by the extracellular electric current in the neuropil , , obtained from the NMM multiplied by the lead field matrix , . This matrix contains information about the geometry and conductivity of the head , and is therefore employed to map the distributed electric sources within the brain to scalp EEG recordings [45]: ( 4 ) The observation function for fMRI is a static nonlinear function of the cerebral blood volume and the concentration of deoxyhemoglobin directly [30]: ( 5 ) The factors , and are dimensionless but depend on the characteristics of the fMRI recording system . For 1 . 5 T and TE of 40 msec , . is the resting blood volume fraction . To link the two main components of the biophysical model , the neural mass model and the Balloon model , we specified three different biologically plausible neurovascular coupling mechanisms based on previous empirical results . These mechanisms are described below: Using EEG-fMRI data in combination with Bayesian inference allows us to estimate the underlying synaptic and spiking activities , along with other parameters of the biophysical framework . Additionally , we can compare the different neurovascular coupling hypotheses using Bayesian model evidence . In Bayesian inference , prior beliefs about parameters , , of model are quantified by the prior density , . Inference on the parameters , , after observing data , , is based on the posterior density . These densities are related through Bayes' rule: ( 12 ) where is the probability of the data ( likelihood ) conditioned upon the model and its parameters . The normalisation factor , , is called the model evidence and plays a central role in model comparison ( see below ) . The posterior density is an optimal combination of prior knowledge and new observations , weighted by their relative precision ( i . e . , inverse variance ) , and provides a complete description of uncertainty about the parameters . Generally , the choice of priors reflects either empirical knowledge ( e . g . , previous measurements ) or formal considerations ( e . g . , biological or physical constraints ) . Here we use empirical knowledge for both the neural mass model parameters and the coupling/hemodynamic parameters , based on estimates obtained by [36] . Under Gaussian assumptions , also known as a fixed-form Laplace approximation [55] , the problem of estimating the posterior density reduces to finding its first two moments , the conditional mean and conditional covariance . The prior density is also assumed to be Gaussian with mean and covariance ( see Table 3 for a list of prior mean values ) . A non-linear model , such as the local electro-vascular ( LEV ) model used here , Eq . ( 3 ) , can be linearised by expanding the observation equation about a working estimate of the conditional mean: ( 13 ) such that , and . In this paper , the error covariance is assumed isotropic over the EEG and fMRI predictions . The linearised model , Eq . ( 13 ) , can be used in a Variational Laplace ( VL ) optimisation scheme that iteratively updates the moments of the conditional density , . VL is a generic approach to estimate the posterior density , and can be formulated by analogy with statistical physics as a gradient ascent on the ‘negative Free Energy’ , , of the system . The full derivation of the algorithm is described in [56] . The maximisation of with respect to in effect maximises a lower bound on the log model evidence , , [57]: ( 14 ) The model evidence is the probability of obtaining observed data , , given model , , and is at the heart of Bayesian Model Selection ( BMS ) . The last term in Eq . ( 14 ) is the Kullback-Leibler ( KL ) divergence between the approximate posterior density , , and the true posterior , . This quantity is always positive , or zero when the densities are identical , and therefore is bounded below by . Through the iterative optimisation described above , the KL divergence is implicitly minimised and becomes an increasingly tighter lower bound on the log-model evidence . Model comparison can then proceed using as a surrogate for the log-model evidence . This approximation to the posterior density has been evaluated using Markov Chain Monte Carlo ( MCMC ) [58] . These schemes are more computationally intensive but allow one to estimate the posterior density without assuming it has a fixed form . Comparison between the model evidence obtained by MCMC methods and by variational approaches showed similar estimates , confirming that the approximations entailed by the variational approach lead to accurate model selection [55] . Again through Bayes' rule we can relate the model evidence to the model posterior probability , : ( 15 ) where is the prior distribution over models . Selecting the optimal model corresponds to choosing the model that maximises the posterior . If no model is favoured a priori then is a uniform distribution , and the model with the highest posterior probability is also the model with the highest evidence , . Given two models , and , we can compare these models using Bayes Factors , [59] , which are defined as the ratio of the corresponding model evidences , or equivalently the difference in their log-evidences: ( 16 ) Bayes factors have been stratified into different ranges deemed to correspond to different strengths of evidence . ‘Strong’ evidence , for example , corresponds to a BF of over 20 ( log-BF over 3 ) [59] in favour of model when compared to model . The equivalent posterior model probability is greater than 0 . 95 [43] . Here we use Bayes factors to compare the neurovascular coupling models defined in the previous section .
In this section , simulations are used to explore the behaviour of the model and its ability to reproduce EEG and BOLD data under the experimental conditions described in the previous section . The response of the three neurovascular coupling models to changes in stimulus frequency is also shown . These synthetic signals are used to test the model inversion routines and to verify that Bayesian model comparison can be used to infer the correct coupling model . The LEV model was numerically integrated using the multi-step Adams-Bashforth-Moulton predictor-corrector algorithm implemented in the MATLAB ( The MathWorks , Inc . ) function ode113 . The integration step used was 1 msec ( 1000 Hz ) for the electrical and vascular states . The integrated signals were then downsampled to 100 Hz in the EEG case and to 0 . 3 Hz for the BOLD signal . The input to the model is described below . Finally we fit the electro-vascular model with the three different coupling mechanisms to the EEG and fMRI data . We used the same ‘multi-step’ inversion procedure described in the previous section . Figure 9 shows the model predictions for EEG , as well as predictions of the coupling models and the BOLD response .
In this paper we used EEG-fMRI data and a biophysically informed mathematical model to investigate the relationship between neuronal activity and the BOLD signal in human visual cortex . In particular , we explored the contributions of synaptic input and spiking output activities to the generation of the BOLD response . We have provided preliminary evidence that the BOLD signal is dependent upon both synaptic and spiking activity but that the relative contribution of these two factors are dependent upon the underlying neuronal firing rate . When the underlying neuronal firing is low then BOLD signals are best explained by synaptic input , in agreement with previous animal studies , such as [3] . This result is also in line with more recent studies , such as [9] and [10] , which show that the BOLD response is only affected by changes in synaptic-related activity ( measured with LFPs ) and not by changes in spiking activity ( measured with MUA ) when these two signals can be dissociated . However , when the neuronal firing rate is high then both synaptic and spiking activity are required to explain the BOLD signal , as observed in , for example [6] , and [15] . We were particularly encouraged to find that a combination of synaptic input and spiking output frequency response curves ( Figure 6 ) can explain the doubly-peaked BOLD response observed by [6] and replicated in our own data . One possible explanation for the increased performance of the output model with higher frequencies comes from neuroenergetic studies such as e . g . [7] and [63] . In these studies brain metabolism was found to depend strongly on neuronal spiking , with increases in oxygen consumption reflecting higher firing rates . More recently [64] , have found that differences in the BOLD response between different brain areas ( motor cortex and thalamus ) could be explained by underlying differences in the firing rates of the corresponding neuronal populations . Our results also support the conclusion that the relationship between synaptic activity , spikes and BOLD signals depends on the specific neuronal circuitry engaged in task processing . Moreover , one can speculate that different coupling mechanisms involving different types of cells and molecules could come into play depending on the task in question . Despite our initial concern about the small number of fMRI samples compared to EEG , our initial results with synthetic data showed that it is possible to make inferences on different hypotheses for the neurovascular coupling using a generative modelling framework and Bayesian model comparison . The issue of different time-scales was addressed by partitioning the estimation of electrical and vascular states into a multi-step approach . In this approach we first estimated the electrical states and parameters from the EEG data and then integrated the full electro-vascular model using these estimates . From the integrated model we extracted the input time-series to the Balloon model , which we then inverted using BOLD data . The last two steps were repeated for each coupling model . This method significantly increases the computational efficiency of the model inversion . However , this multi-step approach is only possible with a deterministic model . In this work we used a deterministic version of the stochastic electro-vascular model proposed by [36] . Under different experimental conditions , which do not induce a large sensory response , the introduction of stochastic effects might be essential to reproduce the empirical data . In this case , other Bayesian inversion frameworks can be employed to estimate the model parameters , such as [65] and [66] . It is also worth noting that despite the fact that the mixture model had more parameters than the input and output models , this extra complexity did not provide a significantly better fit to the data in the low-frequency analysis than the input model . This complexity is correctly penalised using Bayesian methods , such as the one used here . One concern about the coupling models defined here regards the definition of NO concentration . As mentioned in the Methods section , NO is thought to have a pre-synaptic synthesis [12] , [13] . However , here and in [36] the concentration of NO is modelled through post-synaptic quantities such as the transmembrane capacitive currents . Although in principle these two phenomena are directly related ( increases in pre-synaptic activity mean larger post-synaptic effects ) this is not always the case . Changes in transmembrane currents at the post-synaptic level can be caused by different processes such as chemical-gated channels , electric-gated channels , and passive leakage , not all of them being related to pre-synaptic activity . Therefore the transmembrane currents are an indirect way of quantifying the amount of NO released during synaptic activity . However , this issue is also encountered in experimental measures of synaptic activity , such as local field potentials . This signal is a surrogate post-synaptic signal , which is also affected by other slow potentials occurring at the cellular level that do not have a purely pre-synaptic origin . A natural extension to this work is the inclusion of multiple cortical units in the model representing multiple brain areas . For instance , sub-cortical areas such as the thalamus and other cortical areas activated by the experimental task could be included . Having more than one area would facilitate the differentiation between input and local processing synaptic activity , such as in [37] . In a recent study [67] , have decomposed the effect of these two types of synaptic activity on hemodynamic signals by reducing the thalamic input to a rodent's cortex . The authors found that although both input and local neuronal processing contribute to BOLD signals , as previously found , this contribution is larger from local processing . Another extension would be to probe the contribution of excitatory and inhibitory neuronal populations to the generation of BOLD signals , such as in [39] . This model-driven approach could , for instance , be used to study the findings of [68] , where a negative BOLD response in deeper cortical layers , adjacent to positive-BOLD areas , was found to be associated with a reduction in local neuronal firing . Very recently [64] , have optically driven genetically modified inhibitory cells and measured a negative BOLD signal in response to this stimulation , in the rat cortex . This result can inform the development of new generative models of neurovascular coupling . To our knowledge this paper presents the first quantitative model comparison of different biologically plausible mechanisms for neurovascular coupling in human cortex using EEG-fMRI data and a realistic biophysical model . However , even though our results were consistent across the three subjects and the majority of sessions , the case study approach adopted here has its limitations . Namely , it does not quantitatively address the issue of inter-subject variability and it therefore precludes inferences at the population level . With a larger sample of subjects , inter-subject variability can be accommodated using the Random-Effects ( RFX ) model selection approach developed by [69] . This approach fits a Bayesian hierarchical model to group model evidence data to obtain the frequencies with which each model is used in the population . This approach can be combined with the methodology developed in this paper . We hope that future studies with other datasets and different experimental conditions will employ our modeling approach so that a balance of evidence can be reached that clearly disambiguates between different hypotheses concerning neurovascular coupling . Understanding the underlying biophysical mechanisms behind the coupling between neuronal activity and the BOLD response is vital not only for improving the interpretability of the BOLD response , but also for relating findings from fMRI research with results from other neuroscientific disciplines . | Functional magnetic resonance imaging ( fMRI ) , with blood oxygenation level-dependent ( BOLD ) contrast , is a widely used technique for studying the human brain . However , the relationship between neuronal activity and blood flow , the basis of fMRI , is still under much debate . A growing body of evidence from animal studies suggests that fMRI signals are more closely coupled to synaptic input activity than to the spiking output of a neuronal population . However , data from neurosurgical patients does not seem to support this view and this hypothesis hasn't yet been tested in the healthy human brain . Here we design a powerful and efficient modelling framework that can be used to non-invasively compare different biologically plausible hypotheses of neurovascular coupling . We use this framework to explore the contribution of these two aspects of neuronal activity ( synaptic and spiking ) to the generation of hemodynamic signals in human visual cortex , with Electroencephalographic ( EEG ) -fMRI data . Our results provide preliminary evidence that depending on the frequency of the visual stimulus and underlying firing rate , fMRI relates closer to synaptic activity ( low-frequencies ) or to both synaptic and spiking activities ( high-frequencies ) . | [
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] | [
"mathematics",
"applied",
"mathematics",
"biology",
"computational",
"biology",
"nonlinear",
"dynamics",
"biophysics",
"neuroscience"
] | 2011 | Bayesian Comparison of Neurovascular Coupling Models Using EEG-fMRI |
Rapidly characterizing the amplitude and variability in transmissibility of novel human influenza strains as they emerge is a key public health priority . However , comparison of early estimates of the basic reproduction number during the 2009 pandemic were challenging because of inconsistent data sources and methods . Here , we define and analyze influenza-like-illness ( ILI ) case data from 2009–2010 for the 50 largest spatially distinct US military installations ( military population defined by zip code , MPZ ) . We used publicly available data from non-military sources to show that patterns of ILI incidence in many of these MPZs closely followed the pattern of their enclosing civilian population . After characterizing the broad patterns of incidence ( e . g . single-peak , double-peak ) , we defined a parsimonious SIR-like model with two possible values for intrinsic transmissibility across three epochs . We fitted the parameters of this model to data from all 50 MPZs , finding them to be reasonably well clustered with a median ( mean ) value of 1 . 39 ( 1 . 57 ) and standard deviation of 0 . 41 . An increasing temporal trend in transmissibility ( , p-value: 0 . 013 ) during the period of our study was robust to the removal of high transmissibility outliers and to the removal of the smaller 20 MPZs . Our results demonstrate the utility of rapidly available – and consistent – data from multiple populations .
Novel strains of influenza emerge as pandemics either from animal reservoirs [1] or from reassortment in humans [2] . Pandemic strains are characterized by low levels of population immunity that permit higher levels of incidence . However , pandemic strains are not necessarily intrinsically more transmissible nor more severe ( at the individual level ) than the previously circulating seasonal strains they often replace [3] . An ability to rapidly and reliably characterize novel strains in terms of their transmissibility is crucial for health planners in both the civilian and military domains: without good estimates for key parameters it is not possible to identify the appropriate strength of interventions [4] , nor is it possible to consider the spatial optimization of interventions based on variability in transmissibility [5] . The basic reproduction number , , quantifies the transmissibility of a pathogen and is defined to be the average number of secondary cases generated by one infectious individual in an otherwise susceptible population [6] . Pathogen-population combinations with typically do not generate large outbreaks following an introduction ( although they may generate self-limiting stuttering chains of transmission [7] ) . The efficacy of an intervention can be measured by the proportion of transmission it is able to avert , thus , high- pathogen-population combinations such as measles in sub-Saharan Africa [8] require highly effective intervention campaigns in order to achieve control . However , although the main utility of is often thought to be in quantifying the strength of intervention needed for control , it is also important in determining the likely efficacy of mitigating interventions [9] in reducing the number of infections [10] when control cannot be achieved . Although often assumed to be a universal constant for a particular pathogen , is variable across time and population for a variety of scales: the within an elementary school may be different from the within a nearby high school in the same way that the within one northern hemisphere city infected during September may be different from the within a second city infected during January . Early estimates of in civilian populations during the 2009 pandemic ranged from 1 . 1 to 3 . 3 and were based on influenza-like-illness data from an ad hoc data gathering process in a single population [11]–[15] . This wide range of values could be explained by one or more of: intrinsic differences between populations , such as host immunity or predisposition to infection; modifications in behavior over time , such as increased or decreased hand hygiene; seasonal climatic variability; methodological differences in parameter estimation; variability in pathogen-specific virulence across regions and/or time; and variability in underlying data-gathering processes . Crucially , because the 2009 pandemic strain was mild , substantive policy uncertainty did not arise from this discrepancy: there was no need to choose between available mitigating interventions because costly strategies were not justified . Nonetheless , should the next emergent human influenza strain be more severe , any estimate of the absolute benefits of transmission blocking interventions would be highly sensitive to variation in of the scale seen in the literature from the 2009 pandemic . Therefore , prior to the start of the next pandemic , there is clear public health value in the timely coupling of routinely collected high-quality data with robust parameter estimation . Such systems could be calibrated each year using data from seasonal influenza epidemics , and would provide useful decision support during severe non-pandemic influenza seasons . In this study , we use data from the Defense Medical Surveillance System ( DMSS ) to: ( 1 ) describe the pandemic profiles observed at military installations; ( 2 ) compare them with available data from the surrounding civilian population to evaluate how much civilian populations drive incidence in military installations; and ( 3 ) use a parsimonious transmission model to estimate installation-specific values . In addition to allowing us to characterize military-specific patterns , our study offers potential insights into their surrounding civilian populations . Possible strengths of analyses based on the DMSS data compared to other data sources for civilian population are: localization ( to within a zip code ) ; consistent reporting over many years; and , potentially , near-realtime availability .
We obtained data from the Armed Forces Health Surveillance Center ( AFHSC ) consisting of outpatient visits to permanent military treatment facilities ( MTFs ) by active duty military personnel for a range of ICD ( international classification of diseases ) -9 codes associated with respiratory-related illnesses between January 1 , 2009 and April 30 , 2011 . For each record , the data contained: a unique study identifier for the individual; ICD-9 codes associated with that visit; and the zip code ( 5 digits ) of the clinic location . We used the zip code of the reporting clinic as a proxy with which to define military installation: we do not explicitly represent military installations or bases , rather , we assume that case reports from the same zip code are from the same military installation . Each record ( an anonymized Study ID ) was assigned as either “ILI-large” ( ) or “ILI-small” ( ) using a set of classifications based on ICD-9 codes [16] . The definition of ILI-large was broader and included non-specific diagnosis such as ‘viral infection’ and ‘acute nasopharyngitis’ ( Table S2 ) . The definition of ILI-small was more constrained and included: ‘Influenza w/other respiratory manifestations’ ( 25 , 293 ) , ‘Influenza with manifestation not elsewhere classified ( NEC ) ’ ( 1006 ) , ‘Infectious upper respiratory , multiple sites , acute NEC’ ( 897 ) , and ‘influenza with pneumonia’ ( ) . See Table S2 for further details . We further trimmed the data temporally to cover the period from April 1 , 2009 through June 1 , 2010 , and ranked these installations by size according to the total number of ILI-small cases they reported . Although the AFHSC DMMS data includes clinic visits by military personnel at many locations around the world , here we focus on the top-50 largest profiles , 47 of which , were located within the USA . Of the remaining three , one was located in Landstuhl , Germany , and two were located in Japan ( Misawa and Yokosuka ) . We obtained civilian data through a variety of means . County-level data were generally acquired directly from the appropriate public health services department or from the CDC . CDC ILI data were obtained from the flu activity and surveillance website [17] . We considered a set of independent deterministic transmission models , one for each military installation . For each , we solved the following set of equations: ( 1 ) ( 2 ) ( 3 ) where represents the number of susceptible individuals , is the number of infectious individuals , is the number of recovered individuals , and is the total active duty population size at each installation . The incidence ( ) is given by , which computationally , is estimated by: ( 4 ) where is the proportion of the infectious active duty population that present themselves to a clinic with ILI-small symptoms , and the integral runs over a week from to . The total population at each military installation , , is arguably a militarily-sensitive parameter . For this study , we estimated it using publicly-available data in the following way . First , we calculated the total number of active duty out-patient visits at each installation for all causes over the period from January 2009 through April 2011 , , which we suggest is proportional to the total population at each installation . To estimate the coefficient of proportionality , , we identified a subset of the installations for which reasonably reliable estimates for the total population have been published ( Figure S3 ) . Estimates of for the top-50 installations are shown in Table S3 , Column 3 . The time-dependent term , , changes from to at time and returns to after an interval . Equivalently , we allowed to change at some point in time , , to a new value . Intuitively , this definition makes sense if we imagine some mechanism , such as school closures on installations , the deployment of troops , or some other behavior modification to drive the effective contact rate down , and , hence , . For purposes of generality , however , we did not impose any requirement that decrease at this time . Even during a pandemic , there are reasons other than influenza infection for cases to present as ILI-small . Therefore , we also included a noise term . It was implemented as a constant added to the model output for incidence during the optimization procedure , resulting in a total of eight parameters . For fitting purposes we further trimmed the data in time from the outside inward and fit to all data bounded by the first non-zero values . Following [18] we define the Akaike Information criterion ( AIC ) , which is a measure of the relative goodness of fit of a model , for a single model at the th military installation to be ( 5 ) where is the value of the maximized log-likelihood over the unknown parameters ( ) , given the data and the model ( Text S1 ) . When the total number of parameters ( ) is large relative to the sample size ( ) , the reduced Akaike Information Criterion is preferred: ( 6 ) Model fits were optimized by first defining a multidimensional hypercube , running the model simulations with the hypergrid parameters and ranking the resulting scores . Each of the top 1 , 000 scores is then used as an initial guess for a multi-dimensional Nelder-Mead ( also known as downhill simplex ) minimization of the Log-Likelihood . The lowest value of these searches is reported . The bounds and resolution of the hypercube are given in Table S1 . We note in passing that while the results presented here relied on pseudo-Poisson log-likelihoods , we also used both and least-squares fits methods to optimize the solutions with no significant differences in results .
We compared time series for both ILI-small and ILI-large with available civilian data from the Centers for Disease Control and Prevention ( CDC , www . cdc . gov , Figure 1a ) for the time period between April 1 , 2009 and March 31 , 2010 . There was substantially greater temporal correlation between the CDC time series and ILI-small ( Pearson correlation 0 . 91 ) than with ILI-large ( Pearson correlation 0 . 80 ) . The time series for ILI-small cases arising from the largest 50 military installations ( as defined in Materials and Methods ) was similar to the total time-series in both trend and amplitude ( Figure 1a ) . For the same time period ( April 1 , 2009 through March 31 , 2010 ) , 13 , 690 out of 21 , 285 ILI-small cases ( 64% ) in the DMSS data occurred in the largest 50 installations . Therefore , we restrict ourselves to ILI-small for the remainder of this study . The aggregate pattern of incidence of ILI-small for the largest 50 installations was driven both by qualitative variation in the shape of incidence curves and by variation in the timing of epidemic peaks ( Figure 1b ) . Broadly speaking , the shape of each installation incidence curve could be described as: ( 1 ) a typical single-peaked epidemic profile , that is , consisting of a single exponential rise , peak , and more gradual decay; ( 2 ) a bimodal profile , consisting of two peaks separated by a month or two; ( 3 ) a very narrow , sharp peak , where the entire outbreak is complete within ∼4weeks; or ( 4 ) a prolonged , noisy , and relatively flat profile , often containing a single-peaked profile within it . For example , the military populations defined by zip code ( MPZ ) 80913 ( Colorado , MPZ-80913 ) experienced a classic epidemic profile for the incidence of ILI-small; taking off in early September , peaking in the middle of November and then dropping to low levels by early January ( Figure 1c ) . In contrast , the profile at MPZ-92134 ( southern California , Figure 1d ) displayed two clear peaks , one in July and another at the end of October 2009 . Finally , at MPZ-22134 ( Quantico , Virginia ) a single , sharp peak was observed in July , with only the hint of a second wave in early November ( Figure 1e ) . The variability of the profiles for the top 50 MPZs is summarized in the heat chart of Figure 1 , which illustrates the variation in timing of the peaks . Individual line plots of incidence for each of the top 50 MPZs are shown in Figure S1 . The peak weeks of incidence during 2009 for individual military installations were clustered primarily around one point during early Autumn 2009 , with a few installations peaking as early as June 2009 ( Figure 2a ) . The timing of peak weeks was not obviously correlated with longitude , latitude , average temperature , precipitation , or with distance from any of the known points of origin for the pandemic strain in the United States ( Figure S2 ) . However , for most military installations for which detailed civilian surveillance data were available for the region containing that zip code , there was a close correspondence between both the timing of the peak of the epidemic in the civilian population and the more detailed incidence profile in those civilian populations ( Figure 2b and 2c ) . In a small number of cases , however , there was a relatively poor correlation ( Figure 2d , see Discussion ) . For each civilian/military profile pairing , we computed temporal cross correlations for the period from April 1st , 2009 through March 31 , 2010 . The correlation coefficients ranged from effectively zero ( MPZ-22134 ) to 0 . 91 ( MPZ-92134 ) , although all but one were . Moreover , the lag that maximized the correlation was typically one week . Thus , the profiles at the military installations were similar in structure to the civilian peaks but delayed by approximately one week ( Figure 2e ) . We calculated an overall Pearson correlation coefficient of 0 . 89 between available pairs of civilian and military populations for the week of peak incidence . Some populations were used more than once in the calculation of the Pearson coefficient because multiple civilian datasets were available for individual military populations . Our modeling framework permitted two alternate models ( one nested within the other , see Methods ) to estimate the transmissibility during the 2009 pandemic at each of the largest 50 MPZs: a one-peak model ( four parameters ) and a two-peak model ( seven parameters ) . As would be expected , the scores for the two-peak model were much better for MPZs that exhibited double peaks of incidence . However , we also found that the two-peak model always provided substantially better support for the data , even when the time series of incidence did not obviously show two separate peaks . Therefore , we report parameter values for the two-peak model for all military installations . In general , we found satisfactory model fits to the military installations ( Figure 3a–d , Table S1 and Figure S1 ) . Usually , was estimated to be greater than 1 , while was less than 1 . However , there were a number of exceptions ( Figure 3e , Text S1 , Table S3 and Figure S6 ) . The fitted values of for the two peak-model model , when fitting to data from all 50 MPZs , were reasonably well clustered with a median ( mean ) value of 1 . 39 ( 1 . 57 ) and standard deviation of 0 . 41 ( Table S4 ) . We checked for any correlation between base size and our estimates of but did not find any ( Figure S5 ) . In most of the two-peaked profiles , decreased at a point in time necessary to drive the initial wave downward , then returned to at the minimum between the two peaks , although this was not always the case . Moreover , for single-peaked profiles was used ( by the model ) prior to the main peak ( e . g . , Figure 3d ) , during the main wave ( e . g . , Figure 3a ) , or even following it ( MPZ-22060 , Table S3 ) . Thus , it was not always obvious which single transmissibility parameter best captured the profile at each installation . To address this , we constructed an “optimum” estimate for , , which was the maximum of or ( Tables S1 and S3 ) . Seven installations had : MPZ-23665 ( Joint Base-Langley-Eustis ) ; MPZ-22134 ( Marine base at Quantico , VA ( Figure 3d ) ) ; MPZ-85309 ( Luke AFB ) ; MPZ-29152 ( Shaw AFB ) ; MPZ-96319 ( Misawa AFB ) ; MPZ-57706 ( Ellsworth AFB , SD ) ; and MPZ-71459 ( Fort Polk ) ( Table S3 ) . It is interesting to note that a disproportionate number of these high- installations are Air Force bases , which , it could be argued , have the most civilian-like policies of any branch of the armed forces . In general , however , ranged from 1 . 0 to 2 . 0 , being strongly biased toward the lower limit ( Figure 3f ) . During the course of the six months , over which the pandemic spread across the military installations , increased from 1 . 1 to 1 . 6 ( Figure 3g , blue line ) . A least-squares fit to the data gave , with a p-value of 0 . 013 . The general increase was present both with and without the outliers and the trend was also captured by the top 30 installations as well as all 50 .
In this study , we have derived incidence curves for individual MPZs using data from the AFHSC DMSS for the 2009 H1N1 pandemic . Comparison of the military incidence profiles with available civilian surveillance and testing data during the same time period suggested that most MPZs were temporally well-synchronized with their enclosing civilian population , but , importantly , tended to lag it by approximately one week . If we assume that the military installation is usually much smaller than the local civilian population , these findings suggest that the local civilian population is driving the timing of peak incidence in many military installations . Using SIR-like transmission models [19] , we described a gradual increase in the transmissibility of influenza during 2009 in these populations . Our study employed a number of assumptions that require careful consideration . First , we estimated the total population at each installation ( ) by assuming that the total number of visits to a clinic for all causes was a reasonable proxy for the total number of active duty personnel at that location ( Figure S3 ) . We estimated the constant of proportionality by comparing this number to published base sizes . However , in addition to intrinsic inaccuracies that these numbers may have , they are also subject to change over time as troops are recruited , deployed , and/or base sizes are changed . Fortunately , these “denominator” data , while undoubtedly sensitive information , are likely well known by military planners . Thus , in the hands of military personnel , these analysis could be easily re-run with significant improvement . Second , our analysis also assumed a constant value for , the proportion of infectious individuals that presented themselves to a clinic . This assumption was made for simplicity , enabling us to address the fundamental properties of the incidence profiles and estimate . is clearly a key parameter that needs to be estimated early in an outbreak to guide policy makers in what types of intervention strategies , if any , should be employed . However , , which is a measure of the severity of the pandemic , is rapidly gaining appreciation . This will be addressed in a forthcoming study . Third , we did not explicitly include age-dependent effects , rates of reporting , nor accurate estimates of the population at-risk , all of which could potentially improve the utility of this approach . However , given the more tightly clustered age distribution within the active component of the military ( typically 18–45 years old ) , together with the smaller number of cases that would define each profile , we suggest that our fitted models have good utility for the characterization of transmissibility . However , with accurate age-specific denominator data for each population we are confident that these methods can be expanded to allow a more finely-resolved study of age-specific transmissibility . Although on average there was good correlation between the military installations and the enclosing civilian populations , this was not universally true . For MPZs that did not track well with the surrounding population , credible explanations can be given . For example , at Elmendorf AFB , just outside Anchorage Alaska , while the profile on-base was relatively simple , the civilian curve was considerably more complex . Alaska's civilian population , however , is modulated substantially with tourists , which over the course of a year outnumber residents by a factor of two . The two installations in Japan – Misawa and Yokusuka – displayed peaks that coincided with the trailing portion of the bimodal Japanese pandemic , one at the end of each wave . It is possible that , here , military personnel were insulated from the civilian population earlier in each wave . In contrast , the installation at Landstuhl , Germany , provided the only example where a military installation peaked significantly earlier than the civilian population . Here , it is quite possible that the pandemic at the installation was brought by troops recently deployed there from the United States . Thus , the lack of synchronicity at foreign installations can be explained by the fact that such troops mix far less frequently with the surrounding populations . More generally , the analysis presented here could act as a starting-point for the development of more detailed models of different types of military populations and for the systematic identification of a subset of installations that act as accurate sentinels for nearby civilian populations . Our estimate for the basic reproduction number ( mean: 1 . 57 , median: 1 . 38 ) is generally consistent with those found using various civilian data ( e . g . , [11] , [20]–[27] ) , and is relatively clustered ( quartiles: 1 . 27 and 1 . 79 , see Table S4 ) . One might have expected higher values , particularly at installations supporting new recruits , or with on-base families and schools , but this does not appear to be the case . Further , our analyses do not suffer from obvious population selection-bias , as is the case with many early-outbreak studies . Rather , these data originate from routine episode recording for health insurance purposes . Similar data-streams exist in the civilian domain but have less uniform spatial coverage and would be more difficult to make available in real time [28] . The trend for to increase with time is interesting in light of recent work on the seasonality of influenza transmission [29] . Although it is likely that media reports may have driven some individuals to seek treatment when they would otherwise not have done so , and that this effect varied over the course of our study , it would not have affected our description of trends in transmissibility to a large degree . Values for in our analyses were driven by the growth rate of incidence , not by the absolute level of incidence . Therefore , a gradual change in the propensity to report over many months would not affect our reported trend in . More rapid increases during the period of exponential growth at a specific base would affect our results and it is certainly possible that such changes in behavior may have occurred during late April 2009 . However , we would expect those changes to bias upwards during the early part of our study , which is not consistent with the pattern we report . It is intriguing that the two-peak model consistently out-performed ( based on AIC results ) the one-peak model , even for profiles that visually appear to display a single , classic profile consisting of a sharp exponential rise , peak and slower decay . This suggests that even for these apparently straight-forward profiles , there may be some underlying mechanism at work that makes use of the freedom of the extra parameters . It is possible , for example , that changes in behavior or exchange of personnel may sufficiently modulate the basic profiles to the point that a seven-parameter model is appropriate . More generally , these results suggest subtle dynamics around the peak of short-time scale respiratory infections not captured by the very simple saturation process of the classic SIR model . | The ability to rapidly and reliably characterize novel strains of influenza in terms of their transmissibility is crucial for health planners: without good estimates for key parameters it is not possible to identify the appropriate strength of interventions or the spatial optimization of interventions based on variability in transmissibility . While the transmission of influenza in civilian societies has been relatively well-studied , it has received considerably less attention within military populations; yet the consequences , particularly during wartime , are arguably far greater . We have investigated the incidence for the 50 largest military installations in the USA , and , to the extent possible , compared them with the profiles of the enclosing civilian populations during the 2009 influenza pandemic . We infer that the local civilian population drove the timing of peak incidence at the military installations . We also developed and applied a two-peak SIR model to capture the essential properties of the pandemic at each installation , finding that transmissibility tended to increase during the course of the pandemic . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [
"medicine",
"infectious",
"diseases",
"population",
"modeling",
"influenza",
"infectious",
"disease",
"modeling",
"biology",
"viral",
"diseases",
"computational",
"biology"
] | 2013 | Multiple Estimates of Transmissibility for the 2009 Influenza Pandemic Based on Influenza-like-Illness Data from Small US Military Populations |
Dengue is the world's most important mosquito-borne viral illness . Successful future management of this disease requires an understanding of the population dynamics of the vector , especially in the context of changing climates . Our capacity to predict future dynamics is reflected in our ability to explain the significant historical changes in the distribution and abundance of the disease and its vector . Here we combine daily weather records with simulation modelling techniques to explain vector ( Aedes aegypti ( L . ) ) persistence within its current and historic ranges in Australia . We show that , in regions where dengue presently occurs in Australia ( the Wet Tropics region of Far North Queensland ) , conditions are persistently suitable for year-round adult Ae . aegypti activity and oviposition . In the historic range , however , the vector is vulnerable to periodic extinction due to the combined influence of adult activity constraints and stochastic loss of suitable oviposition sites . These results , together with changes in water-storage behaviour by humans , can explain the observed historical range contraction of the disease vector . For these reasons , future eradication of dengue in wet tropical regions will be extremely difficult through classical mosquito control methods alone . However , control of Ae . aegypti in sub-tropical and temperate regions will be greatly facilitated by government policy regulating domestic water-storage . Exploitation of the natural vulnerabilities of dengue vectors ( e . g . , habitat specificity , climatic limitations ) should be integrated with the emerging novel transgenic and symbiotic bacterial control techniques to develop future control and elimination strategies .
Dengue fever is a public health problem of global importance , producing a spectrum of disease spanning febrile arthralgia to hemorrhagic death . Dengue viruses are transmitted between human hosts almost exclusively by Aedes ( Stegomyia ) aegypti and Aedes ( Stg . ) albopictus mosquitoes , both of which are well adapted to using artificial containers for larval habitat . Many urban areas in the tropical world are subject to dengue transmission [1] , the geographic range of which is limited by the distribution of the vectors . However , these ranges are not static , with numerous expansions and retractions recorded through time . Despite great progress in the development of novel control techniques for Ae . aegypti [2] , [3] , our understanding of how dengue and its vectors become extinct is poor . The principal vector , Ae . aegypti , is thought to have originated in Africa and extended its range globally with the expansion of commercial shipping in the 17th and 18th centuries [4] , [5] . While this range was significantly reduced by numerous eradication programs in the Americas from the 1930s to the 1970s [6] , [7] , Ae . aegypti soon regained much of its former range after these programs ceased [7] . An ultimate cause of such range plasticity is human activity . The production of suitable larval habitats ( i . e . artificial containers ) and human-facilitated transport has encouraged the dispersal and establishment of these mosquitoes . Increased urbanisation without properly planned waste management and water handling systems has also created ideal conditions for mosquito breeding [7] . Human activity is thus a key determinant of dengue vector populations . In Australia , dengue transmission is currently restricted to tropical north Queensland ( Qld ) ( Fig . 1 ) . The vector there , Ae . aegypti , is most abundant and active year-round in the tropics , yet its distribution extends into sub-tropical coastal central Qld , and some arid inland areas [8] . Dengue has been recorded in Australia from as early as 1873 [9] , and although outbreaks have been most common in the tropics , sporadic activity has also occurred in the subtropics and temperate regions [10] . This was due in part to the distribution of Ae . aegypti formerly extending well into temperate regions ( up to 33°S in Western Australia ( WA ) ) . However , a range retraction occurred in the last half of the 20th century , with the last collections from New South Wales ( NSW ) in 1948 , and WA in 1970 . The last records from the Northern Territory ( NT ) were from 1956 [11] , with established ( incursant ) Ae . aegypti populations not discovered again until 2004 and 2006 [12] . It also disappeared from southern Qld in the 1950s . The current range has been relatively stable for the last three decades at least . The cause of Ae . aegypti range retraction in Australia has not been resolved , yet is probably related to a number of social improvement factors . In particular , a reduced prevalence of larval habitats ( i . e . water-filled containers ) due to improved reticulated water supplies and a concomitant decrease in domestic rainwater tanks , the decline of steam rail with its attendant water storage infrastructure and potential for dispersal , and , in rural areas in particular , the gradual replacement of domestic food storage cabinets ( e . g . Coolgardie safes with their associated water containers ) by kerosene and then electric refrigerators . Additionally , adult mosquito productivity and survival may have been reduced by greater yard sanitation with the advent of motor mowers limiting trash containers and adult resting sites and the development of residual insecticides ( such as DDT , BHC and dieldrin ) for domestic use . Furthermore , there was enhanced organization of vector control operations by local governments with the return in the late-1940s of well trained public and environmental health officers from military service who were rigorous in their destruction of breeding sites [10] , and the relatively small human population sizes of Ae . aegypti infested areas in many parts of Australia may have facilitated extinction in some places . Finally , there may have been various biological factors that contributed , in some regions at least , to displacement or extinction , such as larval habitat competition from the indigenous ‘container mosquito’ Aedes ( Finlaya ) notoscriptus that was becoming domesticated and gradually spreading westwards in NSW from its native coastal habitats [10] , [13] . However , many of these possible causes of Ae . aegypti range retraction remain speculative and not readily testable . We used computer-based simulation modelling to investigate why Ae . aegypti may have disappeared from much of its former range in Australia that appears still to be climatically favourable [14] , [15] and to determine how well it may persist if reintroduced in the future . Extinction processes have been previously studied through the use of mathematical models [16] , [17] . In a recent application of computer-based modelling [18] , a mechanistic approach was adopted to explain Ae . aegypti distribution in Australia , and described the historic range of the species in terms of its ability to survive in large breeding sites ( rainwater tanks ) . This work demonstrated that large parts of coastal Australia could support survival of the species if such tanks were present , consistent with the historic range , but did not go as far as explaining range retraction . Furthermore , the model used by those authors made use of historic mean climate data , an approach that does not incorporate the stochasticity of daily weather variation that may contribute to extinction processes . Here we describe the use of the Container Inhabiting Mosquito Simulation ( CIMSiM ) to determine the persistence of Ae . aegypti throughout Australia in its current and historic ranges . CIMSiM is a weather-driven depiction of the Ae . aegypti larval and adult habitat that describes the interaction between the mosquitoes and their environment [19] , and has been validated for use in Australia [20] . In addition , we compared the performance of Ae . aegypti in terms of productivity throughout its range , examined the relative prevalence of life stages ( i . e . eggs , larvae , adults ) over time , and examined the relative prevalence of eggs in different habitats for selected localities . In doing this we hoped to explain its current range compared with its more extensive historic one in terms of climate suitability , and to comment on future risk of establishment in areas of Australia that are currently dengue free .
CIMSiM [19] , which accurately models Ae . aegypti population dynamics in Qld [20] , generates daily estimates of egg , larval , pupal and adult numbers per hectare by integrating daily meteorological observations with information about available breeding habitats . Thirteen study locations were selected from both the current and historic Ae . aegypti range [10] , [21] . Simulations were performed for 10 years ( 1998–2007 ) . Model parameters for larval habitats [20] are provided ( Table S1 ) . All other model settings for CIMSiM were default values [19] with the exception of egg survivorship parameters which were modified ( Table S1 ) . The following daily weather observations were used: maximum , minimum and average daily temperature , relative humidity , saturation deficit , and rainfall . These were obtained for each study location from the Australian Bureau of Meteorology ( www . bom . gov . au ) . Ten simulations of 10 years length were performed for each location ( with the exception of Harvey ( WA ) , for which only six years of meteorological data were available ) , realising a total of 1170 simulated years . For each study location , we aimed to characterise the following performance measures for Ae . aegypti:
The simulations described here have allowed a quantitative assessment of Ae . aegypti performance and persistence at localities inside and outside the current range in Australia . Coupled with information about the ecology of larval habitats derived here , explanations of the current and historic range of this disease vector in Australia are possible . The continued presence of Ae . aegypti in its current range in Qld can be explained by its continuing year round adult and larval activity . This is facilitated by the continuous presence of suitable larval habitats , which remain wet enough ( often due to constantly wet containers such as pot plant saucers in these simulations ) and warm enough for year-round activity of all life stages ( mean daily temperatures exceed 15°C year-round in the current north Qld range , www . bom . gov . au ) . An analysis of egg densities at Cairns throughout the dry season ( May – Nov ) ( Fig . 2 ) revealed continued oviposition activity during this period . Conversely , there are a number of factors which make the species vulnerable at localities in the historic range . In tropical Darwin where Ae . aegypti is now extinct , our modelling showed the species to be heavily reliant on manually-filled ( i . e . continuously wet ) containers for activity through the dry season . In our simulations , pot plant saucers ( manually filled ) were the major container type for eggs during the dry season . This represents vulnerability for Ae . aegypti in such locations , in that source reduction activities incorporating vector control and public education programs selectively targeting artificially flooded breeding sites could have a large negative impact on population growth . Kay and others [25] demonstrated that selective control of Ae . aegypti in continuously flooded subterranean wells in Qld reduced recolonization of surface containers during the wet season , reducing overall populations . Such source reduction was evident in Darwin after World War II , when health officers returning from military service set about removal of rainwater tanks concomitant with the establishment of reticulated water supplies ( Peter Whelan , NT Health Department , pers . comm . 22 Apr 2009 ) . Rainwater tanks , while not manually filled , are preferentially filled from rainfall run-off and retain water for extended periods when naturally filled containers may have dried out . We were not able to simulate rainwater tanks here , which limits our ability to interpret their contribution to persistence . However , our inclusion of pot plant saucers ( albeit much smaller but very productive sites ) allow us to assess the importance of continuously wet larval habitats . Field productivity values for Ae . aegypti in rainwater tanks would be useful for future simulation modelling . The importance of continuously wet containers for the persistence of Ae . aegypti in Darwin described here illustrates how vulnerable the species may have been when rainwater tanks were removed en masse post-war . Coupled with the small human population size in Darwin at the time ( approx . 8016 people in the NT in 1948 ) [26] , extinction by a combined action of habitat specificity and a lack of that habitat during the dry season due to source reduction is plausible . Insecticide application does not appear to have played a significant role in this process [11] . Thus , the local extinction of Ae . aegypti at Darwin was likely due to a synergistic combination of processes . This would have included a primary extinction driver ( loss of habitat ) , combined with secondary drivers such as the specialisation of the species for a narrow range of habitat types and physiological vulnerability to dry conditions . Synergistic effects of extinction processes have been well studied for extinction dynamics in other species [27] , [28] . Habitat and host specificity were both factors identified as being significant in the extinction of butterflies [16] . Such specificity is evident in Ae . aegypti , in its strong preference for artificial containers and blood-feeding on humans . While such associations promote the proliferation of the species in human habitats , they also render it vulnerable to changes in such habitat . In modelling for Brisbane , the cool conditions through the austral winter were shown to preclude adult activity , making the species vulnerable in all habitats . Mean daily temperatures at this location are below the threshold for adult activity ( 15°C ) [24] for June – Aug ( www . bom . gov . au ) . Persistence is greatest in continuously wet containers . Considering the timing of its apparent extinction in Brisbane , during the 1950s , the decreasing prevalence of rainwater tanks concomitant with increased reticulated water supplies might explain the disappearance of Ae . aegypti from an area in which it was vulnerable to extinction . Clearly , more than just strong seasonal productivity of Ae . aegypti is required for persistence at a location . This is evident in the very similar productivity values for the species in Brisbane ( where extinction occurs in the model ) and at Charters Towers ( where it does not ) ( Table S2 ) . In our modelling , the main difference between the locations is temperature , which is slightly higher at Charters Towers , permitting longer periods of adult activity and oviposition . This reduces extinction risk due to egg die-off as the egg-only periods are shorter ( Fig . S1 ) . Here we confirm the ability of this species to survive in areas where it no longer exists ( Fig . 1 ) ; a finding consistent with previous distribution reports [10] . However , by demonstrating extinction at some locations , our work challenges the idea that the historic range is climatically suitable for long-term Ae . aegypti survival as has been indicated [14] , [15] . The former presence of Ae . aegypti in many parts of Australia is not questioned here; rather , the ability of CIMSiM to simulate strong periodic productivity in areas where the species was once considered seasonally common but is now extinct provides validation for our approach . The finding that Ae . aegypti passes several months of the year only as eggs ( Fig . 1 ) is consistent with early field reports from southern temperate regions of NSW [29] . A number of possible factors for the retraction of the Ae . aegypti range in Australia have been suggested [10] . Each of these factors could have separately or in common with others plausibly reduced the size of local Ae . aegypti populations and the daily survivorship probability of adult mosquitoes , thereby contributing to local extinction . However , while the widespread introduction of town water reticulation in rural and regional areas has often been proposed as a crucial factor in the disappearance of Ae . aegypti from many southern localities/regions , many houses in many towns in these southern rural areas retained their water tanks throughout and following the period during which the mosquito disappeared , indicating there is no simple explanation that covers all situations . The simulations performed here reveal that when adult daily survivorship probabilities are held high ( 0 . 91 in these simulations ) and suitable breeding containers are available , Ae . aegypti is still vulnerable to extinction , particularly in southern Australia . In the first half of the 20th Century , Ae . aegypti populations were widely sustained in southern Australia , no doubt with the aid of increased numbers of larval habitats and high adult survivorship probability . When these two factors became less prominent , the natural vulnerability of the species ( as demonstrated by simulation here ) could plausibly have led to extinction . Our sensitivity analysis revealed the importance of egg survivorship . Changes in the construction/materials of breeding containers over time ( e . g . a greater proportion of plastic containers with time ) could also have reduced egg survivorship rates . We acknowledge that in applying a CIMSiM model field validated for north Qld across an entire continent we assume that Ae . aegypti performance in relation to temperature , humidity and rainfall remains constant . Furthermore , we also assume a constant breeding site diversity and density throughout Australia , and identical amounts of organic material ( i . e . larval food ) falling into containers . Naturally , we do not anticipate that in the field such generalities will hold true; there will almost certainly be some site-specific variation in local container-breeding mosquito ecology . However , such local scale differences would be very difficult to define accurately , and our approach in applying a constant set of model parameters at different locations ( only differing with local meteorological data ) was the only plausible way for us to model such a range of localities . From our sensitivity analyses , we now understand that changing container densities by up to 20% is unlikely to influence persistence ( at least using the containers simulated here ) . However , persistence is sensitive to changes in egg survivorship rates . Therefore , understanding how such rates are influenced in the field is critical for determining how persistence at a location may vary . Furthermore , the relationship between ambient weather conditions and water temperature in the various container situations that form Ae . aegypti larval habitat has only recently become the focus of study in Australia [18] . The conversion factors for ambient to water temperature built into the CIMSiM program [19] accord well with actual observed temperatures in field-deployed tyres and buckets , albeit with some overestimation of maximum temperature on some days ( MRK unpubl . data ) . Thus , when the water temperature is less than the thermal optimum for Ae . aegypti development , Ae . aegypti productivity in CIMSiM will be overestimated , and when above this threshold , productivity could be underestimated . In addition , our choice of criterion for determining extinction at a location; densities of eggs , larvae and adults <0 . 5 per ha . , could be scrutinized for the absence of pupae . Pupal densities were not included in the criterion , given the relatively short duration of this life stage ( typically 1–3 d ) . However , it is possible ( albeit improbable ) that the pupal stage alone could facilitate persistence at a location when other life stages are at their nadir . In applying the CIMSiM model so widely , we have assumed that on balance , our predictions of Ae . aegypti performance and persistence are satisfactory mid-range estimates that are useful for the kind of population-level analysis presented here . Previous examination of Ae . aegypti range by climate-driven modelling indicated that this species could persist at locations in the historic range ( such as Brisbane Qld and Darwin NT ) in rainwater tanks ( which always retained at least 1 cm of water depth ) , but not in small buckets , which frequently became dry [18] . Modelling of Ae . aegypti distribution using a genetic algorithm [21] also showed suitability of the historic range in the current climate . Our findings , in which Ae . aegypti eggs were most common in Darwin in manually-filled pot plant saucers during the dry season , were consistent with those of previous studies [18] which found that continuously wet habitats were required for persistence at this location . Domestic water storage in tanks is increasing in southern Australia [30] , and in the simulations presented here for southern locations ( Harvey WA , Horsham Vic , Gosford and Wagga Wagga NSW , and Brisbane Qld ) , Ae . aegypti was reduced to existing as eggs only in continuously wet containers . Thus , any increase in water storage behaviour could improve the probabilities of survival of dengue vectors outside of its current range [21] . For this reason , the regulation of water storage behaviour to minimise mosquito breeding is crucial . Areas of northern Australia where Ae . aegypti has become extinct ( e . g . Darwin NT ) remain vulnerable to re-establishment of the species , as evidenced by recent infestations at Tennant Creek and Groote Eylandt ( NT ) . The absence of Ae . aegypti from these areas can only be maintained by adequate surveillance and source reduction activities targeted at manually filled containers ( such as pot plant saucers ) and domestic water storage . According to our modelling , the introduction of a single cohort of Ae . aegypti into southern parts of the historic range in Australia is unlikely to result in a persistent population based on current climate , with container densities similar to that in the current range . Conversely , introductions into northern regions of the historic range ( e . g . Darwin ) may readily lead to persistence of the species . The failure of classical mosquito control methodologies ( e . g . source reduction , insecticide application ) for restricting dengue has stimulated the development of novel molecular strategies [2] , [3] . While there is no doubt such strategies will be integral to the future of dengue control , the natural vulnerability of dengue vectors to extinction should not be forsaken . Incorporating extinction processes into integrated dengue control strategies in the future will ensure a greater probability of success . Furthermore , in subtropical and temperate regions where dengue is a problem , there may be no need for novel , biologically-engineered solutions . | Dengue transmission has not always been confined to tropical areas . In some cases , this has been due to a reduced geographic range of the mosquitoes that are able to carry dengue viruses . In Australia , Aedes aegypti mosquitoes once occurred throughout temperate , drier parts of the country but are now restricted to the wet tropics . We used a computer modelling approach to determine whether these mosquitoes could inhabit their former range . This was done by simulating dengue mosquito populations in virtual environments that experienced 10 years of actual daily weather conditions ( 1998–2007 ) obtained for 13 locations inside and outside the current tropical range . We discovered that in areas outside the Australian wet tropics , Ae . aegypti often becomes extinct , particularly when conditions are too cool for year-round egg-laying activity , and/or too dry for eggs to hatch . Thus , despite being a global pest and disease vector , Ae . aegypti mosquitoes are naturally vulnerable to extinction in certain conditions . Such vulnerability should be exploited in vector control programs . | [
"Abstract",
"Introduction",
"Methods",
"Discussion"
] | [
"infectious",
"diseases/viral",
"infections",
"ecology/population",
"ecology"
] | 2010 | The Extinction of Dengue through Natural Vulnerability of Its Vectors |
With the ability to observe the activity from large numbers of neurons simultaneously using modern recording technologies , the chance to identify sub-networks involved in coordinated processing increases . Sequences of synchronous spike events ( SSEs ) constitute one type of such coordinated spiking that propagates activity in a temporally precise manner . The synfire chain was proposed as one potential model for such network processing . Previous work introduced a method for visualization of SSEs in massively parallel spike trains , based on an intersection matrix that contains in each entry the degree of overlap of active neurons in two corresponding time bins . Repeated SSEs are reflected in the matrix as diagonal structures of high overlap values . The method as such , however , leaves the task of identifying these diagonal structures to visual inspection rather than to a quantitative analysis . Here we present ASSET ( Analysis of Sequences of Synchronous EvenTs ) , an improved , fully automated method which determines diagonal structures in the intersection matrix by a robust mathematical procedure . The method consists of a sequence of steps that i ) assess which entries in the matrix potentially belong to a diagonal structure , ii ) cluster these entries into individual diagonal structures and iii ) determine the neurons composing the associated SSEs . We employ parallel point processes generated by stochastic simulations as test data to demonstrate the performance of the method under a wide range of realistic scenarios , including different types of non-stationarity of the spiking activity and different correlation structures . Finally , the ability of the method to discover SSEs is demonstrated on complex data from large network simulations with embedded synfire chains . Thus , ASSET represents an effective and efficient tool to analyze massively parallel spike data for temporal sequences of synchronous activity .
Synchronous input spikes to a receiving neuron are considered most effective in generating an output spike , as predicted by theoretical studies coining the term coincidence detector [1] . The argument rests on the premise that excitatory post-synaptic potentials ( EPSPs ) in the cortex are typically small in relation to the firing threshold , so that many EPSPs need to overlap to produce an output spike . Due to leak currents in the neuronal membrane , the firing threshold is reached with fewer spikes when these arrive synchronously at the post-synaptic neuron rather than sparsely , thus making the neuron behave like a coincidence detector . Experimental studies provide evidence for the existence of coincidence detectors ( e . g . , [2] ) and relate them to various mechanisms of spike-timing dependent plasticity [3 , 4] as well as to different encoding and decoding schemes [5 , 6] . Cortical anatomy supports such considerations . Individual neurons receive synaptic connections from a large number of neurons ( on the order of 10 , 000 in the human cortex , see e . g . [7] ) and project to a similar number of other cells . Such a connectivity structure combined with suitable synaptic delays may produce spatio-temporal spike patterns , as proposed in [8–11] . A simple case is represented by a temporal sequence of synchronous events ( SSE ) , each event consisting of synchronous spikes from a group of neurons . The synfire chain model [8] is a neural network model that has been proposed to exhibit such activity through a suitable wiring of the neurons in the network in successive groups interconnected in a highly divergent and convergent manner . Each neuron of one group projects to several neurons of the next group , thus forming a chain structure . Under the assumption that the synaptic transmission delays from neurons of one group to neurons of the next group are identical , the synchronous stimulation of neurons in the first group leads to robust propagation of synchronous spiking activity through the chain [10] even in the presence of noise . The activation of a synfire chain would thus lead to the occurrence of an SSE . In order to assess whether SSEs are indeed observed in the brain and have a functional role , the spiking activity of several neurons needs to be recorded simultaneously , analyzed for temporal correlation and related to behavior . Under the assumption that an SSE occurs sparsely in a given data set , for instance in relation to a specific behavior , pairwise correlations between neurons engaged in the activity are weak and will not be detected by means of pairwise correlation analyses . For this reason , an analysis method was proposed in [12] that directly searches for SSEs in massively parallel spike data . The basic idea of the method presented in [12] is the following: after discretizing time in bins of a few ms ( see Fig 1A ) , a synchronous event which repeats at two different time bins leads to a large number of neurons that are active in both bins . Building an intersection matrix I where each entry Iij represents the number of neurons active in both bins i and j , a group of synchronous events occurring in bins i and j results in a large overlap Iij compared to other entries of the matrix . An SSE which occurs twice produces in the matrix I a sequence of high-valued entries , which we name diagonal structures ( DS ) , aligned parallel to the main diagonal ( Fig 1B ) . A diagonal filter applied to the matrix I enhances the contrast between the DS and surrounding entries , mapping I into a filtered matrix F . The method was calibrated using data from large-scale synfire chain network simulations . When the neurons of a full chain or a large portion of it are observed , the method reveals the associated DS in the intersection matrix . However , when only a few hundreds of neurons ( comparable to the number of cells that can be recorded simultaneously in vivo with modern electrophysiological techniques ) are randomly sampled from the full network , the DSs are less visible and less likely to be continuous , making it impossible to isolate them from the surrounding entries . Indeed , this method leaves the judgment of whether individual entries are large or small , as well as the grouping of proximal entries into DSs , to visual inspection . As a consequence , the results are prone to subjectiveness and the procedure is not open to automation . In this paper we present a method , named ASSET ( Analysis of Sequences of Synchronous EvenTs ) , which improves the approach proposed in [12] by providing a mathematical and fully automated detection of SSEs in parallel spike train data . The analysis features i ) a statistical assessment of membership of individual entries of the intersection matrix to a DS , ii ) a rigorous construction of individual DSs and associated SSEs by clustering , and iii ) the reconstruction of the neuronal composition and occurrence times of each event in the found SSEs . The manuscript is organized as follows . In “Methods” we derive a statistical assessment of the membership of matrix entries to a DS based on two statistical tests for the significance of each individual entry and the joint significance of an entry and its neighbors , respectively . Entries passing both tests are then grouped together into individual DSs depending on their reciprocal distance by a clustering procedure . Once the DSs are identified , it is readily possible to reconstruct their neuronal composition . In “Results” we assess the performance and robustness of ASSET on various types of simulated data that replicate typical features of electrophysiological recordings . These include firing rate heterogeneity across neurons , variability over time and different types of correlation structure in the spiking activity . In addition , we demonstrate the ability of the ASSET analysis to find repeated SSE activity in simulated data generated by a neural network model of overlapping synfire chains as introduced in [12] . We conclude by discussing benefits and limitations of the ASSET method , open problems and future plans for the analysis of electrophysiological data .
Our approach builds on the notion of intersection matrix defined in [12] . Here we introduce this concept mathematically and provide definitions . Given a set of N parallel spike trains observed in the time interval [0 , T] , synchronous spike events across neurons are determined by discretizing the time interval into B adjacent time bins b1 , b2 , … , bB of identical width Δ = T/B ( typically of a few ms ) , as illustrated in Fig 1A . Each set Si of spikes falling in time bin bi forms a synchronous event . The value |Si ∩ Sj| represents the number of neurons being active at both time bins bi and bj . The intersection matrix I is defined by I i j ≔ | S i ∩ S j | ∀ i , j = 1 , 2 , … , B . In this setting I is a symmetric matrix , whose main diagonal contains the population histogram , i . e . the time histogram of the number of neurons simultaneously active in each time bin . A synchronous spike event occurring at two bins i and j , e . g . as a result of the repeated activation of the same synfire chain , results in a larger value for Iij compared to the chance level . A repeated SSE composed of lSSE successive synchronous events , each of which occurs at time bins ( bir , bjr ) , r = 1 , 2… , lSSE , determines a sequence of large-valued entries Iir , jr in I—which we term a diagonal structure ( DS ) —as sketched in Fig 1B . To account for the variability of firing rates over time when comparing different entries of I , in [12] each entry of Iij was normalized by the number of neurons active in each bin bi and bj . After normalization , the entries take value 1 for a complete overlap and value 0 for no overlap . To enhance the contrast between high-valued entries belonging to the same diagonal structure and the surrounding entries , the matrix was further filtered by a linear kernel having an orientation parallel to the main diagonal . We take here a different approach , as outlined in the following section . To derive a measure of overlap that is independent of firing rates , we first need the probability mass function ( pmf ) pij ( ⋅ ) of each individual entry Iij in the intersection matrix , under the null hypothesis H0 that the spike trains under consideration are realizations of mutually independent Poisson processes . If the null hypothesis is rejected , i . e . if the observed value ξ taken by Iij is too large to be interpreted as chance , we classify the overlap Si ∩ Sj as a statistically significant repeated synchronous event . In “Results” we show that the statistics of the method are robust to deviations from Poissonianity as well as to the presence of various types of correlations other than repeated SSE activity . Under the stated hypotheses , the distribution of Iij is determined by the firing rate of each neuron k at the time bins bi and bj . Iij represents the ( stochastic ) number of neurons firing simultaneously in both bins and can thus be expressed as I i j = ∑ k = 1 N I i j ( k ) , ( 1 ) where I i j ( k ) is a Bernoulli random variable taking value 1 if neuron k fires in both bins i and j , and 0 otherwise . The probability parameter p i j ( k ) of I i j ( k ) is related to the local firing rate λ i ( k ) of neuron k at bin bi and λ j ( k ) at bin bj by p i j ( k ) = ( 1 - e - λ i ( k ) Δ ) ( 1 - e - λ j ( k ) Δ ) for each i ≠ j , where each factor 1 - e - λ i ( k ) Δ is the probability for neuron k to emit at least one spike in the time bin i . The knowledge of the firing rate profiles of all neurons is therefore a prerequisite for the exact computation of the pmf pij ( ⋅ ) of Iij , i , j = 1 , … , N . In light of Eq 1 , Iij is a Poisson Bernoulli random variable [13] . Its pmf pij ( ⋅ ) is analytically given by p i j ( ξ ) = ∑ A ∈ P N ; ξ ∏ k ∈ A p i j ( k ) ∏ h ∈ A C ( 1 - p i j ( h ) ) , ( 2 ) where P N ; ξ is the family of all possible subsets of ξ elements that can be extracted from the set {1 , … , N} , A is one such subset and AC = {1 , … , N}∖A is the complement of A . Thus pij ( ξ ) is a summation of ( N ξ ) addenda , for a total of 2N terms needed to compute pij ( ⋅ ) . The computation is feasible for small N , but soon becomes prohibitive as N increases beyond a few dozens , as in the applications to large parallel recordings we are interested in . By use of Le Cam’s theorem [13] , we approximate pij ( ⋅ ) by a Poisson density function p ( λ ) with rate parameter λ = ∑k λk p i j ( ξ ) ≃ p ( λ ) ( ξ ) = λ ξ e - λ ξ ! . ( 3 ) The approximation error grows quadratically with the p i j ( k ) ’s and stays low if the p i j ( k ) ’s are sufficiently small , as shown in Fig 2 . Given the firing rate profile ( i . e . the rate at each time bin ) of each neuron , we can thus calculate the pmf pij ( ⋅ ) and its cumulative distribution function ( cdf ) P i j ( · ) ≔ ∑ ξ < · p i j ( ξ ) , either exactly from Eq 2 or approximately relying on Le Cam’s approximation in Eq 3 . Transforming each entry Iij by its respective cdf P i j ( · ) , we map the observed overlaps Iij to cumulative probabilities P i j ≔ P i j ( I i j ) = ∑ ξ < I i j λ ξ e - λ ξ ! , ( 4 ) obtaining the probability matrix P ≔ ( Pij ) ij , as illustrated in Fig 3A . If the null hypothesis holds , then P i j ( · ) is the true probability distribution of the amount of overlap between bins bi and bj , and Iij is a realization from that probability distribution . If so , P i j : = P i j ( I i j ) takes N + 1 values x ∈ [0 , 1 ) ( as many as the intersection values from 0 to N ) , and its cdf is the identity function over the domain of this set of values: Pr ( Pij < x ) = x for any x . If Pij is large ( close to 1 ) , the null hypothesis is rejected in favor of the alternative hypothesis that the observed overlap reflects active synchronization between the involved neurons at time bins bi and bj . After setting a significance threshold α1 ( e . g . α1 = 0 . 99 ) , we classify all entries Iij for which Pij > α1 as statistically significant , along with the associated repeated synchronous events . Note that the sum in Eq 4 includes only values of ξ strictly lower than Iij . This choice , which assigns the observed value Iij to the critical region of the hypothesis test , ensures that 1 − Pij retains the property of a p-value , namely that Pr ( 1 − Pij ≤ y ) = y for each y . 1 − Pij reflects the probability , computed under the null hypothesis , that Iij would take a value equal to or exceeding the observed value . We reject the null hypothesis if this probability is lower than 1 − α1 = 0 . 01 . A DS resulting from a repeated SSE , as sketched in Fig 1B , differs from the surrounding entries of the intersection matrix I due to not just one , but a sequence of entries with large values aligned around a diagonal . We devise a statistical test that exploits this feature to detect DSs in the intersection matrix . In the previous section we have already derived the probability matrix P as a transformation of I that normalizes raw intersection values by the local neuronal firing rates . We can now look for DSs in P rather than in I . Each entry Iij in the raw intersection matrix is tested for its individual significance and for the joint significance of its neighbors . If both tests pass , i . e . if Pij > α1 and Jij > α2 , the entry is classified as belonging to one DS . Such entries are collected in a binary masked matrix M that takes value 1 if both tests pass , and 0 otherwise , M i j ≔ 1 { I i j > α 1 } · 1 { J i j > α 2 } , ( 8 ) as illustrated in Fig 3C . It remains to be established which entries belong together to the same DS . Intuitively , entries in the masked matrix that take value 1 belong to the same DS if close-by , and to different DSs if far apart . The masked matrix in Fig 3C shows for instance two clearly separated DSs . A suitable notion of “distance” for matrix entries should make entries falling in the same diagonal , i . e . aligned along the natural direction of a DS , closer together than entries aligned along the anti-diagonal . We introduce the following elliptic distance between any two matrix entries ( i1 , j1 ) and ( i2 , j2 ) : d ρ ( ( i 1 , j 1 ) , ( i 2 , j 2 ) ) ≔ 1 + ( ρ - 1 ) · sin θ - π 4 · ( i 2 - i 1 ) 2 + ( j 2 - j 1 ) 2 / 2 , where θ = arctan ( j 2 - j 1 i 2 - i 1 ) is the angular coefficient of the line intersecting ( i1 , j1 ) and ( i2 , j2 ) , the first square root factor is the Euclidean distance between the two points and ρ ≥ 1 is a stretching factor for angular coefficients deviating from π/4 . d ( ⋅ ) grows as θ approaches 3π/4 or − π/4 , i . e . the anti-diagonal orientation . For instance , dρ ( ( i , j ) , ( i+k , j+k ) ) = k for any ρ , dρ ( ( i , j ) , ( i+k , j − k ) ) = ρk and d ρ ( ( i , j ) , ( i , j + k ) ) = [ 1 + 2 2 ( ρ - 1 ) ] k . We set ρ = 5 and , based on the distance d5 ( ⋅ , ⋅ ) of their positions in the matrix , group all entries Iij with Mij = 1 ( see Eq 8 ) into clusters via a density based scanning ( DBSCAN ) algorithm [14] . The algorithm considers two entries as part of the same neighborhood if their distance is not larger than a maximum value ε . Neighborhoods sharing an entry are joined together and eventually classified as a cluster if they contain a minimum number l0 of entries . We set ε = 3 . 5 , thus allowing for a maximum of ⌊ε⌋ − 1 = 2 holes between two consecutive entries of a DS along the main diagonal , and l0 = 3 , thus requiring a DS to reflect at least 3 repeated synchronous events . The elliptical neighborhood used when clustering should be contained into the kernel used to build the joint probability matrix J . The reason is that the first defines the “immediate” neighbors of the entry , while the second is meant to cover all entries that may belong to the same DS . Thus , the parameter ε should not be chosen larger than the kernel length lK , while the stretching coefficient ρ should be chosen such that the shorter axis of the ellipse fits into the kernel width wK . The values ε = 3 . 5 and ρ = 5 we set for the validation satisfy these requirements . Fig 6 illustrates the matrix entries falling inside the ellipse ( red dots ) for various choices of the parameters ρ and ε , and the kernel ( gray area ) centered around the same entry . Entries in the matrix M not belonging to any cluster are discarded as events that do not reflect repeated SSE activity . Fig 3D shows the cluster matrix C assigning value 1 ( colored in black ) to entries belonging to a cluster , and 0 ( white ) to the others . The matrix contains two clusters composed of 5 entries each , corresponding to the 5 synchronous events highlighted in Fig 1A . Calculating the entries of the probability matrix as given by Eqs 2 and 3 requires the knowledge of the firing rate of each neuron over time . Firing rate estimation is a problem that has been targeted by a number of studies . The peri-stimulus time histogram ( PSTH , [15] ) is an estimate of the firing rate performed by discretizing time into adjacent bins and by counting the number of spikes falling into each bin . The bin width for the PSTH is typically larger ( tens of ms ) than the one used to define synchrony , as firing rates change on a slower time scale . The larger the bin width , the coarser but less biased the estimate . Normalizing the spike count in each bin by the bin width yields the rate of the process , i . e . the number of spikes per time unit . Kernel convolution [16] replaces each spike with a kernel ( a probability density function ) centered around the spike time and estimates the firing rate by the sum of these distributions . Formally , this is done by representing the spike time by a Dirac delta function centered around the spike time t* and by convolving it with the kernel . Following a standard choice , we specify the kernel as a normal distribution with assigned standard deviation σ , truncated at ±2 . 7σ to yield a finite support . When setting the kernel width w* = 5 . 4σ to a fixed value , we employ w* = 200 ms . Both PSTH and kernel convolution can be applied to the case when multiple independent , identically distributed trials of the activity of a neuron are available , by averaging the estimates obtained for each trial . For identical bin and kernel widths , the PSTH typically better represents sharp changes in the firing rate from one bin to the next , while kernel convolution yields smoother curves . Both estimates are parametric and require the choice of a bin- ( kernel- ) width . Methods have been recently proposed to determine the optimal bin- or kernel-width by minimizing the error between the true ( unknown ) rate and its estimate in some statistical sense ( see [17 , 18] ) . These methods have been shown to outperform their fixed width variants and are particularly helpful when analyzing parallel spike train data from different neurons with different rates , where the optimal bin width varies across neurons . In the “Results” we compare the performance of ASSET employing either PSTH or kernel convolution or optimized-width kernel convolution [18] estimates of the firing rate profiles over an increasing number of trials .
To assess the quality of the method , we first measure its performance in the case when all assumptions entering the derivation of Eq 2 ( alternatively Eq 3 ) and Eq 6 are met . In addition , we test the robustness of the statistics of the method with respect to deviations from these assumptions which are typically found in experimental data . In particular , we investigate how the following features of the data affect the performance of the method and test if these lead to false positive ( FP ) outcomes . The first four features relate to aspects of firing rates , such as various types of non-stationarities and rate correlations , i . e . correlations between the spike trains on a slower time scale than SSEs . The last two features relate to spike synchrony , however with a different organization than in SSEs . For the concrete test cases we formulate 10 different stochastic models of spiking activity ( see Fig 8 as examples of the realizations ) , each including only one or a combination of the above-mentioned features , as summarized in Table 1 . We use these models to generate background activity into which we subsequently embed the spiking activity corresponding to a repeated SSE . We provide here the definition of each model . The computational cost of the algorithm is almost entirely determined by the time required to evaluate the joint probabilities defined in Eq 6 . The expression involves a nested sum of several terms . The number of terms grows with the number n of matrix entries covered by the kernel ( determined by the kernel length lK and the kernel width wK ) and with the number d of largest neighbors among which the joint significance is computed . These are free parameters of the analysis , whereas features of the data , such as the number of neurons or their firing rates , do not influence this step of the computation . As an estimate , for the values we employed in the manuscript ( lK = 5 , wK = 5 , d = 5 ) the evaluation of a single entry took about 10 ms on a single core of a dual AMD 12-core Opteron 6174 machine with 64GB RAM using the Python code provided with this manuscript . Thus , the evaluation of a full matrix J of 200 × 200 entries took on average less than 7 minutes . However , the fact that single entries are evaluated independently may be easily exploited by parallelizing the analysis on multi-core machines or computer clusters , where each worker process is assigned to perform the computation for a subset of the matrix entries ( see [37] ) .
Temporal sequences of synchronous spike events ( SSEs ) have been postulated as a working mechanism of activity propagation in the cortex [38–40] . The present manuscript introduces a novel statistical method for the detection of SSEs in massively parallel spike train data , named ASSET ( Analysis of Synchronous Spike EvenTs ) . The method is inspired by a visual technique first proposed in [12] , which represents the repeated occurrence of an SSE as a sequence of large entries along the diagonals of an intersection matrix ( diagonal structure , or DS ) that indicates for any two time bins the number of neurons firing in both bins . ASSET automatizes the detection process of the original visual technique , assesses the statistical significance of SSEs by exploiting the multiple evidence of its events to derive their joint significance , and determines the structure and neuronal composition of the identified SSEs . In evaluating the null distribution , the method accounts for the temporal profiles of the firing rates of the observed neurons . As such , it detects SSEs which cannot be explained on the basis of rate coding mechanisms , and thus arise from spike correlations on a shorter temporal scale . Rate correlation is understood as a conceptually different mechanism of information processing than spike synchrony ( as for SSEs ) , because it corresponds to stochastic ( probability <1 ) rather than reliable ( probability 1 ) neuron activation [29 , 41–43] . We assessed the performance of ASSET in terms of false positive ( FP ) DSs found in various types of stochastic models which mimicked typical features of neuronal spike trains , such as variable firing rates in time or across neurons , different inter-spike interval distributions , and correlation structures differing from SSEs . We then additionally injected repeated SSE activity in the data to assess the power of the method in terms of true positive ( TP ) detections . The analysis performs two statistical tests on each pair of time bins , which amounts to tens of thousands of tests for a stochastic simulation of 1 s binned at 5 ms . In addition , entries passing the two tests have to lie close to each other in the matrix in order to be clustered into a common DS . To avoid FPs , the statistical threshold needs to be set to low values ( here , 10−5 ) . This is possible without incurring into large levels of FNs because the joint tail probabilities associated to the second test are very low ( typically <10−5 , and as low as 10−12 ) for the entries corresponding to the embedded SSEs . Indeed , the method shows high performance , i . e . FP and FN rates both close to 0 . We did not need to further correct the statistical thresholds by the amount of tests performed ( e . g . by Bonferroni or FDR correction ) , because already the set of combined requirements that need to be fulfilled to identify a DS ( success of the two tests for matrix entries , and proximity of these entries in order to be clustered into a DS ) makes the analysis very conservative . The underlying null hypothesis assumes independent Poisson spike trains , which enables an analytical formulation of the test statistics . The method proves to be robust to deviations from Poissonianity , such as a higher regularity of the inter-spike intervals , which may be observed in experimental data [22 , 44] . It is also selectively sensitive to SSEs , but not to other models of spike correlation , such as synchronous events not organized in a temporal sequence . Nevertheless , anticipating scenarios where strong correlations not forming an SSE might indeed bias the statistics , we proposed a Monte-Carlo approach to account for these correlations . To this end , we constructed the null hypothesis by estimating the probability to find a certain degree of pattern overlap from the repeated generation of surrogate intersection matrices . In our test data the Monte-Carlo approach yielded results comparable to the analytical approach , yet at a considerably higher computational cost . Furthermore , the method was able to distinguish SSEs from repeated precise temporal sequences of sharp , local rate transients from one group of neurons to another ( rate propagation ) . Rate peaks increased the probability of the involved neurons to spike , which remained nevertheless a stochastic event , in contrast to SSE activity . Propagation of rate transients thus generated spike patterns that were different in composition ( due to the stochastic activation of neurons ) and less precisely timed than SSEs , but more closely resembled the latter as the rate modulation became higher and faster . It is likely that , for extremely high firing rates and very short rate jumps , ASSET would not distinguish the two models . The distinction between rate correlation and synchrony correlation ( see e . g . [29] ) has been formally questioned in [30] . Users of ASSET may want to identify waves of co-modulating rates like those defined in our model 6 as SSEs , rather than rejecting them . This is possible by decreasing the statistical thresholds α1 and α2 to less strict values than those used in this manuscript . ASSET critically relies on the estimation of the firing rate profile of each neuron to compute the expected overlap of neuron IDs in the intersection matrix and thus to estimate the significance of the observed overlap . Estimating firing rate profiles on the basis of single trial spike data typically requires some kind of sliding window approach , such as convolution of each spike with a temporal kernel [16] . Temporal averaging smears out peaks of the underlying original firing rate that occur on a shorter time scale than the window width , and creates artificial peaks if the window width is excessively short . Single-trial rate estimates obtained by kernel convolution in the presence of time-stationary firing rates yielded high performance of ASSET , but led to impaired performance when the rates were non-stationary in time on a fast time scale , in particular if the rate excursion was coherent across neurons . The reason is that smoothing by convolution underestimates positive rate peaks and thus the expected overlap , yielding FPs . Estimating the firing rates on the basis of trial averages solved the problem . We here tested three such approaches , namely the peri-stimulus time histogram ( PSTH , [15] ) , a trial-averaged kernel convolution with fixed kernel width [16] and a trial-averaged kernel convolution with an optimized kernel width [18] . Already a small number of trials reduced the FP rate considerably for all three methods , although best performance was achieved for the optimized-width kernel convolution . Importantly , cross-trial rate estimation works under the assumption of identical rate profiles across trials . Deviations from this assumption lead to a wrong estimation of the rate in single trials , that is required to calculate the probability matrix , and thereby enhances the FPs . This bias is amplified if neurons exhibit cross-trial variability in a coherent manner [19] . Latency variability is a special instance of cross-trial non-stationarity which causes a mis-estimation around the rate onset . In some cases the onset variability of rates can be corrected for by choosing a more proper alignment of trials , e . g . to the stimulus or behavioral event related to the rate change [45] . If this is not possible , we suggest to generate the probability matrix P on the basis of surrogate spike data , e . g . by spike train shifting or spike dithering [20] . The details of such an approach , however , still need to be explored . We further investigated how the performance of ASSET relates to various other parameters of the SSEs , such as the number of its sequential synchronous events composing the SSE , the number of neurons in each synchronous event , and the total number of observed neurons . SSEs were statistically more significant and therefore easier to detect when they involved a larger fraction of the total neurons . However , they could be retrieved even when employing sub-optimal parameters such as a kernel length smaller than the length of the DS , given that the SSE involved enough neurons . In contrast , the method did not detect spatio-temporal patterns , i . e . a special case of SSE where each event is composed by a single spike only . Spatio-temporal patterns in a more general sense ( with different time intervals between spikes ) were suggested as signatures of synfire chain activity in data of low numbers of simultaneously recorded neurons [40 , 46 , 47] . A less constrained model of cortical processing ( synfire braid ) was proposed in [48] , which incorporates synchronous input to individual neurons as in the synfire chain model , however transferred by connections of different temporal delays compensated by differences in their activation times . This model was further analyzed in [11] and termed polychronization . Since spike synchrony occurs with a temporal lag in this framework , one expects that , although ASSET is not designed to capture this type of coordinated spiking activity , it may still detect signatures of such activity by choosing a correspondingly larger bin width . We aim to explore such a scenario and other applications of ASSET to data from different types of network simulations that exhibit correlations on a fine temporal scale to study the potential of our method in identifying features of the underlying network model . Importantly , increasing the number of neurons to be analyzed does not increase the computational cost of the method . Indeed , in order to evaluate significance values , the method relies on expressions involving a mere sum of the firing rates of individual neurons , which are virtually instantaneous to evaluate . Rapid advances in electrophysiology will soon enable the simultaneous recording of thousands of neurons [49] . These data promise to expose concerted mechanisms of neuronal coding that remained invisible so far . Our method is designed to keep up with these advances , and to be applicable to the next generation of large-scale recordings of spike data . If an SSE occurs more than two times , it generates multiple DSs in the intersection matrix , each corresponding to one pair of occurrences . In [32] it was suggested to compute the overlap between triplets ( or n-tuples ) of bins rather than pairs , and generate a corresponding n-dimensional intersection matrix to visualize DSs in three ( or n ) dimensions . It is possible to extend this approach to ASSET and to exploit this higher-order evidence to increase the power of the method . This extension will be considered in future work . Finally , we demonstrated the efficacy of ASSET on data of large-scale simulations of a balanced random network with embedded synfire chains , which were previously generated and analyzed in [12] with the original visual method . ASSET fully reconstructed the synfire chains active in the considered time period and did not return additional FPs . Differently from the original technique in [12] , SSEs were here determined on the basis of their statistical significance and the results were obtained by an automated analysis workflow . When analyzing real data , some parameters of the analysis such as the statistical thresholds α1 and α2 should be chosen optimally to minimize the risk of FPs , while at the same time not being excessively strict and thus missing true SSEs . Optimal values for the statistical thresholds can be inferred from independent surrogates of the original data , which can be created by one of several approaches , such as spike train shifting or spike dithering [50] . Such surrogate data contain slightly displaced spikes as compared to the original data , such that correlations in the original data ( and in particular SSEs ) are intentionally destroyed while other features of the data ( e . g . firing rates or inter-spike interval regularities ) are preserved . These uncorrelated data can be used to determine the expected value of entries in the probability and joint probability matrices under independence , and therefore to determine lower bounds for the thresholds α1 and α2 which ensure avoidance of FPs . Taking the least conservative of such values ( the lower bounds ) also minimizes the risk for FNs . Suitable values for other analysis parameters can be determined analogously . As illustrated in “Methods” , some of these parameters are tightly associated to putative features of the searched SSEs ( for instance , the kernel length and kernel width to the temporal span and the wiggliness of the SSE , respectively ) , and may therefore be tuned in order to optimize the detection of SSE swith specific characteristics . Taken together , these results demonstrate that ASSET is a reliable and effective tool to detect and identify repeated sequences of synchronous spiking activity in massively parallel spike train data . Whether synchrony propagation constitutes a mechanism of information processing in the neuronal circuitry still remains an open question , that belongs to the more general debate about the role of fine temporal coding versus rate coding . Convincing theoretical arguments as well as experimental evidence have been provided for both processing mechanisms ( see e . g . [2 , 6 , 51] vs [26 , 52 , 53] ) . However , observing fine-scale temporal correlations requires the simultaneous observation and analysis of sufficiently large portions of the involved neuronal circuitry: the severe subsampling of such circuitry so far characterizing most available data prevents this analysis . Next-generation recording technology promises to expose this concerted activity , whose analysis may finally resolve the long-standing question about the role of millisecond-precise spike correlations in cognitive processing . Our work gives a contribution in this direction by providing , to the best of our knowledge , the first tool for a statistical analysis of synchrony propagation applicable to data of hundreds or thousands of simultaneously recorded neurons . The ASSET analysis method is available as part of the Electrophysiology Analysis Toolkit ( Elephant; http://neuralensemble . org/elephant/ ) . In future work we plan to employ the method to investigate the presence of SSE activity in electrophysiological recordings from awake animals and study SSE occurrence in relation to behavior . | Neurons in the cerebral cortex are highly interconnected . However , the mechanisms of coordinated processing in the neuronal network are not yet understood . Theoretical studies have proposed synchronized electrical impulses ( spikes ) propagating between groups of nerve cells as a basis of cortical processing . Indeed , animal studies provide experimental evidence that spike synchronization occurs in relation to behavior . However , the observation of sequences of synchronous activity has not been reported so far , presumably due to two fundamental problems . First , the long-standing lack of simultaneous recordings of large populations of neurons , which only recently were enabled by advances in recording technology . Second , the absence of proper tools required to find these activity patterns in such high-dimensional data . Addressing the second issue , we introduce here a fully automatized mathematical method that advances an existing visual approach to identify sequences of synchronous events in large data sets . We demonstrate the efficacy of our method on a range of simulated test data that capture the characteristics and variability of experimental data . Our tool will serve future studies in their search for spike time coordination at millisecond precision in the brain . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [
"action",
"potentials",
"medicine",
"and",
"health",
"sciences",
"neural",
"networks",
"applied",
"mathematics",
"membrane",
"potential",
"electrophysiology",
"neuroscience",
"simulation",
"and",
"modeling",
"algorithms",
"probability",
"distribution",
"mathematics",
"convolution",
"research",
"and",
"analysis",
"methods",
"computer",
"and",
"information",
"sciences",
"mathematical",
"functions",
"animal",
"cells",
"mathematical",
"and",
"statistical",
"techniques",
"probability",
"theory",
"cellular",
"neuroscience",
"cell",
"biology",
"kernel",
"methods",
"neurons",
"physiology",
"biology",
"and",
"life",
"sciences",
"cellular",
"types",
"physical",
"sciences",
"cognitive",
"science",
"neurophysiology"
] | 2016 | ASSET: Analysis of Sequences of Synchronous Events in Massively Parallel Spike Trains |
Schistosomiasis is one of the most disabling neglected tropical diseases , ranking second in terms of years lived with disability . While treatment with the drug praziquantel can have immediate beneficial effects , reinfection can occur rapidly if people are in contact with cercaria-infested water . Water treatment for schistosomiasis control seeks to eliminate viable cercariae from water , thereby providing safe alternative water supplies for recreational and domestic activities including laundry and bathing . This provision may reduce contact with infested water , which is crucial for reducing reinfection following chemotherapy and cutting schistosome transmission . A qualitative systematic review was carried out to summarize the existing knowledge on the effectiveness of water treatment in removing or inactivating human schistosome cercariae . Four online databases were searched . Studies were screened and categorized into five water treatment processes: storage , heating , chlorination , filtration , and ultraviolet ( UV ) disinfection . All five water treatment methods can remove or inactivate cercariae in water , and hence produce cercaria-free water . However , reliable design guidelines for treating water do not exist as there are insufficient data . Overall , the review found that cercariae are inactivated when storing water for 10–72 hours ( depending on temperature ) , or with chlorination values of 3–30 mg-min/l . UV fluences between 3–60 mJ/cm2 may significantly damage or kill cercariae , and sand filters with 0 . 18–0 . 35 mm grain size have been shown to remove cercariae . This systematic review identified 67 studies about water treatment and schistosomiasis published in the past 106 years . It highlights the many factors that influence the results of water treatment experiments , which include different water quality conditions and methods for measuring key parameters . Variation in these factors limit comparability , and therefore currently available information is insufficient for providing complete water treatment design recommendations .
Schistosomiasis is a water-borne helminthic disease caused by schistosomes , which are parasitic worms . Infection occurs through dermal-contact with cercaria-infested freshwater . Cercariae are released by snails infected with miracidia which hatch from the eggs in human urine and feces . Despite efforts to control the disease , it remains a major public health problem . Disability-Adjusted Life Year ( DALY ) estimates have in fact increased over the past 20 years , exceeding 3 . 3 million DALYs according to the Global Burden of Disease Study 2010 [1 , 2] . In 2012 , the World Health Assembly ( WHA ) adopted resolution 65 . 21 aiming for the elimination of schistosomiasis . This resolution emphasizes the need for multi-sectoral rationale including chemotherapy , strengthening health systems , WASH , and snail control . The need for an integral approach to tackling schistosomiasis has been echoed by many researchers [3–8] . Schistosome cercariae are approximately 500 μm in length , up to 64 μm in width , and are released into freshwater by intermediate host snails [9 , 10] . These snails can shed hundreds of cercariae a day–approximately 200 for S . haematobium , 250–600 for S . mansoni , and 15–160 for S . japonicum [11 , 12] . Schistosome cercariae are attracted to the human host through skin chemicals , temperature gradient , and turbulence [13 , 14] . They are non-feeders , and depend on endogenous glycogen reserves [15] . Cercariae have a forked tail which they use to propel forward and penetrate the host skin . Inside the host , cercariae develop into schistosomula , which eventually develop into schistosomes . The full schistosome lifecycle was first described in 1908 [16] , and interest grew subsequently as troops fighting on African territory were becoming infected [17] . Initially , the control measures focused on mollusciciding and preventing contact with contaminated water ( such as described in the War Memoranda from 1919 [18] ) . The enthusiasm for snail control led to the development of molluscicides ( pesticides against snails ) , which were used to kill all schistosome lifecycle stages in water [19] . The research on water treatment and snail control was slowed by the development of effective orally-administered drugs in the late 1970’s [17 , 19] . Ever since , chemotherapy has been the focus of schistosomiasis control programs [8] . The schistosome lifecycle can be cut through chemotherapy , intermediate host control , and WASH . The orally administered drug praziquantel is used to treat schistosomiasis by killing adult worms in the human host . However , it does not prevent reinfection . Snail control reduces the number of intermediate host snails , and is commonly applied through mollusciciding . Although snail control has been shown to reduce disease burden , there are environmental risks associated with molluscicides [19–21] . Universal sanitation aims to prevent eggs found in human feces from entering water bodies , thereby lowering the number of miracidia . However , Grimes et al . found no significant association between school sanitation adequacy and S . mansoni infection intensity in a national survey in Ethiopia [22] . The positive effect of sanitation may only be realized when the entire community adopts the sanitation infrastructure; one miracidium is sufficient to infect a snail , which may release up to 20 , 000 cercariae in its lifetime , and as a result , only few miracidia are needed to maintain the schistosome lifecycle [12] . Increased hygiene may also impact schistosomiasis control due to the toxicity of soaps to miracidia and cercariae . Use of soaps has been linked with lower prevalence , especially in women who are more exposed to activities such as doing laundry which use soap [23] . Similarly , access to safe water has been found to reduce schistosomiasis prevalence ( safe water implying cercaria-free water ) , as it reduces skin contact with contaminated water [24–26] . The options for providing cercaria-free water are either treating the water , or providing an alternate safe source of water ( such as boreholes or rainwater ) . In endemic regions , there is often no safe alternative water source , and hence water treatment is required to provide a safe water supply . Access to safe water reduces the contact with cercaria-infested water , and may also reduce the risk of water being contaminated with human excreta , as people are spending less time at transmission sites . Water treatment can provide safe water for domestic use such as washing , laundry , and bathing—activities which are associated with long water-contact periods , and therefore higher risk of schistosomiasis contraction . The community may still have to collect water from transmission sites in order to treat it , but water collection only exposes small areas of the body for short periods and is therefore likely to be less critical . Providing a community with water treatment infrastructure may consequently remove household and recreational activities from transmission sites . This has been demonstrated by numerous studies which observed significantly lower infection rates following the installation of water treatment plants or water recreation areas [24–27] . Water treatment infrastructure can also provide safe water for animals which act as reservoir hosts , particularly for S . japonicum , and which can play a significant part in transmission [28 , 29] . Water treatment infrastructure should not replace other interventions , such as preventive chemotherapy or snail control . On the contrary , it should be implemented alongside chemotherapy-based and behavioral change programs to reduce water contact and thereby minimize the likelihood of reinfection following drug administration . Occupational activities such as fishing and sand harvesting rely on natural water bodies , and exposure to contaminated water will not be affected by the introduction of a safe water supply . Alternative preventive measures may be needed to reduce such forms of exposure ( e . g . boots and gloves for fishermen ) . Even with the provision of safe water , the community may continue to access transmission sites for reasons including preference for untreated water , overcrowding , or lack of privacy at the safe water source , so health education and campaigns to raise awareness of the dangers of exposure to infested water should always accompany water infrastructure implementation . While eliminating contact with infested water bodies completely and forever may be unachievable , strategies could at least be targeted at reducing this contact during and after PC programs , to reduce reinfection rates and hopefully allow a move towards complete elimination of infection in the community . Water treatment infrastructure also needs careful planning to ensure the technology is sustainable and acceptable to the community . Firstly , however , research is needed to determine which water treatment processes are effective at removing or inactivating cercariae , which is the motivation for this systematic review .
The first and second authors independently classified the papers and extracted relevant information . A classification system that assigns a code to each paper was developed to reduce bias in the classification process ( S1 Table ) . First , duplicates were removed . Then , titles were classified and papers were excluded when titles indicated that they were not about water treatment or cercariae ( such as code 4 –this paper is about diagnostic tests ) . Abstracts of the remaining papers were read , and excluded abstracts were classified ( such as code 11 –this paper is primarily about shedding cercariae , not the survival of cercariae ) . Finally , the remaining papers were read in full and classified . Papers with code 13–19 were used for the systematic review as these papers discuss the effect of water treatment on schistosome cercariae . All codes and papers ( included and excluded ) are listed in S1 Table , and S2 Fig depicts the flow diagram of the selection process . Full texts were sought from the Imperial College London Library , The Wellcome Library , and the British Library . Papers in other languages were translated by Imperial College London students , and the results discussed with the first author . The two assessors met to compare and discuss the classifications , and any discrepancies were resolved . The assessors summarized the papers and extracted specific information regarding each treatment method . To ensure that the same data were extracted , a table with variables relating to each water treatment method was developed ( S1 Table ) . The assessors reviewed each other’s summaries to ensure the correct information was captured .
The studies in this review evaluated the effect of water treatment processes on cercariae , however the measure of effectiveness varied significantly between papers . There were four main aspects that authors assessed–the presence , motility , infectivity , and viability of cercariae . A range of methods exist to evaluate these aspects ( summarized in Table 2 ) with differing results . Presence of cercariae is used to assess the effectiveness of filtration . Microscopes are used to detect cercariae in water or on a recovery filter . Cercariae may be stained with a dye prior to being filtered to aid with the detection . Some studies fixed cercariae before filtration to facilitate the counting [30 , 31] . This may affect results as cercariae are highly mobile and may be able to move through pores that dead cercariae would not pass through [31] . In addition , fixatives have been found to make cercariae sticky , thereby increasing the effectiveness of the filter [31] . For water treatment methods besides physical filtration , the motility of cercariae was most commonly used to assess the effectiveness of water treatment . Cercarial movement is observed under the microscope , often with dyes to facilitate observations . Initially , cercariae in treated water slow their movement , sink , and turn opaque and fluffy . They may lose their tail before the complete cessation of movement [32] . Eventually they may disintegrate , which is a sign of mortality . Most authors used motility as a measure of death . However , it has since been found that immotile or even tailless cercariae may still be infective [33] , suggesting that the death of cercariae may have been determined prematurely with the motility measure . Conversely , it is also possible that cercariae that are motile are no longer infective . Cercariae of varying human schistosome species were found to have a functional longevity ( time that cercariae are alive but not infective ) of 50% of their survival time [34] . This would affect experiments that run longer than the functional longevity . Infectivity can be used as a measure for testing the cercarial ability to penetrate skin . It is crucial to note that infective cercariae may die before developing into schistosomula or schistosomes , and thereby not necessarily lead to schistosomiasis infection . The attachment of cercariae to skin is visually evaluated under the microscope , or cercariae are counted before and after exposure to the host , the difference assumed to have penetrated the skin . Animal testing on mice , rats and guinea pigs is common , but cultured cells and human skin donors have also been used . Cercariae are able to determine between different types of skin , and hence the results may differ depending on the skin used [35] . Bartlett et al . showed that cultured primary keratinocytes are an alternative to animal testing , as the attachment rate is similar to that of human skin [36] . A promising method for assessing cercarial infectivity is a biosensor , as has been developed by Webb et al . [37] . Schistosome cercariae secrete elastase when penetrating the host skin , and the biosensor communicates the detection of this elastase through a loss of color . Only viable cercariae can complete the lifecycle , and therefore transmission is cut if cercariae are rendered non-viable . Cercariae are considered viable if they can infect the human host and develop into schistosomes . Animal testing is often used to determine if cercariae developed into worms . The number of worms and severity of infection can only be determined by perfusing animals for worm counts . Another method for testing the viability of cercariae is the fluorescein diacetate ( FDA ) method . Viable cercariae are stained green as they take up and retain the chemical , whereas non-viable cercariae lose the green stain due to lacking esterase activity [33] . This method was initially developed to test schistosomula viability [38] , and has since been used to test cercarial viability [33] . Ultimately , cercarial infectivity and viability are of greatest interest , as this determines if cercariae can infect a human host with schistosomiasis . Immobile cercariae may still be infective if in direct contact with human skin [33] . Similarly , infective cercariae may not develop into schistosomula or schistosomes and hence are not viable [33] . However , if cercariae are not infective or not viable , schistosomiasis will not be transmitted . Leiper ( 1915 ) was the first to study the longevity of schistosome cercariae and found that they could survive up to 36 hours in water [39] . A more conservative estimate was used by the Great Army Medical Services in 1919 who suggested that troops fighting in tropical areas should store water for 48 hours to render it safe [18] . More recent studies found that cercariae live between 10 and 40 hours in natural conditions where water temperature varied between 20 and 30°C [15 , 40 , 41] , but at lower temperatures their lifespan has been shown to exceed 100 hours [32] . It is now commonly accepted that cercariae can survive up to two to three days in environmental conditions [42] . Cercariae are non-feeders and depend on internal , non-renewable glycogen levels . As a result , their glycogen content decreases exponentially post-emergence [15] . However , energy levels do not correlate directly with infectivity: Whitfield et al . demonstrated that cercariae remain highly infective for up to nine hours , despite glycogen levels decreasing continually [43] . Differences in cercarial survival may be due to water temperature and water quality . In 1966 , Frick and Hillyer tested the effect of various water media on cercariae , and found that cercariae were immobile in four hours in deionized and steam-distilled water [44] . Mecham et al . tested synthetic waters to see which could maintain cercarial life the longest [45] . Soft and distilled water had the fastest cercarial die-off , aquarium water followed , and synthetic hard water could support cercarial life for up to 50 hours ( pH , temperature and dissolved oxygen were controlled ) . Similarly , Asch showed that cercarial lifespan decreases with the addition of sodium chloride , and increases when adding glucose [40] . Overall , these studies demonstrate that the water matrix ( the components of the water ) can significantly affect results . In 1911 , the impact of temperature on schistosome transmission was first described . Studies found that the disease was acquired from thermal springs in Gafsa that reached temperatures of 45°C , but not from nearby hot springs that were at least 50°C [39 , 46 , 47] . Since this discovery , it has been shown that the lifespan of cercariae decreases with temperature , when considering the same storage times [15 , 48] . This can be due to thermal intolerance or energy depletion . At temperatures exceeding 45°C , extreme temperature causes direct cercarial mortality , known as the thermal death point . The thermal death points for different schistosome species are summarized in Table 3 . The point of death was visually examined through cessation of movement . At lower temperatures ( below 45°C ) , increased cercarial death is a consequence of energy depletion . Cercarial lifespan is dependent on glycogen reserves and the rate of utilization , which is affected by temperature [49] . As temperature rises , cercariae become more active , indicated by greater displacement [50] and hence deplete their glycogen levels faster , resulting in shorter lifespans [15] . Overall , it can be concluded that the time needed to reach zero cercarial survival decreases exponentially with temperature , as shown in Fig 1 . An exponential regression model was fit to the data with an R2 of 0 . 81 , indicating a moderately good fit . Cercarial survival is expected to decrease exponentially with temperature , as the depletion of glycogen is also exponential , and the rate of biological processes often increases exponentially with temperature [51] . However , only 14 data points were found and therefore more data are needed to improve the regression model . Cercariae died within minutes at 50°C , and within hours at temperatures below 44°C , though there may be some variation between species . In 1938 , Porter carried out the only detailed experiments on S . haematobium cercariae survival with temperature [52] . The limited results indicate that the species is more resistant at 45°C than S . mansoni , however more data are needed to verify these results [52] . External factors such as darkness , oxygen or pH ( pH 4 . 6–10 ) did not appear to affect cercarial longevity [32 , 48] . The effect of temperature on cercarial infectivity ( ability to penetrate host skin ) and viability ( ability to develop into schistosomes ) follows a similar trend . It has been shown that significantly more cercariae die during host penetration at extreme temperatures ( below 10°C and above 40°C ) than at 25–27°C for S . mansoni and S . haematobium [54] . Similarly , worm recovery increased with temperature to a maximum at 24–28°C for S . mansoni , followed by a steep decrease [55 , 56] . In 1964 , Foster was the first to study the relationship between temperature and infectivity , and found that in environmental conditions ( 18–30°C ) temperature did not affect skin penetration and maturation to schistosomula [57] . Similarly , Lee et al . found that attachment is more dependent on cercarial age than environmental temperature ( 20–30°C ) [14] . The effect of temperature on the functional longevity of cercariae remains to be studied . Filtration is an effective water treatment method that works by physically retaining particles while allowing water to pass . Filters can retain solid matter much smaller than the pores of the filter material due to mechanical and biological processes . Granular filters , which use sand or other granular material , are suitable for use in less developed areas as the filtering media is generally locally available and no external power input is required to pass the water through the filter . Initial studies on sand filters found that they were unable to remove cercariae from water , and troops were advised to use Doulton candles instead [18 , 39 , 58] . Nonetheless , Kawata showed that cercariae can be removed from water using sand with an effective size of 0 . 3 mm or less , and sand depth of 1 . 2 m [59] . Two studies confirmed that vertical sand filters are effective , however the discrepancy with earlier studies may be due to the heavy dependence on experimental conditions [60 , 61] . Primarily , the size of filters varied significantly . Kawata and Fadel tested filter depths up to 1 . 2 m and 1 . 4 m respectively with positive results [59 , 61] , whereas Witenberg and Yofe tested filters up to 0 . 75 m depth , which could explain the passing of cercariae [58] . Studies found that slow filtration ( 0 . 04–0 . 19 m/h ) retained more cercariae than more rapid filtration ( 0 . 27–0 . 4 m/h ) , as cercariae were not washed through the filter [59 , 61] . These studies also showed that conditioning of filters , which develops a biologically active Schmutzdecke ( biological layer formed on the surface of a sand filter ) , can improve the overall efficacy of sand filters . The filtration results are summarized in Table 4 . In vertical filters , water flows from top to bottom ( as opposed to horizontal sand filters in which water moves laterally downwards from one side to the other ) . Benarde and Johnson tested horizontal sand filters , which are often constructed in channels filled with sand [63] . Cercariae were unable to pass horizontal sand filters with 0 . 35 mm effective grain size and 0 . 9 m bed depth . These filters seem to achieve better results than vertical sand filters , the hypothesis being that the natural movement of cercariae is vertically up and down , so using a filter where water flows horizontally helps filter out cercariae [63] . Further research is needed to confirm this hypothesis . Generally , filters used one grain size , but a filter with layers of different sized sand may prove to be beneficial and reduce the overall depth of the filter . Pre-filters would also reduce clogging and allow the filters to be used in the field with turbid water . Further research is needed to test the effect of sand size , flow rate , filter depth , filtration volume and water quality on sand filtration as a method to completely remove cercariae from water . Other filter materials that have been tested include diatomaceous silica , a naturally occurring rock that is often used in filtration as a powder . Jones and Brady tested three types of diatomaceous silica filters and recovered no cercariae from the effluent , although no filter material sizes were specified [64 , 65] . Diatomaceous silica has small , irregular pores which helps prevent cercariae from passing through the filter . Membrane filters have been used for cercariometry ( to determine the density of cercariae in water . Micro-fabrics with pore sizes of 23–30 μm have successfully filtered out cercariae [30 , 31 , 66–71] . When filtering natural water , several pre-filters were used to prevent the clogging of the filter . While a slow filtration rate was shown to be beneficial in sand filtration , others found that more rapid flow rates immobilized cercariae and prevented them from moving themselves through the membrane [31] . Although experiments varied in cercaria numbers , water volume , water quality , filtration rate and time , the results are consistent . Water chlorination is an effective water treatment method that kills most bacteria and viruses . Chlorine is an oxidizing agent and kills pathogens through oxidation . Chlorine disintegrates the cell wall of pathogens , rendering them unviable . In chlorination , a ‘CT’ value indicates the residual chlorine dose ( C ) and contact time ( T ) required to inactivate a pathogen , and is calculated as the product of the two variables . CT values have been determined for many waterborne pathogens ( see Table 5 ) , and can be used as a process monitoring parameter , i . e . simply ensuring that the required CT is being achieved means that constant pathogen monitoring is not required . Higher CT values indicate a higher chlorine tolerance . The effect of chlorine on schistosome cercariae has been researched throughout the past century , with varying conclusions . In 1920 , Mahnson-Bahr and Fairley found that some cercariae ( unknown percentage ) were still alive after 2 . 5 hours in water containing 4 mg/l of chlorine [10] , whereas Magath found that 0 . 2 mg/l after 30 minutes was sufficient to kill all cercariae [73] . More recently , the WHO stated that a chlorine residual of 1 mg/l after 30 minutes is sufficient to render cercaria-free water [74] . Table 6 summarizes some of the chlorination results found in this review . Note that the contact time is the time until inactivation of all cercariae ( 100% kill ) , which in all studies was measured by complete cessation of movement of all cercariae in the sample . The data sets of all studies have been plotted in Fig 2 to show the effect of varying chlorine doses on cercariae ( a more detailed graph is shown in S3 Fig ) . Note that these are doses of chlorine ( the chlorine added to the water ) , not residual chlorine concentrations in the water ( accounting for the reaction of chlorine with other water constituents ) ; the lack of reporting of the chlorine residual concentration is a major weakness of many prior studies . It is evident that in all studies the effect of chlorine increased exponentially with chlorine dose . High doses , generally above 1 . 5 mg/l , killed all cercariae within minutes . The data follows a clear trend and can be fit to an exponential regression model with R2 of 0 . 68 ( Fig 2 ) . Data is consistent at low lethal chlorine doses , but there are discrepancies at short inactivation times . This is particularly the case for inactivation times less than 30 minutes ( see S3 Fig ) where the lethal chlorine dose at 3 minutes , for example , varies between 1 . 1 mg/l and 5 mg/l . This is enforced by the variation of calculated CT values in Table 6 , which vary from 3 to 30 mg-min/l . The variation may be due to differing experimental conditions , such as pH , temperature , water quality ( e . g . organic matter which will react with chlorine ) , the form of chlorine used , chlorine measurement techniques , and schistosome species , or because of incomplete mixing of the chlorine into the water at very short contact times . The type of water and water quality used in the experiments varied greatly . This is noteworthy , as water temperature and pH significantly affect the CT value . When sodium hypochlorite is added to water , it reacts to form hypochlorous acid . It may react further , especially under alkaline conditions , to form hypochlorite ions . The latter is a much weaker disinfectant and as a result , more chlorine is needed at higher pH to achieve the desired level of disinfection . Generally , a higher CT value is needed at lower temperature and higher pH . Lower temperature results in lower rates of chemical disinfection . In 1942 , Braune tested ways to increase the efficiency of chlorine against cercariae [77] . He found that the addition of citric acid heavily accelerated the chlorine disinfection process , and suggested adding 100 mg/l of citric acid to the water . Frick and Hillyer followed in 1966 with a detailed study on the effect of chlorination on cercariae with respect to water temperature and pH [44] . They found that pH had a much greater effect on chlorine efficiency than temperature; in tap water at 20°C , S . mansoni cercariae were inactivated in 30 minutes with chlorine concentrations of 0 . 3 mg/l at pH 5 , 0 . 6 mg/l at pH 7 . 5 , and 5 mg/l at pH 10 . The chlorine demand is the difference between the amount of chlorine added to water and the residual chlorine after a given contact time . Generally , turbid water has a higher chlorine demand and consequently the CT value may be expected to increase with turbidity . Pathogens may attach to particulate matter in the water , which may encapsulate them . For this reason , CT values are usually calculated for low turbidity water , and values are then doubled in the field to account for the variability in water quality [72] . Experimental studies should account for the chlorine demand of the waters that they are testing and not simply report the chlorine dose added to the water . It is critical to measure the residual chlorine levels as chlorine decomposes rapidly , and when added to water reacts with chemicals , organic matter , and pathogens . These reactions remove some of the chlorine , hence the residual chlorine is lower than the chlorine dose added . Water that is highly contaminated therefore requires more chlorine to inactivate pathogens , to overcome the chlorine demand of the water matrix . Some authors did reference the available chlorine residual , but the time of measurement did not always coincide with the inactivation time [76 , 77] . The form of chlorine used in experiments included chloramine , chlorine gas , bleaching powder , and calcium hypochlorite . These may have differing cercaricidal effects , even when used in equal available chlorine doses . Authors have reported that chloramine was less effective than chlorine , but more powerful than gaseous chlorine , and that the method of preparing chlorine also affected the results [58 , 76] . Similarly , the stability of chloramine varied depending on how it was prepared . The Great Britain Army Medical Services advised against using water chlorination as a means to disinfect cercaria-infected water because chlorinated lime degraded too quickly , and the presence of sufficient free chlorine could not be ensured [18] . In addition , the method for measuring residual chlorine may be a source of error , as the DPD titration , orthotolidine method , and colorimetric test kits used in the studies have varying accuracies [79] . The species appear to impact results . S . haematobium has been found to be more sensitive to chloramine , and S . mansoni more sensitive to gaseous chlorine [58 , 76] . Out of the 12 chlorine studies identified in this systematic review , only three experimented with S . haematobium or S . japonicum cercariae , stressing the need to research these Schistosoma species . In addition , it was only these three studies that simultaneously tested two species [58 , 65 , 76] . They found that S . haematobium and S . mansoni had differing sensitivities to chlorine , which highlights the need to conduct further research on all species under a variety of water matrix conditions . The age of the cercariae may also play an important role in chlorination , but no research has evaluated this . Generally , studies used freshly hatched ( <1 hour ) cercariae . Sproule stated that fresh cercariae were more resistant to chlorine than cercariae older than 1 hour [80] . This could be due to cercariae hatching within mucus from the snail [43] , which may act as a protective layer . All authors used motility as a measure of death , and the points on Fig 2 show the time until the last cercaria stopped moving . As previously mentioned , cessation of movement does not necessarily indicate death , so the values in Fig 2 may be an underestimation of lethal chlorine doses . In water disinfection studies , log-inactivation is commonly used to indicate what percentage of pathogens has been inactivated . As the lethal chlorine doses were recorded in previous studies when all cercariae were dead , the experiments reported a 100% kill . However , due to the small cercaria samples used ( i . e . 100 or fewer cercariae ) , the highest log-kill that could be claimed was a 2-log kill ( 99% kill ) . Samples exceeding 1000 cercariae would need to be tested to claim a 3-log inactivation ( 99 . 9% kill ) , which may be challenging if evaluating the motility . Also , survival curves for cercariae follow a reverse-sigmoid form [15] , which suggests that if a small sample is used , the results may be less reliable . This could explain the differences between Witenberg and Yofe , and Frick and Hillyer , as the former tested up to 200 cercariae [44] , whereas the latter used as little 20 cercariae in experiments [58] . UV disinfection uses radiation in the UV wavelength range to inactivate microorganisms . This range is often referred to as germicidal , and spans wavelengths of 200–300 nm . UV is absorbed by the DNA and RNA of microorganisms , damaging cell membranes and thereby inhibiting reproduction . The sensitivity of microorganisms to UV is determined by the UV fluence ( mJ/cm2 ) –the product of the fluence rate and exposure time [81] . In this review , two types of UV studies were found: immunization studies and water treatment studies . Immunization studies research the possibility of using UV-attenuated cercariae to develop a vaccine against schistosomiasis . The process is as follows: cercariae are irradiated with UV to a level that damages them but that does not prevent host penetration . Irradiated cercariae infect the host by penetrating the skin , but due to the UV damage are unable to develop into adult worms . Nonetheless , antibody production is triggered , and as a result , the host builds up immunity . Lighter , succeeding infections is proof that partial resistance is induced [82 , 83] . The other type of UV studies found in this review was water treatment studies . These aim at using UV to disinfect cercaria-infested water and kill or inactivate cercariae . Both types of studies therefore use UV to damage cercariae , their protocol is however very different . Immunization studies irradiate cercariae so that their movement is not impaired , and they are still able to penetrate the host , whereas water treatment studies generally use higher UV fluences to instantly kill cercariae . The effect of UV on cercariae and worm burden is summarized in Table 7 . Cercariae were irradiated with fluences between 3 mJ/cm2 and 1890 mJ/cm2 , all at 254nm , but the direct effect ( motility , infectivity ) on cercariae was often not evaluated as the focus was on worm burden ( viability ) . Krakower was the first to research the effect of UV radiation on schistosome cercariae and found that cercariae appeared physically injured after 30 minutes , and motionless after 60 minutes exposure to sunlight when kept in a shallow container [32] . UV fluences as low as 3 mJ/cm2 have been shown to damage cercariae , slow their movement and reduce their survival [84 , 85] . Higher fluences were tested in a series of experiments aimed to identify the UV fluence that triggers the highest level of host immunity [84] . The authors tested fluences up to 60 mJ/cm2 and found that 54 mJ/cm2 instantly reduced cercarial motility and damaged their physiology by inhibiting glycoprotein synthesis . A further increase to 60 mJ/cm2 made cercariae immotile and not infective . It was determined that the appropriate UV fluence for immunization studies was 18–24 mJ/cm2 , as it does not inhibit cercarial penetration but results in significantly lower worm burden , as confirmed by numerous other studies [86–88] . In addition , the migration potential of adult worms is reduced , hampering their ability to migrate from skin to organs [84 , 89] . Overall , cercarial movement , infectivity , and viability decreased with UV fluence . The study by Ghandour and Webbe clearly demonstrates the importance of evaluating the different aspects of cercariae ( motility , infectivity , viability ) [82] . Irradiation for less than 0 . 3 minutes showed no effect on cercarial movement . However , the impact was visible when cercariae were exposed to skin; increased mortality rates during skin penetration resulted in reduced cercarial infectivity . Furthermore , the authors studied cercarial viability by examining schistosomula migration and worm burden . Most schistosomula died within days and did not develop into schistosomes , and were thus not viable [82 , 95] . Similarly , Ariyo and Oyerinde showed that the damaging effect of radiation may not be immediately apparent at low UV fluences [85] . Within the first hour post-radiation , there was no effect on cercarial movement , and the attachment rate ranged between 93% and 100% , similar to the control . Nonetheless , the effect of UV became apparent in the stages following host penetration . Significantly fewer cercariae developed from schistosomula into schistosomes , and those that did had reduced fecundity , indicated by the decrease in egg load and egg viability . The increased cercarial death associated with UV irradiation may be partially due to the increased water temperature . At the time of the studies , UV lamps emitted significant amounts of heat , easily able to increase the temperature of the water within the running time of the experiment . The risk of temperature affecting results is especially high for studies that used small water samples and long irradiation times . This could explain the positive results by Dean et al . who recovered no worms at fluence 19 . 8 mJ/cm2 and 3 minutes [87] , whereas other studies using higher UV fluences but shorter exposure times simply achieved a worm reduction [28 , 88] . The heat production and changes in temperature of the water were not recorded in the studies . It appears that S . mansoni and S . japonicum are equally sensitive to UV , which may be explained by similar distribution of cercariae in water columns; their cercariae accumulate at the water surface , whereas S . haematobium cercariae prefer to remain beneath the surface [49] . This characteristic could lead to S . haematobium cercariae being less damaged by UV radiation , as accumulating at the bottom of the water column provides a protective water layer . Only one study tested S . haematobium , and three studies simultaneously tested two species , making it difficult to compare the results of all species . UV radiation has been shown to inhibit the migration within the host , hampering the worms’ ability to migrate from skin to organs [82 , 86] . This may be due to the physical damage of worms that develop from irradiated cercariae . One study examined the physical effect of UV on S . mansoni worms , and found that worms developed from cercariae that were irradiated for 1 minute with a Mineralight lamp ( unknown fluence ) had lesions , torn tubercles , and lost their spikes . This resulted in sexual anomalies and sterility [96] . The experiments generally irradiated cercariae in a shallow container with tap or distilled water . Therefore , the UV fluence measured at the surface of the water was assumed to be the UV fluence reaching the cercaria surface . This , however , is not the case since suspended particles and water itself absorb UV light , reducing the depth of UV penetration . The turbidity and depth of a water sample increase the UV fluence needed to kill cercariae , and this needs to be considered in future research . The studies demonstrate that UV radiation damages cercariae , however the lack of measurements and controlled variables makes it difficult to draw concrete conclusions . Numerous studies did not measure the UV fluence or irradiation time , as seen in Table 7 . Generally , the fluence was measured at the surface of the water with a UV meter . However , occasionally the distance at which the fluence was measured did not correspond to the surface of the water , and was hence not applicable [90] . The time of taking UV measurements may also affect the accuracy of UV fluences . Some authors took UV readings instantly after turning on the lamp , whereas others allowed the bulb to warm up for 15–30 minutes before taking measurements . Cercariae were irradiated in a variety of containers which included quartz cuvettes , glass slides and well plates , which may also affect the level of UV radiation . The density of cercariae varied greatly between experiments , and may have affected results as cercariae may shield each other in high densities . It is crucial to carry out experiments under well-defined and controlled UV fluences to truly determine the UV-sensitivity of schistosome cercariae . The results of this systematic review indicate that there are many differences in assessing the effects of water treatment processes on schistosome cercariae ( i . e . motility , infectivity , viability ) , making it difficult to directly compare the results of studies that have used different criteria of death . It is necessary to determine at what level of motility cercariae stop being infective and viable . Methods like the biosensor and FDA described in this review may prove to be useful , as they assess the viability of cercarial populations in a water sample , not of each individual cercaria . All treatment methods should be evaluated with the same criterion that ensures cercariae are not infective or viable . The level of variability in results differs between treatment processes . Previous studies on the effect of water storage and temperature are relatively in agreement , and indicate that water quality and pH do not greatly affect these two treatment methods . Cercariae are killed when storing water for more than one to three days , depending on temperature . The results of other water treatment processes , filtration , chlorination , and UV , are much more variable because the processes are heavily affected by experimental conditions , and require further research to obtain reproducible information . It is evident that the water matrix used in experiments hugely impacts cercaria survival . Most studies conducted experiments using lab-hatched cercariae and tap or filtered water . This builds a foundation for understanding the effect of water treatment processes on cercariae . However , it is essential to also conduct experiments under real water matrix conditions . In addition , water treatment processes should be tested individually as well as in succession of each other , such as filtration followed by chlorination , as this may achieve higher levels of cercaria removal than the treatments applied in isolation . The variability in findings is likely also due to differences in experimental protocols and discrepancies in measuring key parameters . Chlorination studies often measured different chlorine values , such as chlorine dose , free chlorine at one-minute , or free chlorine at point of cercarial death . As there is no conversion between these measurements , the studies are incomparable . UV fluence measurements often did not correspond to the fluence reaching cercariae in the water . Filtration studies varied immensely in terms of scale and running time , making it difficult to compare the results . Experiments for all treatment processes should measure the variables needed to reliably design water treatment infrastructure . The review has highlighted the lack of water treatment studies testing S . haematobium and S . japonicum . The majority of studies ( 57% ) tested S . mansoni , 15% S . japonicum , and 3% tested S . haematobium . Only 12% of the studies tested more than one species ( either S . mansoni and S . haematobium , or S . mansoni and S . japonicum ) , making it difficult to draw conclusions regarding their relative resistance to water treatment processes . The remaining 13% did not indicate which species were being tested , primarily older studies . This review focused on the three main human Schistosoma species and no studies about water treatment processes and animal reservoir hosts were found in this review . The only studies involving animals were immunization trials using UV-attenuated cercariae . Animal reservoirs , including domestic livestock , can play a critical role in schistosomiasis transmission , especially of S . japonicum [28 , 29] . Using treated water for livestock water use may contribute to lower infection rates in both humans and animals , and is therefore of medical and veterinary importance . Implementing water treatment infrastructure could remove household water activities such as laundry , washing and recreational swimming from transmission sites . Depending on the level of treatment required to remove cercariae , other water-borne pathogens may also be inactivated , rendering water safe for drinking ( e . g . if chlorination at a sufficient contact time is used ) . Successfully implementing and sustaining the infrastructure will require careful consideration of issues of equity of access and affordability , and it is crucial that educational programs run alongside the implementation to ensure understanding of the operation and maintenance needs . All five reviewed water treatment processes can be implemented as low-cost household or community-scale water treatment systems , yet the cost is dependent on local conditions such as availability of materials or supply chains . It is important to note that the provision of water treatment infrastructure will not remove occupational water exposure , such as fishing , from transmission sites . Nonetheless , an overall reduction in the population’s exposure to contaminated water may result in lower odds of infection and reinfection following chemotherapy , a relationship that needs to be more carefully studied and quantified . Safe water infrastructure may accelerate progress towards control and elimination of schistosomiasis by being incorporated into PC-based strategies , where it would serve to reduce the need for contact with infested water . We propose the following research priorities: The literature demonstrates that all the reviewed water treatment methods have the potential to effectively remove or kill cercariae and thereby produce safe water . However , the database of required water treatment is insufficient to be able to devise guidelines for the design of water infrastructure for schistosomiasis-endemic regions . As countries target the control and elimination of schistosomiasis , it will be crucial to develop such water treatment guidelines , and to better link activities of the WASH sector to PC-based control programs . | Schistosomiasis control currently focuses on preventive chemotherapy ( PC ) with praziquantel , which is effective , safe , and inexpensive . However , this treatment does not prevent subsequent reinfection . As schistosomiasis control targets become more ambitious and move towards elimination , interest is increasing in the potentially complementary roles of water , sanitation , and hygiene ( WASH ) interventions which may disrupt transmission of the parasite , thereby slowing reinfection following treatment . Water treatment for schistosomiasis control seeks to eliminate viable schistosome cercariae from water . We carried out a systematic review to summarize the existing knowledge on the effectiveness of water treatment for the removal or inactivation of cercariae , by processes including chlorination , filtration , UV disinfection , heating , and water storage . This is the first review of its kind and provides a concise summary of what is known to-date regarding water treatment against cercariae of different Schistosoma species . The review also identifies gaps in knowledge and provides crucial and timely guidance for the control and elimination of schistosomiasis , by highlighting the requirements for designing effective and sustainable water infrastructure for schistosomiasis-endemic regions . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"and",
"discussion"
] | [
"schistosoma",
"invertebrates",
"schistosoma",
"mansoni",
"medicine",
"and",
"health",
"sciences",
"ultraviolet",
"radiation",
"helminths",
"tropical",
"diseases",
"database",
"searching",
"light",
"parasitic",
"diseases",
"animals",
"electromagnetic",
"radiation",
"neglected",
"tropical",
"diseases",
"research",
"and",
"analysis",
"methods",
"schistosoma",
"japonicum",
"schistosoma",
"haematobium",
"chemistry",
"chlorine",
"physics",
"helminth",
"infections",
"schistosomiasis",
"chemical",
"elements",
"eukaryota",
"database",
"and",
"informatics",
"methods",
"biology",
"and",
"life",
"sciences",
"physical",
"sciences",
"organisms"
] | 2018 | The effectiveness of water treatment processes against schistosome cercariae: A systematic review |
Thrombocytopenia is a hallmark of dengue infection , and bleeding is a dreaded complication of dengue fever . Prophylactic platelet transfusion has been used to prevent bleeding in the management of dengue fever , although the evidence for its benefit is lacking . In adult dengue patients with platelet count <20 , 000/mm3 without bleeding , we aimed to assess if prophylactic platelet transfusion was effective in reducing clinical bleeding and other outcomes . We conducted a retrospective non-randomised observational study of dengue patients with platelet count < 20 , 000/mm3 without bleeding ( except petechiae ) admitted to Tan Tock Seng Hospital from January 2005 to December 2008 . Baseline characteristics and clinical outcomes were compared between the non-transfused vs . transfused groups . Outcomes studied were clinical bleeding , platelet increment , hospital length of stay , intensive care unit admission and death . Of the 788 patients included , 486 received prophylactic platelet transfusion . There was no significant difference in the presence of clinical bleeding in the two groups ( 18 . 2% in non-transfused group vs . 23 . 5% in transfused group; P = 0 . 08 ) . Patients in the transfused group took a median of 1 day longer than the non-transfused group to increase their platelet count to 50 , 000/mm3 or more ( 3 days vs . 2 days , P <0 . 0001 ) . The median duration of hospital stay in the non-transfused group was 5 days vs . 6 days in the transfused group ( P< 0 . 0001 ) . There was no significant difference in the proportion requiring ICU admission ( non-transfused 0 . 66% vs . transfused 1 . 23% , P = 0 . 44 ) and death ( non-transfused 0% vs . transfused 0 . 2% , P = 0 . 43 ) . Platelet transfusion in absence of bleeding in adult dengue with platelet count <20 , 000/mm3 did not reduce bleeding or expedite platelet recovery . There was potential harm by slowing recovery of platelet count to >50 , 000/mm3 and increasing length of hospitalization .
Dengue is estimated to cause 390 million infections annually of which a quarter are symptomatic [1] . In the last 5 decades , the incidence of dengue cases has increased 30 -fold with expanding geographical distribution , leading to major international public health concern[2] . Dengue infection results in a spectrum of clinical syndromes ranging from a mild flu-like illness to life-threatening dengue shock with bleeding and multi-organ failure[3] . The most striking laboratory finding in dengue is thrombocytopenia . Thrombocytopenia was seen in 99% of dengue patients in a study from Trinidad and Tobago [4] and in a study from Taiwan , up to 85% of patients with dengue fever had platelet count of <100 , 000/mm3[5] . Platelet count < 20 , 000/mm3 was found in up to 45% of subjects in the Trinidad and Tobago study [4] . Consequently , platelet transfusion may be a plausible means of preventing hemorrhagic manifestations in dengue fever . Platelet transfusion for thrombocytopenia in dengue fever is a common practice . The proportion of dengue patients receiving platelet transfusion ranged from 7% to 50 . 3% in studies from Trinidad and Tobago , India , Taiwan and Singapore[4 , 6–9] . This wide range reflects varying local practices and general lack of consensus with regards to the management of thrombocytopenia in dengue . In a global survey , 190 out of 306 ( 62 . 1% ) respondents did not advocate prophylactic platelet transfusion in the absence of bleeding [10] . However , there was also a wide geographic variation in their responses . A few studies investigated the role of prophylactic platelet transfusion in dengue fever . In a study of pediatric patients with dengue shock syndrome and platelet count of <30 , 000/mm3 , preventive transfusion did not reduce bleeding . There was significantly higher fluid balance , incidence of pulmonary edema and increased length of hospital stay associated with preventive transfusion [11] . In an observational cohort study of adult patients with platelet count of <20 , 000/mm3 , transfusion of platelets in 188 out of 256 subjects had no effect on incidence of bleeding , rate of platelet increment and length of hospital stay [9] . A restrictive policy adopted in a study from Martinique resulted in platelet transfusion in only 9 out of 350 patients ( 2 . 6% ) . 3 deaths ( 2 were given platelet transfusion ) were reported but they were not related to hemorrhage [12] . We conducted a retrospective observational study to investigate the effect of prophylactic platelet transfusion in the management of thrombocytopenia <20 , 000/mm3 in adult dengue fever and evaluate its usefulness in the prevention of hemorrhagic manifestations .
We conducted a retrospective analysis of all adult patients who were admitted to Communicable Disease Centre , Tan Tock Seng Hospital , Singapore between January 2005 and December 2008 . All patients were tested positive for dengue polymerase chain reaction ( PCR ) [13] or dengue IgM/IgG ( Dengue Duo IgM & IgG Rapid Strip , Panbio Diagnostic , Queensland , Australia ) [14] together with probable dengue criteria based on World Health Organization ( WHO ) 1997 or 2009 dengue guidelines [15 , 2] . Patients who developed platelet count <20 , 000/mm3 , without any bleeding manifestations ( with the exception of petechiae ) were assessed . Decision to provide or defer prophylactic platelet transfusion was based on clinical judgment of the physicians in-charge . Patients admitted for dengue were managed according to a clinical care pathway developed in Tan Tock Seng Hospital . Clinical data collected were recorded in the pathway . Data collected from all patients included age , sex , fever duration , medical co-morbidities , signs and symptoms , diagnoses of dengue hemorrhagic fever ( DHF ) based on WHO 1997 guidelines [15] , dengue with warning signs and severe dengue ( SD ) based on WHO 2009 guidelines [2] , hematological and biochemical results , treatment including intravenous fluid , platelet and blood transfusion , progress in hospital including admission to intensive care , length of hospitalization and death . Patients in the prophylactic platelet transfusion and non-transfusion groups were analyzed for clinical outcomes . Outcomes studied were: i ) clinical bleeding which was defined as any bleeding excluding the presence of petechiae , ii ) mucosal bleeding from gums , nose or vagina , iii ) internal bleeding defined as intracranial , retroperitoneal , gastrointestinal tract bleeding , hemoptysis or hematuria , iv ) platelet increment the day after transfusion , v ) time for platelet count to exceed 50 , 000/mm3 , vi ) length of stay ( LOS ) , vii ) intensive care unit ( ICU ) admission , and viii ) death . The chi-squared was used to test for univariate associations between categorical variables . Fisher’s exact test was used if the expected cell frequencies fell below 5 . Mann-Whitney U test was used to test for differences in continuous variables . Variables were imputed with their group median if less than 10% of data were missing . Variables with 10% missing data were excluded from the analysis . Propensity score matching ( PSM ) was used to analyze the effects of platelet transfusion on bleeding in view of different baseline characteristics in patients who had undergone the treatment regimen . Variables that we believed may influence a clinician's decision to transfuse platelets were included as well . Variables used in the PSM included age , year of infection , DHF , SD , diabetes mellitus , cardiac disease , fever day at presentation , presence of any warning signs , temperature , leucocyte , count neutrophil percentage , platelet count , gender , systolic blood pressure <90mmHg and whether aspartate aminotransferase ( AST ) , prothrombin time ( PT ) and partial thromboplastin time ( PTT ) were taken . Logistic regression was used to identify the risk factors of clinical bleeding . Twenty-six predictors with p<0 . 2 in the univariate analysis and/or were clinically relevant were entered into the logistic model . Age , gender , Charlson's co-morbidity score [16] , hematocrit and whether AST was taken were identified as potential confounders and adjusted for in the model . Patients who did not bleed were randomly split into 4 approximately equal subsets and merged with the cases . Logistic regression was performed on the 4 subsets , and each model was validated against the other 3 subsets . Manual backward elimination was performed to get the most parsimonious model . The results were further validated with stepwise regression using Akaike information criterion . The model with the best average sensitivity and specificity was deemed the most appropriate and further validated with the whole dataset . All statistical analyses were performed using R version 3 . 0 . 2 [17] and Stata 13 . 0 ( Stata Corporation , Texas , U . S . A . ) . All tests were carried out at a 5% significance level . The study was approved by the National Healthcare Group Domain Specific Review Board ( DSRB/E/2008/00567 ) with a waiver of informed consent for the collection of anonymized data .
Of 7500 patients with dengue fever studied , 788 ( 10 . 5% ) developed platelet count < 20 , 000/mm3 with no clinical bleeding . Of these , 486 ( 61 . 7% ) were given prophylactic platelet transfusion . The median volume transfused was 240 ml ( range 100-618ml ) . The demographic data , dengue severity , symptoms and signs , and laboratory data when patients developed platelet count <20 , 000/mm3 were presented in Table 1 . Patients with and without prophylactic platelet transfusion did not differ significantly in terms of gender , median age , co-morbidities and DHF . However , patients who received prophylactic platelet transfusion were more likely to have severe dengue . There were many significant differences in terms of clinical features and laboratory data between the two groups , although absolute differences were minor and unlikely to be of clinical significance . Compared with patients who did not receive prophylactic platelet transfusion , transfused patients had statistically significant shorter duration of fever , higher median temperature and pulse rate; reported more anorexia , nausea and vomiting; had higher median serum hematocrit , neutrophil proportion , serum albumin , AST and ALT , and lower median leukocyte and platelet counts . With regards to the clinical outcome ( Table 2 ) , there was no significant difference in the presence of clinical bleeding in the two groups ( 18 . 2% in non-transfused group vs . 23 . 5% in transfused group; P = 0 . 08 ) . Surprisingly , the occurrence of internal bleeding after platelet transfusion was slightly more common in transfused patients albeit statistically not significant ( 1 . 3% in non-transfused vs . 3 . 4% in transfused; P = 0 . 07 ) . Likewise , surprisingly there was more mucosal bleeding after platelet transfusion in transfused ( 18 . 5% ) versus non-transfused group ( 9 . 3% , P<0 . 001 ) . The median time to any bleeding was 1 day after patients developed platelet count less than 20 , 000/ mm3 ( range 1–3 days ) with no significant difference between the two groups ( P = 0 . 77 ) . Of those who had clinical bleeding , 91% of patients did so within 48 hours of reaching nadir platelet count regardless of platelet transfusion . 2 patients developed liver failure ( platelet transfused ) and 1 patient developed renal failure ( platelet transfused ) . There was no patient with pulmonary edema . While platelet count increased significantly more the next day in the transfused group ( by 8 , 000/ mm3 ) than in the non-transfused group ( by 5 , 000/mm3 , P<0 . 0001 ) , patients in the transfused group took a median of 1 day longer than the non-transfused group to increase their platelet count to 50 , 000/mm3 or more ( 3 days vs . 2 days , P <0 . 0001 ) . This finding was illustrated in Fig 1 which showed that platelet recovery was slower in the transfusion group . Consequently , the median hospital stay in the non-transfused group was 5 days vs . 6 days in the transfused group ( P< 0 . 0001 ) . There was no significant difference in the proportion requiring ICU admission ( non-transfused 0 . 66% vs . transfused 1 . 23% , P = 0 . 44 ) and death ( non-transfused 0% vs . transfused 0 . 2% , P = 0 . 43 ) . Given the baseline differences , and the excess mucosal bleeding in the transfused group , we performed propensity score matching analysis to examine the effect of prophylactic platelet transfusion and potential biases in treating physicians’ decision to provide prophylactic platelet transfusion . After adjustment for potential confounders of age , year of infection , DHF , SD , diabetes mellitus , cardiac disease , fever day at presentation , presence of any warning signs , temperature , leukocyte count , neutrophil percentage , platelet count , gender , systolic blood pressure < 90mmHg and whether AST , PT and PTT were taken , prophylactic platelet transfusion no longer had any effect in clinical , mucosal and internal bleeding ( Fig 2 ) . We compared the two groups and adjusted for age , gender , dengue severity , Charlson’s co-morbidity score , serum hematocrit and AST . Independent risk factors for any clinical bleeding were presence of fever , lower leukocyte count and higher neutrophil proportion ( Table 3 ) .
Thrombocytopenia is common in dengue infection [18] and may be the result of influence from cytokines [19] , decreased megakaryopoiesis [20] , an increase in the number of dysfunctional megakaryocytes [21] , increased destruction of platelets in the liver and spleen [21] , destruction of platelets following the binding of dengue-specific antibodies to virus-infected platelets [22 , 23] and sequestration of platelets by dengue-infected endothelial cells [24] . Bleeding is a dreaded clinical manifestation of severe dengue . This has long been attributed to thrombocytopenia associated with dengue fever . Prophylactic platelet transfusion has been performed in many instances in an attempt to mitigate this risk . A 1992 study by American Association of Blood Banks' Transfusion Practice Committee reported that over 70% of hospitals transfused platelets primarily for prophylaxis with an arbitrary threshold of 20 , 000/mm3 or higher in 80% of these hospitals [25] . This threshold was widely adopted for many years after published clinical data in 1962 despite lack of clinical evidence that 20 , 000/mm3 was the appropriate transfusion threshold [26] . In a study conducted by Lye et al , no significant relationship was demonstrated between clinical bleeding and platelet count in adult dengue [9] . In our cohort of patients with platelet count <20 , 000/mm3 without any bleeding , prophylactic transfusion was administered in 486 of 788 ( 61 . 6% ) cases . Our analysis showed that this practice did not reduce bleeding risk in this group of patients who were thought to be at high risk of bleeding . Platelet transfusion was shown to be associated with slower platelet count recovery with the group receiving transfusion taking one day longer to achieve a platelet count of >50 , 000/mm3 . This consequently resulted in an increase in length of stay by one day as platelet recovery to >50 , 000/mm3 was part of the discharge criteria during the study period . Thrombocytopenia in dengue correlated with high thrombopoietin level which stimulated platelet recovery [27] . The transient increase in platelet count from the transfusion could have caused a reduction in serum thrombopoietin level , thereby slowing the endogenous production of platelets from megakaryocytes [28] . Studies suggested that risk factors for bleeding in dengue included degree of thrombocytopenia [29] , older age [30] , female gender [31] , high hematocrit and elevated APTT [32] , and high absolute lymphocyte count [31] . While low platelet count was not associated with bleeding in dengue in our study , several correlations for bleeding were identified in our cohort . They were the presence of fever on the day of platelet count <20 , 000/mm3 , low white cell count and higher neutrophil proportion . Identification and analysis of other risk factors may contribute to the development of a bleeding risk calculator for the management of dengue patients . There were limitations to our retrospective study . Firstly , the lack of randomization may have resulted in treatment bias since the decision to transfuse platelets prophylactically was solely based on treating physician's decision . We attempted to correct possible confounders using propensity score matching which confirmed the lack of benefit in platelet transfusion . There were low numbers of severe clinical outcomes such as ICU admission and death in our cohort . Hence the effect of platelet transfusion on these outcomes could not be determined reliably . Secondly , although full blood count , renal and liver panel , PT/APTT were done on admission and FBC was repeated at least daily , further investigations were performed only when clinically indicated . To our knowledge , our cohort is the largest to date showing a lack of efficacy of prophylactic platelet transfusion in the prevention of bleeding in adult dengue with platelet count <20 , 000/mm3 . It resulted in slower platelet recovery and longer hospital stay . We currently reserve platelet transfusion in dengue fever to those with clinically significant bleeding manifestations . This will reduce the use of precious blood products and associated risks of transfusion . Additional research from a randomized trial is needed to address the role of prophylactic platelet transfusion in dengue . | Thrombocytopenia is one of the most prominent clinical features in dengue infection . This may manifest clinically as bleeding , which carries significant morbidity and mortality . For many years , clinicians have given dengue patients platelet transfusion in a bid to increase their platelet counts to prevent hemorrhagic manifestations . However , this practice has not been proven to reduce the risk of bleeding in dengue . Conversely , transfusion of blood products may be detrimental to patients as it can carry risks of fluid overload , transmission of infectious diseases and transfusion reactions . In this study , the authors found that platelet transfusion in the absence of bleeding in adult dengue patients did not prevent bleeding . Instead , this practice is associated with a slower platelet recovery and increased length of hospitalisation . These findings may contribute to better clinical management of thrombocytopenia in dengue patients and limit the use of these precious blood products . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [
"blood",
"cells",
"medicine",
"and",
"health",
"sciences",
"clinical",
"laboratory",
"sciences",
"body",
"fluids",
"pathology",
"and",
"laboratory",
"medicine",
"blood",
"counts",
"immunology",
"tropical",
"diseases",
"preventive",
"medicine",
"signs",
"and",
"symptoms",
"platelets",
"neglected",
"tropical",
"diseases",
"transfusion",
"medicine",
"vaccination",
"and",
"immunization",
"public",
"and",
"occupational",
"health",
"infectious",
"diseases",
"thrombocytopenia",
"animal",
"cells",
"dengue",
"fever",
"hematology",
"diagnostic",
"medicine",
"blood",
"anatomy",
"cell",
"biology",
"prophylaxis",
"fevers",
"physiology",
"hemorrhage",
"biology",
"and",
"life",
"sciences",
"cellular",
"types",
"viral",
"diseases",
"vascular",
"medicine",
"blood",
"transfusion"
] | 2016 | Potential Harm of Prophylactic Platelet Transfusion in Adult Dengue Patients |
Place cells in the mammalian hippocampus signal self-location with sparse spatially stable firing fields . Based on observation of place cell activity it is possible to accurately decode an animal’s location . The precision of this decoding sets a lower bound for the amount of information that the hippocampal population conveys about the location of the animal . In this work we use a novel recurrent neural network ( RNN ) decoder to infer the location of freely moving rats from single unit hippocampal recordings . RNNs are biologically plausible models of neural circuits that learn to incorporate relevant temporal context without the need to make complicated assumptions about the use of prior information to predict the current state . When decoding animal position from spike counts in 1D and 2D-environments , we show that the RNN consistently outperforms a standard Bayesian approach with either flat priors or with memory . In addition , we also conducted a set of sensitivity analysis on the RNN decoder to determine which neurons and sections of firing fields were the most influential . We found that the application of RNNs to neural data allowed flexible integration of temporal context , yielding improved accuracy relative to the more commonly used Bayesian approaches and opens new avenues for exploration of the neural code .
Place cells , pyramidal neurons found in CA1 and CA3 of the mammalian hippocampus [1–4] , exhibit spatially constrained receptive fields , referred to as place fields . In general , the activity of place cells is considered to be stable [5 , 6]; place fields are typically robust to the removal of specific environmental cues [7 , 8] , persist between visits to a location [9] , and in the open field do not strongly depend upon an animal’s behaviour [2 , 5] . Upon exposure to a novel enclosure the firing correlates of place cells rapidly ‘remap’; place fields change their firing rate and relative position , forming a distinct representation for the new space [10–12] . For these reasons place cells are widely held to provide the neural basis of self-location , signalling the position of an animal relative to its environment and thus being a necessary element for the control of spatial behaviours , such as navigation , and the retention of spatial memories [2] . Unsurprisingly then , given information about the activity of a population of place cells , it is possible to decode the location of an animal with a relatively high degree of accuracy [13 , 14] . However , although place cell activity is strongly modulated by self-location this relationship is non-trivial and not exclusive . For example , during rest and brief pauses , but also during motion , the place code can decouple from an animal’s current location and recapitulate trajectories through the enclosure [15]; ‘replaying’ previous experience [16] or , perhaps , foreshadowing future actions [17] . Similarly , when animals run on linear runways or perform constrained navigational tasks , such as T-maze alternation , place cell activity becomes strongly modulated by behaviour , disambiguating direction of travel [18] , prospective and retrospective trajectories [19 , 20] , and the degree of engagement with a task [21] . Furthermore , although place fields are repeatable they are not static . Even though remapping occurs rapidly in a novel environment , the newly formed firing fields continue to be refined during subsequent experience , a process that appears to persist for several hours [10 , 13 , 22 , 23] . Even in familiar environments , that animals have visited many times , the spatial activity of place cells is known to exhibit incremental changes that can result in the generation of distinct spatial codes [23–26] , which might be important for encoding goal locations [27] or other non-spatial variables [28] . As such , although hippocampal activity provides considerable information about an animal’s self-location the representation is dynamic: accumulating changes and sometimes encoding other variables both spatial and non-spatial . A common approach used to interrogate neural representations , such as that of place cells , is decoding; the accuracy with which a variable , such as self-location , can be decoded from the brain , places a useful lower limit on the amount of information present [13 , 14] . In the case of place cells , decoding methodologies typically apply a Bayesian framework to calculate a probability distribution over the the animal’s position , given the observed neural data [14 , 16 , 29] . Decoding to a specific location is then accomplished via a maximum likelihood estimator applied to the probability distribution . However , the accuracy of Bayesian methods depends on accurate information about the expected activity of neurons . For place cells , activity recorded over the course of tens of minutes is typically used to estimate the firing rate of each cell at different points in the animal’s enclosure , with instantaneous rates assumed to exhibit Poisson dynamics . However , for the reasons outlined above , it is not clear that hippocampal activity can be modelled in this way . Indeed , the variability of place cell firing rates is known to greatly exceed that expected from a Poisson process [30] . As such , it is likely that Bayesian methods , as currently applied , do not provide an accurate reflection of the accuracy with which the hippocampus encodes self-location . To better understand these constraints , we trained a deep recurrent neural network ( RNN ) [31–33] to decode rodent location from the firing rates of CA1 neurons . At each time step the network was presented with a vector corresponding to the spike counts of hippocampal cells within a given time window . After accumulating information for 100 time-steps the network was required to predict the animal’s location—supervision being provided in the form of the animal’s true location . We found that decoding with the trained RNN was consistently more accurate than a standard Bayesian approach [14 , 29] with flat priors ( essentially a MLE ) as well as a Bayesian decoder with priors informed by the animals’ historic activity [14 , 16] . This demonstrates that RNNs are able to capture the relationship between a temporal sequence of neural activity and an encoded variable without the necessity of explicit assumptions about the underlying noise model or complicated hand-coded priors . Further , inspection of the trained network allowed us to identify both the relative importance of individual neurons for accurate decoding and the locations at which they were most informative . Thus , not only does the accuracy of the RNN set a new limit for the amount of information about self-location encoded by place cells but more generally this work suggests that RNNs provide a useful approach for neural decoding and provide a means to explore the neural code .
To test the RNN’s ability to decode rodent location based on hippocampal activity we first characterized the decoding error for a single animal foraging in a 2D arena ( 1m x 1m square ) . Single unit recordings were made using tetrodes from region CA1 of five rats . In all animals less than 10% of the recorded neurons were interneurons , characterized by narrow waveforms and high firing rates . Rat R2192 yielded the greatest number of simultaneously recorded hippocampal neurons ( n = 63 ) . Since the number of recorded neurons is expected to correlate with decoding accuracy , we first focused on this particular animal . Neural data was processed to extract action potentials and these were assigned to individual neurons using the amplitude difference between tetrode channels [35] ( see Methods ) . The input features for the RNN-decoder then consisted of spike counts for each neuron within a set of time windows . The length of time windows was parametrically varied between 200 ms and 4000 ms in 200 ms increments . Each consecutive window started 200 ms later than previous one ( this means 0% overlap for 200 ms windows , 50% overlap for 400 ms windows , 80% overlap for 1000 ms windows , etc . See “Feature extraction” in Methods ) . The network was presented with spike counts from 100 windows before being asked to predict the animal’s location at the center of the latest window . As the RNN training process is stochastic , 10-fold cross validation ( CV ) procedure was run multiple times for each window size . For each of these runs we trained 10 models ( for each fold of CV ) and extracted the mean and median results across the folds . Black dots on Fig 1 correspond to these different realizations of the 10-fold CV procedure ( notice multiple dots per window size ) . 10-fold cross validation was also applied to the Bayesian decoders . For both the mean ( Fig 1a ) and median ( Fig 1b ) of the validation errors , the error curve was convex with lowest errors obtained at intermediate values . Best median decoding accuracy was achieved with time window of 1200 ms ( median error = 10 . 18±0 . 23 cm ) . Best mean decoding was achieved for a time window of 1400 ms ( mean error = 12 . 50±0 . 39 ) cm ) . Using longer or shorter time windows lead to higher errors—likely because spike counts from shorter windows are increasingly noisy , while the animal’s CA1 activity is less specific to a particular location for longer windows . For all time windows , the accuracy of the RNN considerably exceeded that of both the simple Bayesian decoder ( dashed red line ) and the Bayesian decoder with memory ( solid red line ) . The lowest median decoding error achieved with the simple Bayesian decoder was 12 . 00 cm ( 17 . 9% higher than for the RNN; this accuracy was obtained with multiple different window sizes ) , lowest mean error was 15 . 83 cm with a 2800 ms window . The Bayesian decoder with priors informed by the animal’s historic activity was more accurate than the naive Bayes approach but was still considerably less accurate than the RNN ( lowest mean decoding error was 15 . 46 cm with 2000 ms windows , and lowest median 11 . 31 cm with 1400ms windows ) . The RNN has the ability to flexibly use information from all 100 input vectors and thus integrates contextual information over time . This results in lower mean and median errors as compared to the two baseline Bayesian approaches . The naive Bayesian method with flat priors does not have access to information about past activity , resulting in lowest accuracy . Equally , the Bayesian decoder with memory incorporates past activity to form an informed prior , but does this in a predefined manner , being less flexible than the RNN . Notice also that the RNN approach achieves its best results for shorter time windows than the Bayesian approaches ( see also Table 1 for optimal window size results from other animals ) . We hypothesize that the RNN’s efficient use of contextual information helps it to overcome the stochastic noise in the spike counting obtained for shorter time windows . Beyond the global descriptors of mean and median error , we also inspected the distribution of decoding error sizes ( Fig 2a ) . The RNN error distribution followed a unimodal curve with most predictions deviating from the rat’s true position by 6-8 cm and few errors were larger than 35 cm ( 1 . 7% of errors > 35 cm , see Fig 2a ) . The Bayesian classifiers achieve more very low ( <2 cm ) errors , but also an abundance of very large ( >50 cm ) errors ( ≈8% of errors > 35 cm , ≈2 . 7%> 50 cm; for both Bayesian classifiers ) . In many cases single unit recordings yield fewer than the 63 neurons identified from R2192 . We hypothesised that the RNN’s ability to use contextual information would be increasingly important in scenarios where neural data was more scarce . To test this prediction we randomly downsampled the dataset available from R2192 , repeating the training and decoding procedure for populations of neurons varying in size from 5 to 55 in increments of 5 . As expected we saw that decoding accuracy reduced as the size of the dataset reduced . However the RNN was considerably more robust to small sample sizes , decoding with an error of 30 . 9 cm with only 5 neurons vs . 46 . 0 cm error for the Bayesian decoder ( Fig 2b ) . In total we analyzed recordings from five animals as they foraged in a 2D open field environment ( 1m x 1m square ) . For each of these 5 datasets , we determined the best performing time window size ( similarly to Fig 1 ) for the RNN architecture ( composed of 2-layers of 512 LSTM units ) , simple Bayesian decoder ( MLE ) , and Bayesian decoder with memory . The optimal time window sizes for the five 2D foraging datasets are given in top half of Table 1 along with the length of the recording and the number of identified neurons . In the 2D decoding task , for different animals , the mean error ( mean across cross validation folds ) ranged between 12 . 5-16 . 3 cm and median between 10 . 3-13 . 1 cm ( Fig 3a and 3b ) . Interestingly , despite some recordings yielding as few as 26 or 33 cells , the decoding accuracy using RNNs is roughly similar . In all cases the mean and median decoding results from the RNN decoder outperformed both the standard Bayesian approach and Bayesian with memory . We also performed decoding on 1D datasets recorded while the same 5 animals shuttled back and forwards on a 600 cm long Z-shaped track for reward placed at the corners and ends ( Table 1 ) [36] . As before we applied RNN and Bayesian decoders to 10-fold cross validated data , selecting in each case the optimal time window size ( Table 1 ) . The RNN decoder greatly outperformed the two Bayesian decoders in all 5 data sets when comparing mean errors ( Fig 3c ) . In the 2D task the largest possible error was 141 . 7 cm ( if the predicted location is in the corner diagonally opposite to the true location ) , whereas in 1D task it is 600 cm ( if the opposite end of the track is predicted ) . In the 1D task a small number of extremely large errors will inflate the mean error , whereas the median will be less affected ( Fig 3c and 3d ) . Examining the median errors we found that RNN outperformed the Bayesian decoders in all cases . However for four of the five animals the difference in error was relatively small ( Fig 3d ) . For the fifth rat with the fewest recorded cells ( R2117 , n = 40 ) , the RNN clearly outperformed the Bayesian approaches , having a median decoding error that was almost half that of what the two types of Bayesian decoders achieved . Next to understand how behavioural and neural variability influenced decoding accuracy we focused on the results obtained from rat R2192 in the 1m square—the animal with the greatest number of neurons and the lowest decoding error . First we examined the decoding error as a function of the rat’s location . It is important to note that the animals’ behaviour is non-uniform—the rats visit some parts of the arena more often than others ( see Fig 4a ) . Since more training data is available for frequently visited regions it is expected that any decoding approach would be most accurate in those locations . The spatial distribution of decoding error for R2192 seems to confirm this conjecture—well sampled bins in the center of the enclosure and portions of its borders are more accurately decoded ( Fig 4b ) . To confirm this , we calculated the correlation between the decoding error and the number of training data points located within 10 cm radius of the predicted data point , finding a significant negative correlation ( Spearman’s Rank Order , r = −0 . 16 , pval ≪ 0 . 001 , dof = 4412 ) . Another important factor influencing the decoding accuracy is the distribution of neural activity across the 2D enclosure . In particular , place fields of the recorded hippocampal cells do not cover the enclosure uniformly . Clearly it would be difficult for the algorithm to differentiate between locations where no cell is active . As such , it is likely that areas where more neurons are activated are decoded with higher precision . Our results confirm that the sum of spike counts across neurons at a given location is strongly anti-correlated with the prediction error made at that location ( Fig 4c , Spearman’s Rank Order , r = −0 . 31 , pval ≪ 0 . 001 , dof = 4412 ) . We also inspected the x and y components of the decoding error separately . Previous work suggests that , in the case of grid cells , contact with an environmental boundary results in a reduction of error in the representation of self-location perpendicular to that wall [37] . Such a relationship would be expected if boundaries function as an elongated spatial cue , used by animals to update their representation of self-location relative to its surface . Accordingly , we found that for RNN decoding based on CA1 neurons , the decoding accuracy orthogonal to environmental boundaries increased with proximity to that boundary ( Fig 4d , Spearman’s Rank Order between error and distance to wall in the region up to 25cm from the wall , r = 0 . 31 , p ≪ 0 . 001 , dof = 3968 ) . The result also held for x ( r = 0 . 35 , p ≪ 0 . 001 , dof = 2101 ) and y ( r = 0 . 25 , p ≪ 0 . 001 , dof = 1855 ) coordinates separately . Conversely , decoding error parallel to the boundary was not modulated by proximity . Furthermore , an additional factor that seemed to influence prediction accuracy was the animal’s motion speed . Predictions were more reliable when the rat was moving as opposed to stationary . The mean prediction error for speeds below 0 . 5 cm/s being 16 . 5 cm , higher than the 12 . 1 cm average error for all speeds above 0 . 5 cm/s ( two-sided Welch’s t-test , t = 10 . 62 , p ≪ 0 . 001 , median errors 8 . 68 cm and 7 . 74 cm accordingly ) . It seems plausible that the lower prediction accuracy during stationary periods might be due to place cells preferentially replaying non-local trajectories during these periods [38] . A second interesting observation is that the prediction error does not increase at higher speeds ( two-sided Welch’s t-test between errors in data points where speed is in range from 0 . 5 cm/s to 10 . 5cm/s and errors in data points with speed above 10 . 5cm/s , t = 0 . 31 , p = 0 . 76 ) . The accuracy of any neural decoder represents a useful lower bound on the information about the decoded state contained by the recorded neurons . Thus , a biologically relevant question is how such information is distributed among the neurons , across space and time . In short we asked which features of the neuronal activity are the most informative at predicting the animal’s position . To this end we conducted two different types of sensitivity analyses to measure robustness to different types of perturbations . For a visualization of the representations learned by the RNN , see the dimensionality reduction analysis ( using t-SNE ) in S1 Text , S2 and S3 Figs .
We have shown that the sequential processing afforded by an artificial recurrent neural network ( RNN ) provides a flexible methodology able to efficiently decode information from a population of neurons . Moreover , since a RNN decoder is a neural network , it represents a biologically relevant model of how neural information is processed . Specifically , when applied to hippocampal neural data from freely moving rats [2] , the network made use of the past neural activity to improve the decoding accuracy of the animals’ positions . In a 2D open field arena ( 1m x 1m ) , the RNN decoder was able to infer position with a median error of between 10 . 3 cm to 13 . 1 cm for 5 different rats . These results represented a marked improvement over both a simple Bayesian decoder using a flat prior [14 , 16 , 29] , which bases its decision solely on spike counts from a single time window centered around the moment of position measurement , as well as a Bayesian decoder incorporating priors informed by the animals’ behaviour and recent spiking history [14] . Bayesian methods are known to be optimal decoders when using appropriate priors [41] . However , when applied to neural decoding it is difficult to determine these appropriate priors—as a result sub-optimal approximations are commonly used . Hence we propose that RNNs offer a practical methodology to incorporate sequential context without the need to choose or estimate specific priors over high-dimensional spaces . The improvement in 2D position decoding observed for the RNN was mirrored by similar results from a 1D decoding task using hippocampal recordings made while animals ran on a 6 meter track . Here again , the RNN decoder achieved equal or better results than the Bayesian approaches . Making use of the past neural activity as contextual information , the RNN seems more robust to noise than the two Bayesian classifiers . In particular when using shorter time windows the spike counts become noisier and the Bayesian models’ prediction accuracy degraded rapidly . In contrast the RNN decoder was more resistant to the variability of spike counts , likely due to its ability to combine information over the complete sequence of past inputs . Similarly , in situations where fewer neurons were available and hence the total amount information was reduced , the RNN exhibited a pronounced advantage over the Bayesian decoders . Equally , in the 1D task the benefit of the RNN was most evident for animal R2217 , which had the fewest recorded neurons . Nevertheless notice that fewer recorded neurons does not necessarily mean lower accuracy . As described in Section 2 . 3 . 1 , the error depends strongly on the amount of training data available ( length of recording ) and the quality of the cells ( amount and location of firing ) . Taken together these results suggest that RNN decoding of neural data may prove to be particularly useful in situations where large populations of neurons are not available or are difficult to stably maintain . Beyond quality and amount of data available , the size of error the RNN decoder made was also seen to depend on the distance of the animal from the walls and its instantaneous speed . At higher speeds ( above 10 . 5 cm/s ) the decoding accuracy does not decrease , but when the animal is immobile ( below 0 . 5 cm/s ) the error was significantly higher than when in motion . We hypothesize that while stationary hippocampal activity may reflect non-local activity associated with sharp-wave ripple states [38] . Beyond providing more accurate decoding , the neural network approach also provides a new means of conducting sensitivity analyses . While knockout-type sensitivity analyses can be applied to both Bayesian and RNN decoders , the latter approach also supports gradient analyses . The two types of sensitivity—knockout and gradient—are correlated , but not identical . By design knockout analyses answers how the system behaves if an input is completely removed , while gradient analyses investigated how the system behaves in response to small perturbations to that input . Having access to the gradients with respect to each spike count makes is possible to pose new questions about the dynamic variability of the information content of individual neurons .
All procedures were approved by the UK Home Office , subject to the restrictions and provisions contained in the Animals Scientific Procedures Act of 1986 . Deep learning is a class of algorithms that learn a hierarchy of representations or transformations of the data that make the problem of classification or regression easier [31 , 33] . In particular , deep neural networks , inspired by biological neural circuits , consist of layers of computational units called neurons or nodes . The deepness means that there are multiple “hidden” layers between the input and output . By tuning the connection weights between its layers a neural network can learn to approximate a function from a set of examples , i . e . , pairs of related input and output data . In this work we are interested in training a neural network to decode the rat spatial coordinates from the activity recorded from its hippocampal cells . Whereas feed-forward neural networks learn to predict an output based on a single input , recurrent neural networks ( RNNs ) can deal with series of inputs and/or outputs [32 , 33] . In particular , a recurrent network can preserve information from previous inputs by means of feedback connections ( loops between its units ) . Having access to past information can be useful to minimize errors in certain tasks . Such memory of past inputs also means that the order in which the inputs are presented to the network may change the eventual predictions , and thus integrate contextual information over time . A naive implementation of RNNs can only maintain information from a few past inputs , making it possible for the network to detect only immediate trends , but not long timescale dependencies . Advanced realizations of recurrent networks , such as long-short term memory ( LSTM ) [34] and gated recurrent units ( GRU ) [42 , 43] have specific architecture and sets of parameters that control to what extent past activity should be remembered or overwritten by a new input [43] . This makes them capable of integrating knowledge over a longer sequence . Through using past inputs as contextual information these networks have achieved outstanding performance with noisy sequential data such as text and speech . Spatial decoding was also implemented using a Bayesian framework [14 , 29 , 49] subject to 10-fold cross validation ( see also the next subsection ) . Specifically , for each fold , 90% of the data was used to generate ratemaps for hippocampal neurons—spike and dwell time data were binned into 2 cm square bins , smoothed with a Gaussian kernel ( σ = 1 . 5 bins ) , and rates calculated by dividing spike numbers by dwell time . Note , for the Z-maze only , positional data was linearised before binning . Next , with the remaining 10% of the data , using temporal windows ( 200 ms to 4000 ms ) each of which overlapped with its neighbours by half , we calculate the probability of the animal’s presence in each spatial bin given the observed spikes—the posterior probability matrix [14 , 16 , 29] . Specifically during a time window ( T ) the spikes generated by N place cells was K = ( k1 , … , ki , … , kN ) , where ki was the number of spikes fired by the i−th cell . The probability of observing K in time T given position ( x ) was taken as: P ( K | x ) = ∏ P o i s s o n ( k i , T α i ( x ) ) = ∏ i = 1 N ( T × α i ( x ) ) k i k i ! × e - T α i ( x ) , where x indexes the 2 cm spatial bins defined on the Z-track/foraging environment and αi ( x ) is the firing rate of the i − th place cell at position x , derived from the ratemaps . In the case of the simple Bayes decoder , to compute the probability of the animal’s position given the observed spikes we applied Bayes’ rule , assuming a flat prior for position ( P ( x ) ) , to give: P ( x | K ) = R [ ∏ i = 1 N α i ( x ) k i ] × e - T ∑ i = 1 N α i ( x ) , where R is a normalizing constant depending on T and the number of spikes emitted . Note in this case we do not use the historic position of the animals’ to constrain P ( x|K ) thus the probability estimate in each T is independent of its neighbours . Finally , position was decoded from the posterior probability matrix using a maximum likelihood method—selecting the bin with the highest probability value . Decoding error was then taken as the Euclidean distance between the centre of the decoded bin and the centre of the bin closest to the animal’s true location . Finally for the Bayes decoder with memory we made two further changes . First for each animal P ( x ) , the probability of being at position x , was calculated directly from the experimental data for the entire trial , giving: P ( x | K ) = R × p ( x ) × [ ∏ i = 1 N α i ( x ) k i ] × e - T ∑ i = 1 N α i ( x ) , Second , following [14] we incorporated a continuity constraint such that information about the animal’s decoded position in the previous time step was used to calculate the conditional probability of P ( xt | spikest , xt−1 ) . P ( x t | s p i k e s t , x t - 1 ) = C × P ( x | K ) × n o r m D i s t ( x t - 1 , s i g m a ) Where C is a normalising constant and normDist is a normal distribution centred on the animal’s decoded position in the previous time step with sigma equal to the mean distance travelled per time step in the previous 15 time steps multiplied by a scaling factor which was set to 1 for the open field decoding and 5 for linear track . The implementations of these two approaches can be found in the Bayesian folder of the GitHub repository https://github . com/NeuroCSUT/RatGPS . Please notice that the data for MLE and Bayesian apporaches must be downloaded and added to the Bayesian/Data folder manually , as the files were too large do be added to GitHub . As instructed in the README files in the repository , the data can be found via DOI ( https://doi . org/10 . 5281/zenodo . 2540921 ) . The reported errors for both Bayesian and RNN approach are measured using a 10-fold cross validation method that divides the D data points between training and validation sets . Due to the overlap between consecutive time windows a random assignment of data points to training and validation sets would imply that for most of the validation data points a highly correlated neighbouring sample can be found in the training set . This would result in an artificially high validation accuracy that does not actually reflect the model’s ability to generalize to new , unseen data . Instead , in our analysis the first fold in cross validation simply corresponds to leaving out the first 10% of the recording time and training the model on the last 90% of data . The second fold , accordingly , assigns the second tenth of recordings to the validation set , and so on . For RNNs we need to additionally discard 99 samples at each border between training and validation sets . Remind that the input for RNNs is a series of 100 spike count vectors—to avoid any overlap between training and test data we remove validation data points that have at least one shared spike count vector with any training data point . For each fold we train a model on the training set and calculate the error on the validation set . All reported errors are the validation errors—errors that the models make on the one tenth of data that was left out of the training procedure . To increase the reliability of the results , we perform 10-fold cross validation procedure multiple times and report the mean and median of the errors . This is done only for the RNN decoder , because the Bayesian decoder is deterministic and repeating cross-validation procedure multiple times is not necessary . | Being able to accurately self-localize is critical for most motile organisms . In mammals , place cells in the hippocampus appear to be a central component of the brain network responsible for this ability . In this work we recorded the activity of a population of hippocampal neurons from freely moving rodents and carried out neural decoding to determine the animals’ locations . We found that a machine learning approach using recurrent neural networks ( RNNs ) allowed us to predict the rodents’ true positions more accurately than a standard Bayesian method with flat priors ( i . e . maximum likelihood estimator , MLE ) as well as a Bayesian approach with memory ( i . e . with priors informed by past activity ) . The RNNs are able to take into account past neural activity without making assumptions about the statistics of neuronal firing . Further , by analyzing the representations learned by the network we were able to determine which neurons , and which aspects of their activity , contributed most strongly to the accurate decoding . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] | [
"machine",
"learning",
"algorithms",
"action",
"potentials",
"medicine",
"and",
"health",
"sciences",
"neural",
"networks",
"applied",
"mathematics",
"membrane",
"potential",
"social",
"sciences",
"electrophysiology",
"neuroscience",
"learning",
"and",
"memory",
"simulation",
"and",
"modeling",
"algorithms",
"recurrent",
"neural",
"networks",
"animal",
"behavior",
"mathematics",
"cognition",
"artificial",
"intelligence",
"memory",
"zoology",
"research",
"and",
"analysis",
"methods",
"computer",
"and",
"information",
"sciences",
"animal",
"cells",
"behavior",
"mathematical",
"and",
"statistical",
"techniques",
"bayesian",
"method",
"cellular",
"neuroscience",
"psychology",
"cell",
"biology",
"physiology",
"neurons",
"biology",
"and",
"life",
"sciences",
"cellular",
"types",
"physical",
"sciences",
"cognitive",
"science",
"neurophysiology",
"machine",
"learning"
] | 2019 | Efficient neural decoding of self-location with a deep recurrent network |
Dengue is prevalent among Malaysia's indigenous peoples , known as the Orang Asli , and it poses a serious health threat to them . The study aims to look at the socio-demographic factors , health beliefs , and knowledge about dengue and its association to dengue prevention practices among Orang Asli communities in Peninsular Malaysia . A cross-sectional survey was conducted in 16 randomly selected Orang Asli villages from eight states in Peninsular Malaysia from April 2012 until February 2013 . A total of 560 Orang Asli were interviewed and 505 completed the survey . Slightly above half of the participants ( n = 280 , 55 . 4% ) had a total dengue prevention score of 51–100 ( of a possible score of 0–100 ) . Multivariate analysis findings showed dengue knowledge , perceived barriers to perform dengue prevention , fogging frequency , and perceived susceptibility to dengue fever as significant factors associated to dengue prevention practices . Participants with a lower dengue knowledge score ( score 0–18 ) were less likely ( OR = 0 . 63 , 95%CI = 0 . 44–0 . 92 vs . score 19–36 , P = 0 . 015 ) to practice dengue prevention . Participants with low perceived barriers to prevent dengue ( score of 1–5 ) were more likely ( OR = 2 . 06 , 95%CI = 1 . 21–3 . 53 , vs . score of 6–10 , P = 0 . 008 ) to practice dengue prevention . Villages that were not fogged ( OR = 0 . 49 , 95%CI = 0 . 24–0 . 99 , P = 0 . 045 ) or rarely fogged ( OR = 0 . 40 , 95%CI = 0 . 22–0 . 75 , P = 0 . 004 ) had lower dengue prevention practices than villages that were fogged often . Participants with low perceived susceptibility of acquiring dengue ( score of 1–5 ) were less likely ( OR = 0 . 54 , 95%CI = 0 . 33–0 . 89 vs . score of 6–10 , P = 0 . 018 ) to practice dengue prevention measures . Findings imply that educational and health programmes should focus on enhancing dengue knowledge and perceived susceptibility of acquiring dengue and reducing perceived barriers to performing dengue prevention practices among the Orang Asli . More outreach on mosquito control campaigns should be carried out especially in villages where mosquito fogging is frequent .
Dengue fever is a serious problem overwhelming the world: annually , there are about 50–100 million dengue infections [1] which include 500 , 000 dengue hemorrhagic fever DHF cases with 22 , 000 deaths , mostly among children [2] . In the year 2085 , is it estimated at least 50–60% of the world population will be at risk of dengue fever [3] . In Malaysia , the morbidity rate for DHF was the highest recorded from 1987 to 1991 among adults aged 20–29 years [4] . WHO recorded an increase of two times the incidence rate of dengue fever from the year 2012 , with 21 , 900 cases , compared to year 2013 , with 49 , 346 cases , in Malaysia . The prevalence of dengue fever in the rural areas of Malaysia was estimated to range from 24% in the Lundu District , Sarawak [5] to about 91% throughout the Malaysian population [6] . The indigenous peoples in Malaysia , known as “Orang Asli” , make up 1% of the total population in Malaysia [7] . Approximately , 61% of the Orang Asli live at the fringes of the jungle or rural areas , while 37% live deep within the jungle , and only 1% are found in or close to urban areas . A recent study conducted in Malaysia , found that prevalence of dengue was significantly higher among the rural areas than in urban areas [8] . A study conducted in 1956 [9] among the Orang Asli population in Peninsular Malaysia showed that virtually most adults above the age of 30 years from the Temuan and Semai community in Hulu Langat , Selangor , had been affected by dengue fever . Another study conducted in 1986 [10] showed that 73% of the Temuan Orang Asli community in Kampung Tanjong Rabok , Selangor had been affected by dengue fever and its related viruses . Human behaviour contributes majorly in controlling the breeding grounds for these mosquitoes and reducing the number of the mosquito population [11] . Vector control is one of the effective method in controlling and preventing dengue fever[12] [13] . Vector control can be done by frequent fogging in endemic areas which is mostly done outdoors . However , the Aedes aegypti mosquito tends to rest hidden indoors , making it hard for insecticide to reach adult mosquitoes [14] . One of the few methods of dengue prevention is eliminating the breeding sites of dengue mosquitoes indoors and outdoors . The success is accorded in Thailand where eliminating mosquito breeding has definitely reduced the number of dengue cases in the region [15] . The success of efforts in dengue prevention and control is mainly from improving public and household environmental sanitation , water supply , and alteration of human behaviour towards dengue fever [16] . Health beliefs were found to influence dengue preventive practices [17] . It is reported that the health belief model ( HBM ) is by far the most commonly used theory in health education and health promotion [18] . The underlying concept of the original HBM is that health behaviour is determined by personal beliefs or perceptions about a disease and the strategies available to decrease its occurrence [19] [20] . The HBM consists of four perceptions that serve as the main constructs of the model: perceived seriousness , perceived susceptibility , perceived benefits , and perceived barriers [21] and will be implemented to assess the health behaviour pattern accordingly . It has been recognized that socio-demographic characteristics have an important impact on dengue prevention practices and control . Younger and married people reported higher prevention practices against dengue fever compared to those from older age groups and those who were single . A study conducted in Malaysia reported that eliminating breeding sites and mosquito prevention practices were higher among the Malaysian rural population compared to the aborigines [22] . This could perhaps be explained by the rural populations having a higher level of education and living nearer to health facilities . Therefore , this shows that socio-demographic characteristics are an important factor in dengue elimination and prevention . Knowledge or awareness has been reported as important in dengue prevention and control . According to a recent study , inadequate knowledge about dengue is a major risk factor faced in the elimination of dengue [23] . A recent study found that inadequate and lack of knowledge about signs and symptoms , transmission of dengue , and preventive practices can increase the spread of dengue fever among the Malaysian population [8] , [22] . A past study conducted among Malaysians found that they generally had good knowledge of dengue fever and its prevention [24] . However , evidence was found that higher knowledge did not necessarily result in adoption of the recommended preventive behaviour [25]; [26] . Therefore , further investigation is important to find out the association between knowledge and dengue prevention practices . One of the important factors that contribute to the spread of dengue fever is the intensity of the dengue causing mosquitoes . Intensity of dengue causing mosquitoes increases when there are more available breeding sites and food . Other than that , the density of vegetation in a surrounding area is a potential habitat for Aedes mosquito breeding [27] . Orang Asli live in jungles and in surroundings of highly dense vegetation where many mosquitoes are found and , therefore , are at risk of diseases caused by mosquitoes . The spread of diseases and viruses caused by the mosquito is most effective in very densely populated areas [28] where it feeds almost exclusively on humans . In a study conducted among the Native Americans , the prevalence of infectious disease caused by vector was likely to be attributed to poor living conditions where house crowding with lack of sanitation is common [29] . House crowding is common among the Orang Asli as all family members live under one roof [30] . Dengue prevention practices and the associated factors have never been explored among the Orang Asli in Peninsular Malaysia . Identifying and understanding factors associated to dengue prevention may provide insight into targeted interventions to enhance dengue prevention practice and facilitate authorities in the management of dengue prevention . This study is aimed to look at these factors ( socio-demographic , theoretical constructs of the HBM , and dengue knowledge ) and their association with dengue fever prevention practices .
The sample for this study were Orang Asli originating and living in Peninsular Malaysia . According to the Department of Orang Asli Development ( JAKOA ) , there are eight states in Malaysia where Orang Asli are found . From these eight states , two villages from each state were randomly selected where JAKOA was able to provide assistance in accessing the Orang Asli of the respective states . The research group approached Orang Asli members with JAKOA's supervision to acquire better acknowledgment and responses . Based on their location , the villages were either 1 ) forest fringe areas–Orang Asli villages which were relocated and have access to basic resources such as electricity and pipe water or 2 ) deep within the jungle–Orang Asli villages where most basic resources were not readily available . In total , 16 Orang Asli villages were selected based on 1 ) accessibility of these villages by land transport and 2 ) permission being granted by the head of the village . A cross-sectional study was performed in each household , two people were surveyed: 1 ) resident aged between 18–40 years old , 2 ) resident aged 41 years old or above . If there was more than one eligible person available in a household , one participant was selected randomly . Each household in the selected villages was approached . If participants refused to be interviewed or if the resident of the house was not present , it was regarded as a non-response . Trained enumerators administered the questionnaire to the participants . Inclusion criteria for the study were: 1 ) Orang Asli above 18 years of age and 2 ) originating from and living in the selected villages . The questionnaire consists of five main sections: socio-demographic characteristic , type of house and surrounding environment , self-reported prevention practices and control against dengue fever , health beliefs regarding dengue fever and knowledge of dengue fever . Health beliefs regarding dengue fever was measured using the Health Belief Model ( HBM ) construct . This construct consists of four mains parts: Perceived Threat consists of two sub-parts which measure the participant’s susceptibility to contracting dengue fever and severity of the seriousness of dengue fever . This was measured on a scale of 1–10 , where a higher score indicates higher susceptibility to dengue fever . Perceived Barrier examines the perceptions of barriers to prevent dengue fever among participants . This was also measured on a scale of 1–10 , where a higher score indicates higher barriers . Self-efficacy is measured by the behaviour of participants that successfully execute dengue prevention measures . This is measured by a four-point Likert scale that ranges from 1 ( strongly agree ) to 4 ( strongly disagree ) . Other Contracts and Cue to Actions measures the mosquito problem , frequency of fogging , community participation and other things which effect an individual’s perception which indirectly influences health-related behaviour . Measurement of knowledge of dengue fever consisted of 43 items divided into five sub-parts: Knowledge about 1 ) Dengue and the Aedes mosquito , 2 ) Transmission of dengue , 3 ) Prevention of dengue , 4 ) Signs and symptoms , and 5 ) Treatment , curability and precautionary measures for people infected with dengue . For each item , the participants could choose between “yes” , “no” , or “don’t know” responses . For the analyses , correct responses were scored as 1 and incorrect or “don’t know” responses were scored as 0 . Scores ranged from 0–43 , where higher scores indicate greater knowledge about dengue fever . Self-reported prevention practices against dengue fever and control was sub-divided into three parts: prevention practices of mosquito breeding , prevention practices of mosquito bites , and prevention practices of dengue transmission . The questions were assessed using nine-item , seven-item , and one-item questions respectively . The options for dengue prevention practices were “not at all” , “rarely” , “sometimes” , “often” , and “not applicable” and were assigned points of 0 , 1 , 2 , 3 , and 0 respectively and calculated based on the number of applicable answers . Scores were calculated based on “Number of points obtained” over “Total points of applicable answer” . Results were reported as percentages , where higher percentages indicate higher dengue preventive practices . This questionnaire was adapted from “Community knowledge , health beliefs , practices and experiences related to dengue fever and its association with dengue prevalence” by Wong et al . , 2014 . The Cronbach’s alpha >0 . 70 was reported . Cronbach’s alpha coefficient measurement for prevention of mosquito breeding and mosquito bite were 0 . 791 and 0 . 898 respectively , demonstrating good internal consistency . Cronbach’s alpha coefficient measurement for dengue knowledge was 0 . 916 , showing high internal consistency . The Cronbach’s alpha coefficient measurement for self-reported preventive practices in this study was 0 . 655 , demonstrating a good internal consistency . The study received special permission from the Department of Orang Asli Development ( JAKOA ) and was approved by the Medical Ethics Committee of the University Malaya Medical Centre , Kuala Lumpur ( MEC Ref . No: 896 . 15 ) . Due to cultural reasons and the sensitivity to outside visitors of the Orang Asli community , a JAKOA representative who was known to the Orang Asli community was present to help during the entire study in the selected villages . Care was taken to safeguard all information from participants who agreed to participate in the study which was voluntary . Informed written and signed consent was obtained prior to beginning the interview . Other than descriptive analyses , the data were tested for significant relationship between the associative variables and the outcome variables using chi-square test , where P = < 0 . 05 . The dependent variable ( Percentage Scores of Dengue Prevention Practices ) was associated to the independent variable ( socio-demographic characteristics , HBM construct , cues-to-action and knowledge score ) using crosstab and chi-square analysis to see how the variables were associated with dengue prevention practices . Logistic multivariate regression models were used to see the independent effect of each of these variables on the dependent variables . In the modelling strategy , the independent variables were included if they had a p<0 . 05 on univariate analysis . Associations were expressed with adjusted odds ratios of 95% confidence intervals for each variable included in the multivariate model . All statistical analyses were performed using SPSS 20 . 0 ( SPSS Inc . , Chicago , IL ) . In all analyses , a p-value of less than 0 . 05 was considered statistically significant .
The study participants consisted of 67 . 9% ( n = 343 ) female and 32 . 1% ( n = 162 ) male participants . A majority of the participants were aged between 18–40 years old ( n = 366 , 72 . 5% ) and only 27 . 5% ( n = 139 ) of the participants were more than 41 years old of age . Approximately , 36 . 0% ( n = 182 ) of the participants were from the Temuan tribe . A majority of the participants were living in the forest fringe areas ( n = 319 , 63 . 2% ) , and a minority of the participants were living deep within the jungle ( n = 186 , 36 . 8% ) . Less than half of the participants were primary school educated ( n = 205 , 40 . 9% ) and most of them were unemployed ( n = 253 , 50 . 1% ) . About 38 . 4% of the participants living deep within the jungle had no formal education ( n = 58 ) . A majority of the participants ( n = 347 , 68 . 7% ) have less than RM500 as an average monthly household income as they work in the village as a helping hand and do odd jobs around the village . Only 31 . 3% ( n = 158 , ) of the villagers have an average monthly income of RM500–RM1200 as the participants work as assistant kindergarten teachers in the village and as bus or tourist drivers . About 39 . 0% ( n = 197 ) of the Orang Asli participants reported low density of plants and vegetation surrounding their houses and most of them live in the forest fringe ( n = 133 , 67 . 5% ) . Slightly less than half of the participants ( n = 252 , 49 . 9% ) reported that the density of mosquitoes in their neighbourhood was severe . One third of the participants ( n = 173 , 34 . 3% ) reported that the authorities fog their village occasionally with insecticide . Only a minority of the participants ( n = 60 , 11 . 9% ) reported that their village was fogged often , while a majority of the participants reported that their village in the forest fringe was fogged occasionally ( n = 136 , 78 . 6% , P = 0 . 001 ) . In the self-reported survey among 505 Orang Asli participants , only 2 . 8% ( n = 14 ) of the participants have had dengue fever . Among the 14 participants , only 85 . 7% ( n = 12 ) of the participants have been hospitalized for dengue fever ( Table 1 ) . Table 2 shows that 85 . 9% ( n = 434 ) of the participants correctly answered that dengue fever is transmitted by a mosquito . Most of them ( n = 324 , 64 . 2% ) know that dengue fever is caused by the Aedes mosquito . Only a minority of the participants ( n = 89 , 17 . 6% ) know that the Aedes mosquitoes do not live in places with a lot of plants and 39 . 8% ( n = 201 ) know that dengue is a virus . Slightly more than half of the participants know that dengue fever can spread by an Aedes mosquito biting an infected person and then biting another ( n = 296 , 58 . 6% ) . The mean total knowledge score for the overall sample was 18 . 4 , ( SD±9 . 45 ) out of a possible score of 43 . In Table 1 , slightly more than half of the participants , 50 . 1% ( n = 253 ) had a range of total dengue knowledge score of 0 to 18 , while 49 . 9% of the participants ( n = 252 ) had a range of total dengue knowledge score of 19 to 36 . Slightly more than half of the participants ( n = 161 , 50 . 5% ) living in the forest fringe had a range of total dengue knowledge score of 19 to 36 , while 51 . 1% of the participants living deep within the jungle ( n = 95 ) had a range of total dengue knowledge score of 0 to 18 . Participants with more than RM500 average monthly income had a range of total dengue knowledge score of 19 to 36 compared to those who earn RM500–RM1200 a month ( n = 96 , 60 . 8% ) . Participants living in village houses ( n = 137 , 56 . 8% ) had a range of dengue knowledge score of 19 to 36 , compared to participants living in single story houses ( n = 115 , 43 . 6% ) under the Housing Project for the Hardcore Poor ( PPRT ) . As shown in Table 1 , the dengue knowledge score was significantly different between religions , tribes , occupations , average monthly incomes , and types of house . The Temuan tribe had a range of knowledge score of 19 to 36 compared to the other tribes investigated ( n = 64 , 71 . 1% ) . Almost 80% of the skilled workers ( n = 16 ) had a range of dengue knowledge score of 19 to 36 compared to the unskilled ( n = 120 , 51 . 7% ) and unemployed ( n = 116 , 45 . 8% ) . The proportion of dengue knowledge score range of 19 to 36 was significantly higher among Orang Asli participants with higher perceived susceptibility of dengue ( level of susceptibility 6–10 ) ( n = 57 , 62 . 6% , P = 0 . 007 ) than participants with lower perceived susceptibility of dengue ( level of susceptibility 1–5 ) . It was found that the proportion of dengue knowledge score range of 19 to 36 was significantly higher in those who agreed on self-efficiency in dengue prevention than those who disagreed on self-efficiency in the prevention of dengue in the chi square test ( n = 85 , 65 . 9% , P = 0 . 001 ) . In this study , more than half of the participants ( n = 280 , 55 . 4% ) had a total dengue prevention practices percentage score range of 51 to 100 . Table 3 shows that most of the participants ( n = 491 , 97 . 2% ) practiced proper disposal of items that can collect rain water . Most of them also practiced proper disposal of household garbage ( n = 479 , 94 . 9% ) and clearing out of the debris that blocked water flow ( n = 476 , 94 . 3% ) . It was also noted that 93 . 9% ( n = 474 ) of the participants practice cleaning the areas surrounding their house frequently as one of the reasons to prevent dengue . Only a minority of the participants ( n = 93 , 18 . 4% ) use Abate or chemicals in water storage containers to prevent dengue mosquito breeding , and 10 . 7% ( n = 54 ) of the participants use mosquito repellent on their body to prevent mosquito bites . The total percentage scores of dengue prevention practices were significantly different among the socio-demographic variables and self-reported variables , specifically the Orang Asli tribes and fogging frequency . As shown in Table 1 , the Semai tribe reported a range of percentage scores of 51–100 for dengue prevention practices ( n = 64 , 71 . 1% ) compared to the other tribes investigated . A significantly higher proportion of participants reported that their village was often fogged had a range of percentage scores of 51–100 for dengue prevention practices ( n = 40 , 66 . 7% ) . The proportion of dengue prevention practices percentage scores in the range of 51–100 was significantly higher among participants with higher perceived susceptibility to dengue fever ( level of susceptibility 6–10 ) ( n = 64 , 70 . 3% ) compared to participants with lower perceived susceptibility to dengue fever ( level of susceptibility 1–5 ) . A significantly higher proportion of participants with a higher perceived barrier to prevention of dengue ( level of barriers 6–10 ) had a range of percentage scores of 0–50 for dengue prevention practices ( n = 41 , 58 . 6% ) compared to participants with lower perceived barriers to prevention of dengue ( level of barriers 1–5 ) ( n = 184 , 42 . 3% ) . It was also found that participants with a range of total dengue knowledge score of 19–36 had significant dengue prevention practices percentage score range of 51–100 ( n = 153 , 60 . 7% ) . In Table 4 , the multiple logistic regressions indicated that participants with a lower dengue knowledge score ( score of 0–18 ) were less likely ( OR = 0 . 63 , 95%CI = 0 . 44–0 . 92 , P = 0 . 015 ) to perform dengue prevention practices compared to the reference group ( participants with a higher dengue knowledge score of 19–36 ) . The results also indicated that the two main constructs of HBM ( perceived susceptibility and perceived barriers ) were significant correlates of dengue prevention practices . Participants with lower perceived susceptibility ( level of susceptibility 1–5 ) was less likely ( OR = 0 . 54 , 95%CI = 0 . 33–0 . 89 , P = 0 . 018 ) to perform dengue prevention practices compared with those with the reference level of susceptibility ( level 6–10 ) . Likewise , those with lower perceived barriers ( level 1–5 ) had a higher likelihood ( OR = 2 . 06 , 95%CI = 1 . 21–3 . 53 , P = 0 . 008 ) to perform dengue prevention practices compared with those with the reference level of barriers ( level 6–10 ) . Orang Asli villages that were not fogged ( OR = 0 . 49 , 95%CI = 0 . 24–0 . 99 , P = 0 . 045 ) or rarely fogged ( OR = 0 . 40 , 95%CI = 0 . 22–0 . 75 , P = 0 . 004 ) had a lower likelihood to perform dengue prevention practices when compared with the villages that were often fogged ( reference group ) .
In Malaysia , the actual number of dengue cases is higher than the number of self-reported dengue cases [2] . In a recent study conducted in Malaysia among the Orang Asli , the Semai Perak community showed the highest prevalence of dengue fever ( > 50% ) . However , in this study , only a minority of the Orang Asli participants reported that they have had dengue fever . The low self-report of dengue experience in this study may indicate that the Orang Asli were not aware that they have had dengue fever . This may imply that they would have had dengue fever and recovered from it without receiving treatment . Hence , it is important to educate the Orang Asli to distinguish the disease and encourage the Orang Asli to visit a health care practitioner to avoid further complication and fatalities due to dengue fever . The participants in this study had a low overall mean knowledge score of 18 . 4 ( out of a possible highest score of 43 ) . Most of the participants were aware that dengue is transmitted by a mosquito and may evidently portray as fever 4–7 days after a mosquito bite . This is in accordance to a study conducted in Jamaica and India , where participants could identify one of the common and obvious sign and symptom is portray of fever [25 , 35] . However , they had low knowledge about joint pains , rashes , and headaches which are the main sign and symptoms of dengue fever . This implies that the majority of the participants did not have good knowledge about the signs and symptoms of dengue fever . Therefore , it is vital to educate the Orang Asli on the signs and symptoms of dengue fever so as to seek immediate treatment to prevent unwarranted death caused by dengue fever . The ideal approach to prevent dengue is to eliminate areas where the dengue mosquito lays its eggs . This study found that the majority of the participants had adequate knowledge of proper disposal of items that can retain water and proper disposal of garbage , which leads to prevention of dengue fever . Cleaning those surroundings of the house which can bring about stagnant water is basic to keeping dengue mosquito from spreading . However , the majority of the participants did not know that the dengue causing mosquito breeds in clean water . Participants presume that it breeds in dirty water , as mostly dengue mosquito breeds outdoors in stagnant water in drains or empty cans which are dirty . The majority of the public associate “dirty” sites outside the house as prominent breeding sites for dengue causing mosquitoes [36] . There appears to be a misconception that dengue mosquitos only breed in dirty water . This implies education should focus on informing the Orang Asli community that in fact dengue mosquitoes prefer to breed in clean stagnant water such as in water jars and flower pots . The perceived susceptibility and perceived barriers constructs of the HBM were significantly associated with prevention practices . Participants with lower perceived susceptibility to dengue fever were less likely to carry out prevention practices against dengue fever . This could be because the majority of the Orang Asli are not aware of the dangers of dengue fever and have not experienced it for themselves . Other studies have shown that if action is likely to occur , the individual perceives the susceptibility of getting the illness [37] . Education programmes need to highlight the risk of getting the disease to create awareness among people who are unconscious of the serious threat of dengue . Testimonials and campaigns can be used from families who have lost a family member due to dengue fever . In this study , participants with lower perceived barriers were significantly associated with higher dengue prevention practices . In order to effectively perform dengue prevention , it is imperative to remove barriers that impede taking action against dengue . Among the barriers are i ) low perceived susceptibility , ii ) lack of self-efficacy , iii ) unsure perceived susceptibility , and iv ) lack of perceived benefit . Authorities should provide facilities to remove these barriers such as increasing community participation to eliminate mosquito breeding sites and increase campaigns to boost responsibilities towards neighbourhood cleanliness to facilitate prevention practices among the Orang Asli communities . It was observed that slightly more than half of the participants took precautions against mosquito bites by using mosquito coils . Participants need to be aware that using mosquito coils should also be practiced at home to prevent mosquito bites as the dengue mosquito can be found everywhere , not only outdoors . In a study conducted by Dieng H . et al . ( 2010 ) it was found that indoor containers contained immature Aedes mosquito eggs which further shows that the Aedes mosquitoes have adapted to breeding indoors due to easy access to a blood source . Therefore , precautions against mosquito bites to prevent dengue fever from spreading should not only be taken outdoors , but in-house prevention is also important . Most of the Orang Asli live below the poverty line , and cannot afford to buy precautionary materials such as mosquito coils or bed nets . Therefore , it is recommended that the government should put more emphasis on introducing cost effective ways of preventing mosquitoes and dengue fever . Only a minority of the Orang Asli were aware of the use of Abate which can prevent mosquito breeding in water containers . Using Abate was found useful in Thailand to reduce Aedes aegypti in water holding containers [38] . The reasons for lack of Abate use among the Orang Asli is due to lack of awareness of Abate in dengue prevention . Another reason is , the Orang Asli are unable to obtain Abate easily as they have to travel into the nearest towns to obtain Abate for their use . A large number of the Orang Asli living in remote areas do not have proper water supply , and therefore mostly depend on containers to store water . Thus , the Orang Asli communities should be educated about Abate in dengue prevention and that the use of Abate is not dangerous to health . The findings from the multivariate analysis found the highest odds of dengue prevention practice among the following: dengue knowledge level , perceived barriers to performing dengue prevention , fogging frequency , and perceived susceptibility to dengue fever . Low knowledge on symptoms and prevention of dengue leads to poorer precautionary practices against dengue fever . Therefore , this study highlights that health education , campaigns , and knowledge of dengue fever is highly needed and necessary in order to boost preventive practices among the Orang Asli community . The multiple logistic regression analysis also found a significant association between fogging frequency in the villages and prevention practices . Fogging is a commonly used method in dengue prevention in many countries [39] . One study proved that fogging has greatly influenced the reduction of dengue cases but is influenced by seasonality and the level of transmission intensity of dengue fever in an area [13] . In Malaysia , fogging is commonly conducted when dengue fever is highly reported in the area . It is likely that fogging creates higher awareness among the community thus triggers higher prevention practices . Therefore , this may imply that fogging may be beneficial both in eradicating adults mosquitoes as well as create awareness and enhance dengue prevention practises among the targeted communities . The study had a few limitations . Orang Asli villages were selected based on accessibility by land transport . This may result in selection bias because of the sample which was not representative of the overall Orang Asli population in Peninsular Malaysia since the Orang Asli living in more remote or inaccessible areas were not surveyed . All information obtained from the interview was self-reported , thus bias towards socially desirable responses and behaviours might exist . Despite some of the limitations in the study , the results provided useful outcomes and knowledge that would guide government officials in the development of programmes and activities to initiate dengue prevention to address the every growing problem of dengue fever . More community based projects should be conducted among the Orang Asli tribes to educate them on dengue fever and its fatal disease . One of the main sources of dengue awareness and education is through mass media . Therefore , more advertisements and bill boards should be put up in outskirts and remote areas emphasizing the seriousness of dengue fever . From this study , several conclusions could be inferred with important implications for dengue fever prevention practices . Firstly , the findings indicate that the level of knowledge about dengue fever , signs and symptoms , and prevention among the participants was low . Secondly there were significant differences of knowledge scores in different religions , states , occupations , average monthly incomes , and types of house . Thirdly , dengue knowledge level , perceived barriers to perform dengue prevention , fogging frequency , and perceived susceptibility to dengue fever were significant factors associated with dengue prevention practices . Therefore , educational and health programmes should focus on enhancing dengue knowledge and perceived susceptibility of acquiring dengue and reducing perceived barriers to perform dengue prevention practices among the Orang Asli communities in Malaysia . Mosquito fogging may create awareness among Orang Asli and enhance their dengue prevention practices . Therefore , less fogged areas need to be the focus of education interventions . Awareness of dengue should be adhered to more enthusiastically within the community and to enhance health beliefs . | Dengue poses a threat to everyone worldwide , including the Orang Asli community in Peninsular Malaysia . Social demographic , knowledge and behavioral factors are essential aspects to control and prevent dengue . This study is aimed to examine these factors and their association with dengue prevention practices . A cross-sectional survey was conducted among 16 randomly selected Orang Asli villages from 8 states in Peninsular Malaysia . Results showed that 1 ) level of dengue knowledge , 2 ) perceived barriers to perform dengue prevention , 3 ) fogging frequency , and 4 ) perceived susceptibility to dengue fever are significant factors of dengue prevention practices . Findings provide important insights into intervention to increase dengue prevention practices among the Orang Asli community . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [] | 2015 | Practices of Dengue Fever Prevention and the Associated Factors among the Orang Asli in Peninsular Malaysia |
A key function of the Vpu protein of HIV-1 is the targeting of newly-synthesized CD4 for proteasomal degradation . This function has been proposed to occur by a mechanism that is fundamentally distinct from the cellular ER-associated degradation ( ERAD ) pathway . However , using a combination of genetic , biochemical and morphological methodologies , we find that CD4 degradation induced by Vpu is dependent on a key component of the ERAD machinery , the VCP-UFD1L-NPL4 complex , as well as on SCFβ-TrCP-dependent ubiquitination of the CD4 cytosolic tail on lysine and serine/threonine residues . When degradation of CD4 is blocked by either inactivation of the VCP-UFD1L-NPL4 complex or prevention of CD4 ubiquitination , Vpu still retains the bulk of CD4 in the ER mainly through transmembrane domain interactions . Addition of a strong ER export signal from the VSV-G protein overrides this retention . Thus , Vpu exerts two distinct activities in the process of downregulating CD4: ER retention followed by targeting to late stages of ERAD . The multiple levels at which Vpu engages these cellular quality control mechanisms underscore the importance of ensuring profound suppression of CD4 to the life cycle of HIV-1 .
Human Immunodeficiency Virus-1 and -2 ( HIV-1 and -2 ) , as well as Simian Immunodeficiency Virus ( SIV ) , selectively target helper T-lymphocytes and macrophages/monocytes by binding of their viral envelope protein , Env , to a combination of two cell-type-specific surface receptors: a type 1 transmembrane protein , CD4 , and a seven-transmembrane chemokine receptor , CXCR4 or CCR5 [1] . An early and lasting effect of infection is the downregulation of CD4 from the host cell surface [2] , [3] . Although it may seem counterproductive for a virus to downregulate its own co-receptor , this event actually promotes the establishment of a robust infection . Indeed , CD4 downregulation prevents ( i ) superinfection by additional virions [4] , ( ii ) retention of newly-synthesized Env precursor in the endoplasmic reticulum ( ER ) [5] , and ( iii ) interference with the release of progeny virions from the cell surface [6] . In addition , CD4 downregulation compromises the ability of T-lymphocytes to become activated in response to immunogenic peptides bound to MHC class II molecules on the surface of antigen-presenting cells [7] . These effects all contribute to propagation of the infection , eventually leading to depletion of CD4-positive cells and development of acquired immunodeficiency syndrome ( AIDS ) in untreated individuals . The most pathogenic of these viruses , HIV-1 , devotes two accessory proteins encoded in its genome , Nef and Vpu , to the task of suppressing CD4 expression [8] , [9] , [10] . Nef is an N-terminally myristoylated , cytosolically-disposed peripheral membrane protein encoded in the genomes of most strains of HIV-1 , HIV-2 and SIV . It is expressed early during infection and functions to accelerate endocytosis of cell surface CD4 by a clathrin/AP-2 pathway [11] , [12] , [13] , followed by delivery of the internalized CD4 to the multivesicular body pathway for eventual degradation in lysosomes [14] . Vpu , on the other hand , is a type III integral membrane protein having a short luminal N-terminal domain ( 3–12 amino acids ) , a single transmembrane span that doubles as an uncleaved signal peptide ( 23 amino acids ) , and a cytosolic C-terminal domain ( 47–59 amino acids ) . Unlike Nef , Vpu is encoded in the genomes of only HIV-1 and a few SIV strains [15] . Vpu is expressed at later stages of infection and acts by targeting newly-synthesized CD4 in the ER for degradation by cytosolic proteasomes [16] , [17] . Together , Nef and Vpu ensure profound and sustained suppression of CD4 expression throughout the HIV-1 infectious cycle [18] , [19] . CD4 downregulation by Vpu depends on an interaction between the cytosolic domains of both proteins [20] . A canonical DpSGxxpS sequence containing two phosphorylated serine ( pS ) residues in the cytosolic domain of Vpu ( residues number 52 and 56 in the NL4-3 variant of HIV-1 used in this study ) then binds β-TrCP1 [21] and β-TrCP2 ( also known as FBXW11/HOS ) [22] , two paralogous F-box adaptor proteins for the cytosolic SCFβ-TrCP E3 ubiquitin ( Ub ) ligase complex . Recruitment of this SCF complex results in ubiquitination of the CD4 cytosolic tail on lysine residues [23] , [24] , marking CD4 for degradation by cytosolic proteasomes [17] . Unlike CD4 , Vpu itself is not ubiquitinated and degraded in this process [25] . Vpu function , therefore , can be likened to that of Ub ligase adaptors , which link substrates to Ub ligases [26] . At first blush , the process of Vpu-induced CD4 degradation evokes the well-known ER-associated degradation ( ERAD ) pathway , which generally functions to dispose of abnormal proteins from the ER [27] , [28] , [29] . Two sets of observations , however , distinguish Vpu function from targeting to typical ERAD . First , the cytosolic SCFβ-TrCP complex does not normally function in ERAD , but is responsible for the ubiquitination and degradation of non-ERAD substrates such as IκBα and β-catenin [30] , [31] . Instead , the ERAD pathway employs several membrane-bound Ub ligases , including the HRD1-SEL1L complex [32] , [33] , TEB4/MARCH-VI [34] , and the GP78-RMA1 complex [35] , [36] ( names and references correspond to the mammalian orthologs ) . Second , genetic analysis involving expression of human CD4 and HIV-1 Vpu in the yeast S . cerevisiae , showed that CD4 degradation in the presence of Vpu is independent of components of the ERAD machinery such as Hrd1p , Hrd3p and Ubc7p ( yeast names; orthologous to the mammalian proteins HRD1 , SEL1L and UBC7 , respectively ) [23] . These observations have led to the notion that the mechanism by which Vpu induces CD4 degradation is fundamentally distinct from ERAD [23] . We have put this notion to the test by examining the requirement of additional components of the ERAD machinery for Vpu-induced CD4 degradation in human cells . Using siRNA and dominant-negative overexpression approaches , we find a requirement for the VCP-UFD1L-NPL4 complex , which is a key component of the ERAD machinery [37] , [38] , [39] , [40] . In addition , we show that degradation depends on ubiquitination of CD4 on not only lysine , but also serine/threonine residues , the first instance in which a cellular Ub ligase activity is implicated in serine/threonine ubiquitination of an ERAD substrate . Inactivation of the VCP-UFD1L-NPL4 complex prevents degradation of CD4 induced by Vpu . Under these conditions , CD4 accumulates in the ER as a properly folded and membrane-associated protein , indicating that the VCP-UFD1L-NPL4 complex plays a role in the dislocation of CD4 from the ER membrane . Dissection of the mechanism by which Vpu retains CD4 in the ER reveals at least two contributing factors: SCFβ-TrCP-dependent ubiquitination of the CD4 cytosolic tail and transmembrane domain ( TMD ) interactions . These findings indicate that Vpu exerts two distinct , separable activities in the process of downregulating CD4: retention in the ER followed by targeting to a variant ERAD pathway .
To determine whether Vpu-induced degradation of CD4 involves any part of the ERAD pathway , we tested for the requirement of key components of the ERAD machinery using a siRNA approach in human cells . To this end , HeLa cells were treated with siRNAs directed to eighteen proteins that mediate various ERAD steps ( Table S1 ) [28] . Human CD4 was then co-expressed with codon-optimized , wild-type HIV-1 Vpu [41] or inactive HIV-1 Vpu bearing mutations of serines 52 and 56 ( Vpu-S52 , 56N ) [25] by transient transfection . Total levels of CD4 were determined by immunoblot analysis . A typical immunoblot for negative and positive siRNA controls is shown in Fig . 1A , and quantification of the results from three independent experiments for all the siRNAs tested is shown in Fig . 1B . In cells treated without ( mock ) or with siRNAs to an irrelevant protein , GAPDH , expression of wild-type Vpu lowered total CD4 levels to <5% of those in control , non-Vpu- or Vpu-S52 , 56N-expressing cells ( Fig . 1 , A and B ) . Combined treatment with siRNAs to β-TrCP1 and β-TrCP2 ( β-TrCP1/2 ) largely protected CD4 from Vpu-induced loss ( to ∼60% of control cells ) , in agreement with previous findings [21] , [22] . The remaining siRNA treatments had various effects , with siRNAs to VCP ( also known as p97 ) , UFD1L or NPL4 causing the greatest degree of protection ( ∼60% of control cells ) . These three cytosolic proteins form a complex that participates in ERAD by extracting or “dislocating” ubiquitinated substrates from the ER membrane [37] , [38] , [39] , [40] . These observations thus pointed to the involvement of at least part of the ERAD pathway in Vpu-induced CD4 degradation . To ascertain that the prevention of CD4 loss by depletion of VCP , UFD1L or NPL4 was due to a block in CD4 degradation , we performed pulse-chase analysis ( Fig . 1 , C–F ) . Cells treated with the corresponding siRNAs and transfected with plasmids encoding CD4 , plus or minus Vpu , were labeled for 2 min with [35S]methionine-cysteine and chased for different times in complete medium . Immunoprecipitation with a conformation-independent antibody showed that expression of Vpu shortened the half-life of CD4 from ∼4 . 8 h ( Fig . S1 , A and B ) [16] to ∼20 min in mock-treated cells ( Fig . 1 , C and D; Fig . S1 , C-F ) . Depletion of VCP , UFD1L or NPL4 largely abrogated the rapid degradation of CD4 induced by Vpu ( Fig . 1 , C and D; Fig . S1 , E and F ) . Use of a conformation-dependent antibody revealed progressive acquisition of a conformational epitope upon folding of the CD4 ectodomain in the absence of Vpu ( Fig . 1 , E and F ) . In mock-treated cells , expression of Vpu prevented accumulation of the folded CD4 species by counteracting folding with degradation ( Fig . 1 , E and F ) . Inhibition of degradation by depletion of VCP , UFD1L or NPL4 restored CD4 ectodomain folding ( Fig . 1 , E and F ) . Subcellular fractionation and Na2CO3 treatment showed that the population of CD4 that was protected from Vpu-induced degradation by VCP , UFD1L or NPL4 depletion remained integrally associated with membranes ( Fig . 1G ) . These experiments thus demonstrated that depletion of the VCP-UFD1L-NPL4 complex prevents degradation of CD4 in the presence of Vpu . Under these conditions , the CD4 ectodomain continues to fold while retaining its association with membranes . Moreover , these findings support the notion that the VCP-UFD1L-NPL4 complex is required for extraction of CD4 from membranes . VCP is a member of the AAA-ATPase superfamily; it comprises an N-terminal domain ( N ) that binds UFD1L and NPL4 , and two AAA-ATPase domains ( D1 and D2 ) ( Fig . 2A ) . To investigate the requirement of these domains in Vpu-induced CD4 degradation , we introduced mutations in the VCP cDNA and overexpressed these mutants together with CD4 and Vpu by transfection into HeLa cells . The fate of CD4 was then examined by pulse-chase analysis . VCP-ΔN is a VCP deletion mutant lacking most of the N domain ( residues 1–185 ) ( Fig . 2A ) [39] . We found that overexpression of this construct had no effect on CD4 degradation ( Fig . 2 , B and C ) , probably because this mutant is unable to assemble with the ubiquitin-binding UFD1L and NPL4 proteins , cannot be targeted to ubiquitinated substrates , and does not compete with endogenous VCP . Mutation of lysine and glutamate residues in the active sites of both ATPase domains to alanine ( VCP-AA ) or glutamine ( VCP-QQ ) residues , respectively ( Fig . 2A ) , is known to prevent ATP binding ( VCP-AA ) or ATP hydrolysis ( VCP-QQ ) [39] . Because of the presence of the N domain , these constructs are capable of assembling with UFD1L and NPL4 , allowing the recruitment of catalytically-inactive VCP to ubiquitinated substrates . The VCP-AA mutant was previously shown to abrogate Vpu-induced CD4 degradation [24] . We confirmed this observation and , in addition , found that the VCP-QQ had a similar effect ( Fig . 2 , B and C ) , indicating that these mutants exerted a potent dominant-negative effect , and that both ATP binding and hydrolysis are required for this process . VCP is thus likely to provide the energy required for extraction of CD4 from membranes . In line with this conclusion , depletion of ATP by incubation of cells in glucose-free medium supplemented with 2-deoxy-D-glucose and sodium azide [42] inhibited CD4 degradation induced by Vpu ( Fig . 2 , D and E ) . Co-precipitation experiments showed that the substrate-trapping VCP-QQ mutant was isolated as a complex with CD4 in the presence of wild-type Vpu but not Vpu-S52 , 56N ( Fig . 2F ) . Because these Vpu constructs differ in their ability to promote CD4 ubiquitination by the SCFβ-TrCP complex ( Fig . S2 ) , our findings are consistent with recruitment of VCP to ubiquitinated CD4 . The recruitment of VCP to CD4 is likely mediated by UFD1L and NPL4 . These proteins form a heterodimer that assembles with a VCP homohexamer [43] . Both UFD1L and NPL4 contain binding sites for each other and for VCP ( Fig . 2A ) [44] , [45] . In addition , UFD1L and NPL4 contain domains that bind K48-linked and K63-linked Ub chains , respectively ( Fig . 2A ) [39] , [46] . Overexpression of an UFD1L mutant lacking the binding site for VCP and NPL4 ( UFD1L-ΔUT6 ) had no effect on Vpu-mediated CD4 degradation ( Fig . 2 , G and H ) , probably because it cannot interfere with the ability of endogenous UFD1L to engage VCP and NPL4 for interaction with ubiquitinated CD4 . Overexpression of an UFD1L mutant lacking the Ub-binding domain ( UFD1L-ΔUT3 ) , on the other hand , inhibited Vpu-induced CD4 degradation ( Fig . 2 , G and H ) , whereas overexpression of an analogous NPL4 mutant ( NPL4-ΔZFD ) had no effect on this process ( Fig . 2 , I and J ) . The dominant-negative effect of UFD1L-ΔUT3 might be explained by the ability of this mutant to bind both VCP and NPL4 but not be recruited to proteins conjugated to K48-linked Ub chains . This is consistent with the specific involvement of K48 Ub linkages in ERAD [39] . We also observed that overexpression of a NPL4 mutant lacking the VCP-binding domain ( NPL4-ΔUBD ) failed to elicit a dominant-negative effect ( Fig . 2 , I and J ) . The inability of these NPL4 mutants to exert a dominant-negative effect stands in sharp contrast with the inhibitory effect of the NPL4 siRNAs on Vpu-induced CD4 degradation ( Fig . 1 , B-D ) . The explanation for this apparent discrepancy lay on the effects of each siRNA on the levels of the three components of the complex ( Fig . 2K ) . Indeed , siRNAs to VCP , UFD1L or NPL4 resulted in depletion of the corresponding target protein without affecting the levels of the other two , with the notable exception of the NPL4 siRNAs , which depleted both NPL4 and UFD1L ( Fig . 2K ) . This indicated that NPL4 is required to stabilize UFD1L , as previously reported [47] . Taken together , these findings ascribe specific functions to each of the components of the VCP-UFD1L-NPL4 complex in Vpu-mediated CD4 degradation: VCP energizes the process through ATP binding and hydrolysis , UFD1L binds ubiquitinated CD4 through recognition of K48 Ub chains , and NPL4 stabilizes UFD1L . We next wondered whether inhibition of Vpu-induced CD4 degradation by interference with the VCP-UFD1L-NPL4 complex allowed transport of newly-synthesized CD4 out of the ER . To assess this possibility , we performed siRNA-mediated depletion of NPL4 ( which , as described above , depleted both NPL4 and UFD1L ) , expressed CD4 in the absence or presence of Vpu , and examined the localization of CD4 by immunofluorescence microscopy ( Fig . 3 , A-L ) , sensitivity to Endo H ( Fig . 3 , P and Q ) , and FACS analysis ( Fig . 3 , R and S ) . Immunofluorescence microscopy showed that CD4 was predominantly at the plasma membrane in the absence of Vpu ( Fig . 3B ) . As expected , expression of Vpu , which localized to the ER as well as juxtanuclear structures corresponding to the trans-Golgi network and endosomes ( Fig . 3 , E and M–O; Fig . S3 , A–C ) , caused a marked loss of CD4 staining in mock-treated cells ( Fig . 3F ) . Interestingly , in NPL4-depleted cells , both Vpu ( Fig . 3I ) and the protected CD4 ( Fig . 3J ) co-localized on the ER , as evidenced by co-staining for the ER-resident protein , calnexin ( Fig . 3K ) . Normally , newly-synthesized CD4 receives two N-linked high-mannose oligosaccharide chains , only one of which acquires complex carbohydrates upon transport through the Golgi complex [16] . Treatment with Endo H removes one high-mannose chain , whereas treatment with PNGase F removes both chains , as detected by SDS-PAGE and immunoblot analysis of total CD4 ( Fig . 3P , top panel; Fig . 3Q ) . The small amount of CD4 that remained upon expression of Vpu in mock-treated cells was completely sensitive to Endo H , indicating that it was localized to the ER ( Fig . 3P , middle panel; Fig . 3Q ) . This phenotype was not due to a general impairment of protein maturation through the biosynthetic pathway because acquisition of Endo H-resistance by the transferrin receptor was not affected by expression of Vpu ( Fig . S3 , D and E ) . Significantly , in NPL4-depleted , Vpu-expressing cells , ∼75% of CD4 remained sensitive to Endo H ( Fig . 3P , lower panel; Fig . 3Q ) , consistent with localization of the majority of CD4 to the ER . Finally , FACS analysis showed that depletion of NPL4 per se did not alter expression of CD4 at the cell surface in the absence of Vpu ( Fig . 3 , R and S ) . However , in line with the experiments described above , Vpu drastically reduced ( ∼80% ) surface CD4 expression even in NPL4-depleted cells ( Fig . 3 , R and S ) . Altogether , these results demonstrated that a large fraction of CD4 remains in the ER in the presence of Vpu when targeting for degradation is blocked . Therefore , Vpu mediates CD4 retention in the ER in addition to degradation . To dissect the mechanism by which Vpu causes CD4 retention in the ER , we examined the effect of disrupting the Vpu-β-TrCP1/2 interaction by either mutating serines 52 and 56 to asparagine in Vpu [21] , [25] or depleting cells of both β-TrCP1 and β-TrCP2 [22] . A quantitative measure of ER retention was obtained by Endo H treatment in conjunction with immunoblotting ( Fig . 4 , A and B ) or pulse-chase analysis ( Fig . 4 , C and D ) . Using these assays , we found that ∼50% of CD4 was retained in the ER upon disruption of the Vpu-β-TrCP1/2 interaction by either method . Consistent with these observations , FACS analysis showed that expression of Vpu-S52 , 56N , depletion of β-TrCP1/2 , or a combination of both , resulted in ∼50% reduction of CD4 surface levels ( Fig . 4 , E and F ) . Comparison to the retention observed upon NPL4 depletion ( ∼75% ) ( Fig . 3 , P–S ) indicated that binding of β-TrCP1/2 to Vpu contributes to CD4 retention in the ER , but is probably not the only factor . Binding of β-TrCP1/2 to Vpu could contribute to CD4 retention in the ER directly through assembly of a transport-incompetent complex or indirectly through ubiquitination mediated by the SCFβ-TrCP complex . To examine the role of ubiquitination in ER retention and degradation , we mutated all potential Ub-acceptor lysine residues to arginine ( CD4-K-less ) or lysine , serine and threonine residues to arginine , alanine and isoleucine , respectively ( CD4-KST-less ) , in the cytosolic tail of CD4 ( Fig . 5A ) . None of these mutations affected the ability of CD4 to interact with Vpu-S52 , 56N , as determined by co-precipitation analysis ( Fig . 5B ) . Ubiquitination was assessed by co-expression of untagged CD4 constructs with FLAG-tagged Ub in the absence or presence of Vpu . Cell extracts were fully denatured prior to immunoprecipitation with a conformation-independent antibody to a luminal CD4 epitope and immunoblotting with an antibody to the FLAG epitope ( Fig . 5C ) . This protocol ensured that ubiquitinated species corresponded to CD4 and not to associated proteins . The amounts of ubiquitinated CD4 were normalized to the amounts of remaining CD4 in each sample ( Fig . 5 , D and E ) . We observed that expression of Vpu enhanced ubiquitination of CD4 ∼39-fold ( Fig . 5 , C and E ) . Mutation of all four cytosolic lysines decreased but did not completely abolish CD4 ubiquitination ( Fig . 5 , C and E ) and only slightly diminished degradation ( Fig . 5 , C and D ) induced by Vpu . Additional mutation of the cytosolic serine and threonine residues abrogated virtually all ubiquitination ( Fig . 5 , C and E ) and degradation ( Fig . 5 , C and D ) induced by Vpu . These results indicated that targeting of CD4 for degradation depends on ubiquitination of lysine and serine/threonine residues in the cytosolic tail . Endo H digestion analysis showed that the CD4-K-less and CD4-KST-less mutants were normally exported from the ER in the absence of Vpu ( Fig . 5 , F and G ) . Expression of Vpu , however , resulted in retention of ∼80% of the CD4-K-less mutant in the ER ( Fig . 5 , F and G ) , similarly to wild-type CD4 ( Fig . 3 , P and Q ) . Interestingly , the CD4-KST-less mutant was retained ∼50% in the ER ( Fig . 5 , F and G ) . This level of retention is similar to that observed upon disruption of the Vpu-β-TrCP1/2 interaction ( Fig . 4 , A and B ) , indicating that the contribution of this interaction to ER retention is likely indirect , through SCFβ-TrCP–mediated ubiquitination of the CD4 cytosolic tail . The fact that Vpu still retains ∼50% of CD4 in the ER independently of the interaction of Vpu with β-TrCP1/2 or the ubiquitination of the CD4 cytosolic tail suggests an additional role for the CD4-Vpu interaction in ER retention of CD4 . This interaction has been previously shown to depend on the cytosolic domains of both CD4 and Vpu [20] . To test whether this particular interaction accounted for the bulk of ER retention , we examined the effect of deleting most of the cytosolic tail from CD4 ( residues 426–458 ) ( CD4-Δcyto construct; Fig . 5A ) . This truncation should eliminate interaction with Vpu and , consequently , all retention and degradation determinants . As expected , CD4-Δcyto was not degraded in the presence of Vpu ( Fig . 6 , A and B ) [20] . Pulse-chase experiments combined with Endo H digestion showed that CD4-Δcyto completely exited the ER , albeit more slowly than full-length CD4 , in the absence of Vpu ( Fig . 6 , C and D ) . These results indicated that the CD4 cytosolic tail has only a weak ER export signal . Surprisingly , expression of Vpu resulted in ∼80% retention of CD4-Δcyto in the ER ( Fig . 6 , E and F ) . Similar results were obtained with a CD4 construct having a more radical deletion of the cytosolic domain ( residues 422–458 ) ( Fig . 5A ) ( data not shown ) . Therefore , interactions other than those between the cytosolic domains must be the main determinants of Vpu-mediated retention of CD4 in the ER . Since Vpu has a very short and variable luminal domain ( 3–12 non-conserved amino acids; 4 in the HIV-1 NL4-3 variant used in this study ) , direct or indirect interactions at the level of the TMDs must be the main determinant of the CD4 retention in the ER mediated by Vpu . The fact that ER retention of CD4-Δcyto by Vpu-S52 , 56N was greater ( ∼75% ) ( Fig . 6 , E and F ) than that of full-length CD4 ( ∼50% ) ( Fig . 4 , A and B ) further supports the occurrence of a weak ER export signal in the CD4 cytosolic tail . These effects at the level of the ER were reflected in changes in surface expression of CD4 analyzed by FACS ( Fig . 6 , G and H ) . Thus , Vpu has an intrinsic ability to retain CD4 in the ER that is independent of the cytosolic domain of CD4 and is likely mediated by TMD interactions . To assess more directly the importance of the Vpu TMD , we replaced it with that of the G protein of vesicular stomatitis virus ( VSV-G ) , resulting in a chimera designated Vpu-VSV-G-TMD ( Fig . 7A ) . Co-precipitation experiments showed that Vpu-VSV-G-TMD interacted with β-TrCP1 ( Fig . 7B ) and CD4 ( Fig . 7C ) to the same extent as wild-type Vpu . However , this chimera largely failed to promote CD4 degradation ( Fig . 7 , C and D ) . This observation is in line with previous reports that substitutions of the Vpu or CD4 TMDs abrogate Vpu-induced CD4 degradation [48] , [49] . The TMDs must therefore play a specific role in ERAD , perhaps in allowing dislocation of CD4 from the ER membrane . Immunofluorescence microscopy ( Fig . 7 , E–P ) and Endo H digestion analysis ( Fig . 7 , Q and R ) showed that Vpu-VSV-G-TMD also failed to prevent exit of the bulk of CD4 from the ER towards the plasma membrane . Only a minor fraction of CD4 ( ∼25% ) remained in the ER in the presence of Vpu-VSV-G-TMD ( Fig . 7 , Q and R ) , presumably because of cytosolic domain interactions with Vpu/β-TrCP1/2 and ensuing ubiquitination ( Figs . 4 and 5 ) . Taken together , these results indicate that TMD interactions play important roles in Vpu-dependent ER retention and degradation of CD4 . Unlike CD4 , VSV-G has a strong ER export signal in its cytosolic tail [50] . Substitution of the VSV-G cytosolic tail for the CD4 cytosolic tail resulted in a chimeric protein ( CD4-VSV-G-cyto mutant; Fig . 8A ) that was not degraded in the presence of Vpu ( Fig . 8 , B and C ) , presumably because the VSV-G cytosolic tail does not interact with Vpu and lacks degradation determinants . Endo H digestion showed that , in contrast to CD4-Δcyto , this chimera was efficiently transported out of the ER both in the absence or presence of Vpu ( Fig . 8 , D and E ) . In addition , FACS analysis showed that this chimera was expressed at the cell surface very efficiently irrespective of the presence or absence of Vpu ( Fig . 8 , F and G ) . Therefore , the VSV-G export signal is capable of overriding ER retention of CD4 mediated by Vpu .
The results of our study shed light on two aspects of Vpu function that remained poorly understood: its ability to retain CD4 in the ER and to target CD4 to the ERAD pathway . The scheme shown in Fig . 8H represents our view of how these aspects are integrated . We think that Vpu first acts to retain CD4 in the ER ( Fig . 3 ) by virtue of TMD interactions ( Figs . 6 and 7 ) . The cytosolic domain of Vpu then interacts with CD4 and recruits the SCFβ-TrCP Ub ligase complex ( Fig . 4 ) [21] , [22] , which mediates the addition of multiple Ub moieties to lysine [23] , [24] and serine/threonine residues ( Fig . 5 ) in the cytosolic tail of CD4 . Ubiquitination further contributes to CD4 retention in the ER , and additionally marks CD4 for delivery to proteasomes ( Fig . 5 ) . This delivery involves recruitment of the VCP-UFD1L-NPL4 complex through recognition by UFD1L of K48-linked poly-Ub chains on the CD4 cytosolic tail ( Figs . 1 and 2 ) . The ATPase activity of VCP then drives dislocation of CD4 from the ER membrane into the cytosol ( Fig . 2 ) [24] for eventual degradation in proteasomes [17] . The multiple levels at which Vpu acts to prevent export of CD4 from the ER underscore the importance of ensuring complete suppression of CD4 for progression of the infection . Our experiments show that a large fraction ( ∼50-80% ) of CD4 remains in the ER in the presence of Vpu when ERAD is blocked by inactivation of the VCP-UFD1L-NPL4 complex ( Fig . 3 ) , disruption of the Vpu-β-TrCP1/2 interaction ( Fig . 4 ) , or mutation of lysine , serine and threonine residues in the cytosolic tail of CD4 ( Fig . 5 ) , in all cases in the absence of Env or any other inhibitor of ER export . Furthermore , this retention is independent of the only interaction between Vpu and CD4 reported to date , which involves the cytosolic domains of both proteins [20] . Indeed , deletion of the cytosolic tail of CD4 does not abrogate but rather enhances ER retention of CD4 upon expression of either wild-type Vpu or Vpu-S52 , 56N ( Fig . 6 ) . Since Vpu and cytosolic tail-less CD4 only overlap at their TMDs , direct or indirect interactions at this level must be the main determinant of ER retention . Indeed , a Vpu mutant containing a heterologous TMD failed to retain CD4 in the ER ( Fig . 7 ) , further supporting a role of TMD interactions in ER retention . How could interaction with Vpu result in CD4 retention in the ER ? We envisaged that assembly with Vpu could either mask an ER export signal on CD4 , or confer on CD4 an intrinsic tendency of Vpu to reside in the ER . Our analysis ruled the first possibility unlikely , as CD4 has only a weak ER export signal in its cytosolic tail ( Fig . 6 ) . Moreover , even this weak signal can wrest a small amount of CD4 away from Vpu and out of the ER , as demonstrated by the higher ER retention of cytosolic tail-less CD4 relative to full-length CD4 in the presence of the Vpu-S52 , 56N mutant ( when neither CD4 construct is degraded ) ( Fig . 6 ) . Finally , replacement of the CD4 cytosolic tail by that of the VSV-G protein , which has a strong ER export signal [50] , completely prevented ER retention of CD4 by Vpu ( Fig . 8 ) . Therefore , we favor the second explanation of retention in trans , according to which Vpu itself has ER retention information that is transmitted to CD4 upon assembly . Indeed , our immunofluorescence and immunoelectron microscopy analysis show that a large fraction of Vpu localizes to the ER ( Figs . 3 and S3 ) , irrespective of the presence or absence of CD4 ( data not shown ) . CD4 may thus be detained in the ER by forming a transport-incompetent complex with Vpu . The conveyance of ER retention information from Vpu to CD4 , however , does not account for the full extent of CD4 retention in the ER . SCFβ-TrCP-mediated ubiquitination on lysine and serine/threonine residues in the CD4 cytosolic tail provides an additional contribution to ER retention , as evidenced by the higher degree of retention elicited by NPL4 depletion ( Fig . 3 ) relative to disruption of the Vpu-β-TrCP1/2 interaction ( Fig . 4 ) or mutation of Ub acceptor residues in the CD4 cytosolic tail ( Fig . 5 ) . Recently , ubiquitination of a palmitoylation-deficient mutant of the lipoprotein receptor-related protein 6 ( LRP6 ) was also shown to mediate ER retention of this protein [51] . The exact mechanism by which ubiquitination can impart ER retention of mutant LRP6 or CD4 in the presence of Vpu remains to be elucidated . Therefore , Vpu retains CD4 in the ER , independently of targeting to degradation , by the additive effects of two distinct mechanisms: TMD-mediated conferral of ER residency ( accounting for ∼50% of the ER retention ) and ubiquitination of the CD4 cytosolic tail ( adding another ∼25% to ER retention ) . In addition to retaining CD4 in the ER , Vpu targets newly-synthesized CD4 for proteasomal degradation [5] , [16] , [17] . This targeting has been proposed to be fundamentally distinct from ERAD , mainly based on a study using yeast as a heterologous expression system [23] . This study showed that the typical yeast ERAD components , Hrd1p , Hrd3p and Ubc7p , are dispensable for Vpu-induced CD4 degradation [23] , a fact that our results have confirmed for the human orthologs , HRD1 , SEL1L and UBC7 , in human cells ( Fig . 1 ) . The participation of a non-ERAD Ub ligase , SCFβ-TrCP , in Vpu-mediated CD4 degradation [21] , [22] has further emphasized the distinctive nature of this process . However , we find that the VCP-UFD1L-NPL4 complex , a key component of the ERAD machinery , is involved in CD4 degradation by Vpu ( Figs . 1 and 2 ) . As previously shown for other ERAD substrates [37] , [38] , [39] , [40] , our results are consistent with VCP-UFD1L-NPL4 mediating extraction of CD4 from the ER membrane ( Fig . 1 ) . Vpu thus appears to bypass the early stages of ERAD , including substrate recognition and ubiquitination by ERAD machinery components , but joins in the later stages , beginning with dislocation by the VCP-UFD1L-NPL4 complex . We speculate that other components that act downstream of VCP-UFD1L-NPL4 in typical ERAD [28] might also participate in Vpu-induced CD4 degradation . Over the past few years , it has become increasingly clear that there is not a single ERAD pathway but several alternative routes for targeting cellular proteins with defects in luminal ( ERAD-L ) , membrane ( ERAD-M ) , and cytosolic domains ( ERAD-C ) [28] . Similarly , various herpesviruses downregulate class I molecules of the major histocompatibility complex ( MHC-I ) from the ER by engaging the ERAD machinery at different levels . For example , the human cytomegalovirus ( HCMV ) immunoevasin , US11 , establishes TMD interactions with Derlin-1 that are required for the ability of this viral protein to downregulate MHC-I [52] . In contrast , another HCMV immunoevasin , US2 , downregulates MHC-I by a mechanism that is Derlin-1-independent [52] but involves cytosolic domain interactions with signal peptide peptidase [53] . The mK3 immunoevasin encoded by murine γ-herpesvirus 68 ( MHV-68 ) exemplifies yet another variation on the mechanisms used by herpesviruses to downregulate MHC-I [54] . Unlike US11 and US2 , mK3 has intrinsic Ub ligase activity that ubiquitinates newly-synthesized MHC-I , leading to its disposal by the proteasome [55] . Importantly , all of these mechanisms eventually merge at the level of the VCP-UFD1L-NPL4 complex for delivery to the proteasome . In this context , Vpu-induced CD4 downregulation represents another adaptation to the use of the same fundamental pathway for the proteasomal degradation of ER-retained proteins . The activities of Vpu are not limited to the ER , but extend to other cellular compartments [10] . Indeed , Vpu has recently been shown to downregulate the restriction factor BST-2/tetherin from the cell surface [56] , [57] . In light of these findings , it will be of interest to determine whether Vpu is capable of directly removing pre-existing CD4 from the cell surface in addition to targeting newly-synthesized CD4 to the ERAD pathway . Mutation of all four lysines in the cytosolic tail of CD4 only partially inhibited Vpu-induced CD4 ubiquitination and degradation ( Fig . 5 ) . This led us to hypothesize that cytosolic tail residues other than lysine could be additional targets for ubiquitination . Previous studies showed that downregulation of MHC-I by mK3 involved ubiquitination of not only lysine , but also serine and threonine residues in the MHC-I cytosolic tail [55] . We tested whether this was also the case for Vpu-induced CD4 downregulation and found that more profound inhibition of CD4 ubiquitination and degradation could be achieved by mutation of all lysine , serine and threonine residues in the CD4 cytosolic tail ( Fig . 5 ) . Moreover , serine and threonine residues in the cytosolic tail of CD4 contributed to its retention in the ER mediated by Vpu ( Fig . 5 ) . These observations are highly significant because Vpu , unlike mK3 , does not have intrinsic Ub ligase activity , indicating that serine/threonine residues could be ubiquitinated by a cellular enzyme . It will now be of interest to investigate whether this modification is mediated by SCFβ-TrCP or another Ub ligase , and whether other ERAD substrates undergo a similar modification .
pcDNA3 . 1-FLAG-Ub was kindly provided by S . Ishikura ( NICHD , NIH ) . pFLAG-CMV2-human β-TrCP1 was obtained from Y . Ben-Neriah ( Hadassah Medical School , Hebrew University ) . pNLA-1 is a derivative of pNL4-3 [58] , lacking the gag and pol genes but expressing all other viral genes . pCMV-human CD4 [20] and pcDNA3 . 1-codon-optimized Vpu ( pcDNA3 . 1-Vphu ) [41] were used as templates for site-directed mutagenesis using a QuikChange II kit ( Stratagene , Cedar Creek , TX ) . pcDNA3 . 1 plasmids encoding RGS-His-tagged mouse wild-type VCP , VCP-ΔN , VCP-AA and VCP-QQ were previously reported [39] , [59] . Mouse wild-type UFD1L , UFD1L-ΔUT3 ( residues 215-307 ) and UFD1L-ΔUT6 ( residues 1-214 ) cDNAs were amplified by PCR using pcDNA3 . 1-FLAG-mouse UFD1L [60] as template and cloned as XhoI/BamHI fragments into the pcDNA3 . 1/myc-His A vector ( Invitrogen , Carlsbad , CA ) . To produce FLAG-tagged rat wild-type NPL4 , NPL4-ΔUBD ( residues 96-608 ) and NPL4-ΔZFD ( residues 1-579 ) , cDNAs were PCR-amplified from pFLAG-His-rat NPL4 [60] and cloned into the pFLAG-CMV-6c vector ( Sigma-Aldrich , Saint Louis , MO ) as EcoRI/BglII fragments . C-terminal FLAG-One-STrEP-tagged CD4 was cloned as an EcoRI/BamHI fragment into the pcDNA3 . 1/myc-His A vector . pCMV-human CD4 and pEGFP-N1-VSV-G [61] were used as templates in a sewing PCR strategy to generate the CD4-VSV-G-cyto chimeric cDNA ( human CD4: bp 1–1260; VSV-G: bp 1447–1536 ) , which was cloned as an EcoRI/NotI fragment into the pCI-Neo vector ( Promega , Madison , WI ) . A sewing PCR approach was also used to create a cDNA encoding Vpu-VSV-G-TMD ( Vpu: bp 1–12; VSV-G: bp 1393-1461; Vpu: 82–246 ) , which was cloned into the pcDNA3 . 1/myc-His A vector as a EcoRI/XhoI fragment . All mutagenesis and cloning products were verified by DNA sequencing . The following mouse monoclonal antibodies were used in this study: 4B12 ( Leica Microsystems , Bannockburn , IL ) , OKT4 ( eBiosciences , San Diego , CA ) or unconjugated and allophycocyanin ( APC ) -conjugated S3 . 5 ( Caltag Laboratories , Burlingame , CA ) to human CD4; Ab-5 to actin , clone 37 to calnexin and clone 19 to UFD1L ( BD Biosciences , San Jose , CA ) ; 58 . 13 . 3 to VCP ( RDI Research Diagnostics , Concord , MA ) ; H68 . 4 to human transferrin receptor ( Zymed , San Francisco , CA ) ; M2 to FLAG ( Sigma-Aldrich ) . A mouse polyclonal antibody to NPL4 was obtained from Novus Biologicals ( Littleton , CO ) . Rabbit polyclonal antibodies to FLAG and myc were from Sigma-Aldrich and Cell Signaling ( Danvers , MA ) , respectively . Rabbit polyclonal antibodies to the human CD4 cytosolic tail ( residues 420 to 447 ) [17] and Vpu ( residues 32 to 81 ) [62] were previously described . Alexa Fluor 488-conjugated donkey anti-rabbit IgG ( H+L ) was from Molecular Probes ( Eugene , OR ) . Alexa Fluor 594- or 647-conjugated donkey anti-mouse IgG2a or IgG1 , respectively , were from Invitrogen . HRP-conjugated donkey anti-mouse IgG and donkey anti-rabbit IgG were from Amersham Biosciences ( Piscataway , NJ ) . HeLa cells ( American Type Culture Collection , Manassas , VA ) were transiently transfected by using Lipofectamine 2000 ( Invitrogen ) . Plasmids encoding human CD4 ( pCMV-CD4 ) and codon-optimized Vpu ( pcDNA3 . 1-Vphu ) were transfected at a 1∶1 ratio . Vpu levels driven from this codon-optimized construct were ∼3-fold higher than those driven from the proviral pNL4-3 construct ( Fig . S4 ) . However , similar levels of CD4 downregulation were attained with amounts of codon-optimized Vpu that were up to 16-fold lower ( Fig . S4 ) . ON-TARGETplus SMARTpool siRNAs and siControl duplex siRNA ( Dharmacon , Lafayette , CO ) at a final concentration of 100 nM were used to knockdown expression of endogenous ERAD targets ( including β-TrCP1/2 ) and GAPDH , respectively ( Table S1 ) . Silencing was achieved by double transfection of 0 . 5×105 HeLa cells with Oligofectamine ( Invitrogen ) . Cells were analyzed 48 h after the second round of transfection . Cells were lysed in ice-cold lysis buffer ( 0 . 5% Triton X-100 , 50 mM Tris-HCl pH 7 . 5 , 150 mM NaCl , 5 mM EDTA ) supplemented with the complete Mini protease inhibitor cocktail ( Roche Diagnostics , Indianapolis , IN ) . Equivalent amounts of cell lysates were subjected to immunoprecipitation as described [63] . Immunoprecipitated proteins or whole cell lysates ( 10 µg ) were subjected to SDS-PAGE using the NuPAGE Bis-Tris Gel system ( Invitrogen ) and immunoblotting as previously described [64] . Membrane bound horseradish peroxidase ( HRP ) -conjugated antibodies were detected using the SuperSignal West Pico Chemiluminescent Substrate from Thermo Scientific ( Rockford , IL ) . Data analysis was performed using the Image J software . Cells grown in 6-well plates were incubated for 30 min at 37°C in methionine- and cysteine-free DMEM ( Invitrogen ) , pulse-labeled with 0 . 2 mCi/ml [35S]methionine-cysteine ( Express Protein Label; Perkin Elmer , Boston , MA ) for 2 min at 37°C , and chased in complete medium ( DMEM supplemented with 10% fetal bovine serum ) supplemented with 5 mM L-methionine and L-cysteine ( Sigma-Aldrich ) . Cellular ATP levels were depleted during the chase period using 20 mM 2-deoxy-D-glucose and 10 mM sodium azide ( Sigma-Aldrich ) in glucose-free DMEM ( Invitrogen ) . At each chase time , cells were extracted by incubation in ice-cold lysis buffer ( 0 . 5% Triton X-100 , 50 mM Tris-HCl pH 7 . 5 , 300 mM NaCl , 5 mM EDTA ) supplemented with a protease inhibitor cocktail , and lysates were then subjected to immunoprecipitation as described [63] . Immunoprecipitated proteins were analyzed by SDS-PAGE and visualized by fluorography on a Typhoon 9200 PhosphorImager ( Amersham Biosciences ) . Data analysis and quantification was performed using the ImageQuant software . Cell fractionation and membrane protein extraction with Na2CO3 were performed as previously described [65] . Indirect immunofluorescence staining of fixed , permeabilized cells was performed as described [64] . Cells were examined with an Olympus FluoView FV1000 laser scanning confocal unit attached to an Olympus IX81 motorized inverted microscope as previously described [66] . Immunoelectron microscopy was performed as reported [14] . Immunoprecipitation of CD4 under denaturing conditions was performed as described [63] . Briefly , HeLa cells expressing FLAG- or myc-tagged Ub were lysed in a denaturing lysis buffer ( 1% SDS , 50 mM Tris-HCl pH 7 . 4 , 5 mM EDTA , 10 mM dithiothreitol , 15 U/ml DNase I , 10 mM α-iodoacetamide , 5 mM N-ethylmaleimide ) supplemented with the complete Mini protease inhibitor cocktail . After heating samples for 10 min at 100°C , the suspensions were diluted 10-fold in a non-denaturing lysis buffer ( 1% Triton X-100 , 50 mM Tris-HCl pH 7 . 5 , 150 mM NaCl , 5 mM EDTA , 10 mM α-iodoacetamide , 5 mM N-ethylmaleimide ) supplemented with the complete Mini protease inhibitor cocktail . CD4 was immunoprecipitated using a conformation-independent anti-CD4 antibody . Alternatively , FLAG-One-STrEP-tagged CD4 was pulled-down using Strep-Tactin sepharose ( IBA , Göttingen ? , Germany ) . Ubiquitinated CD4 was then analyzed by immunoblotting with antibodies to the FLAG or myc epitopes , respectively . Cell lysates made in 1% SDS , 0 . 1 M Tris-HCl pH 8 were processed for Endo H and PNGase F digestion as previously reported [67] . Non-permeabilized cells were stained with an APC-conjugated antibody to the human CD4 ectodomain and prepared for FACS analysis as described [13] . The HIV-1 Vpu and human CD4 clones used in this study correspond to Swiss-Prot entries P05923 and P01730 , respectively . | HIV-1 devotes two accessory proteins , Nef and Vpu , to the task of removing the viral receptor , CD4 , from the cell surface . Whereas Nef delivers surface CD4 for degradation in lysosomes , Vpu targets newly-made CD4 in the endoplasmic reticulum for degradation by cytosolic proteasomes . This latter process was thought to be fundamentally distinct from that used for the disposal of abnormal cellular proteins from the endoplasmic reticulum . Contrary to this notion , however , we show that Vpu utilizes at least part of the endoplasmic reticulum-associated degradation machinery to dispose of CD4 . Disabling this machinery prevents CD4 degradation induced by Vpu but , surprisingly , does not allow transport of CD4 to the cell surface . This is due to a second function of Vpu: retention of CD4 in the endoplasmic reticulum . These two functions of Vpu are mediated by different parts of the Vpu molecule and involve distinct mechanisms . This functional redundancy underscores the importance of suppressing CD4 expression for HIV-1 to thrive in the infected cells . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"cell",
"biology/membranes",
"and",
"sorting",
"infectious",
"diseases/hiv",
"infection",
"and",
"aids",
"virology/virulence",
"factors",
"and",
"mechanisms"
] | 2010 | Multilayered Mechanism of CD4 Downregulation by HIV-1 Vpu Involving Distinct ER Retention and ERAD Targeting Steps |
Vertebrate genes often play functionally distinct roles in different subsets of cells; however , tools to study the cell-specific function of gene products are poorly developed . Therefore , we have established a novel mouse model that enables the visualization and manipulation of defined subpopulations of neurons . To demonstrate the power of our system , we dissected genetic cascades in which Pax6 is central to control tangentially migrating neurons of the mouse brainstem . Several Pax6 downstream genes were identified and their function was analyzed by over-expression and knock-down experiments . One of these , Pou4f2 , induces a prolonged midline arrest of growth cones to influence the proportion of ipsilaterally versus contralaterally settling neurons . These results demonstrate that our approach serves as a versatile tool to study the function of genes involved in cell migration , axonal pathfinding , and patterning processes . Our model will also serve as a general tool to specifically over-express any gene in a defined subpopulation of neurons and should easily be adapted to a wide range of applications .
Understanding cell-specific regulatory mechanisms is a major challenge in the post-genome era . Particularly in mammals , the reiterated usage of the same transcription factor in distinct subsets of cells or during distinct developmental time points provides the basis to generate thousands of individual cell types with a relatively small number of genes . A single transcription factor may therefore elicit variable downstream effects depending on the context of its expression . Tissue-specific knockout strategies based e . g . on the Cre-lox-system , or promoter-driven transgenic models allow a cell-specific manipulation of genes . However , as these techniques rely on the generation of new transgenic animals for each gene-combination analyzed they are laborious and time-consuming . Here , we combined a transgenic model with tissue-specific transfection protocols and organotypic cultures to enable the quick analysis of numerous genes in a cell-specific manner . As a proof of principle we applied our system to decode molecular pathways initiated by the transcription factor Pax6 which is involved in neuronal cell migration and axonal pathfinding processes . Pax6 , a homeodomain and paired domain containing transcription factor , is a major determinant of visual and olfactory sensory structures and is essential for a variety of patterning and pathfinding processes throughout the nervous system [1]–[3] . Depending on the context and area of expression Pax6 initiates varying downstream effects . Homozygous small eye ( Pax6Sey/Sey ) mouse and rat embryos , which lack functional Pax6 , do neither generate eye nor nasal structures and are deficient in ventral diencephalic structures [4]–[9] . In the ventral hindbrain and spinal cord , Pax6 controls the dorso-ventral patterning of motorneurons and of interneurons [5] , [10] . In the cerebral cortex Pax6 determines the neurogenic potential of radial glial cells [11] , [12] . Throughout the developing nervous system , with the exception of the midbrain , Pax6 is expressed in a ventral and a dorsal pool of progenitor cells . Although the dorsal Pax6 expression domain has achieved much less attention than the ventral domain there is evidence that Pax6 plays a pivotal role in the specification and migration of neurons derived from this domain [13]–[16] . The dorsal domain of Pax6 positive neuronal precursors of the hindbrain includes the rhombic lip ( RL ) [14] , [16] which comprises the interface between the dorsal neuroepithelium and the roof plate . The RL is the source of several tangentially migrating neurons ( see also Figure 1A ) [14] , [17]–[24] . The most notable are the neurons of the marginal migratory stream ( mms; also pes ) which migrate from the rhombic lip circumferentially around the medulla towards their contralateral destinations to settle in the ECN ( external cuneate nuclei ) and the LRN ( lateral reticular nuclei ) [17] , [25] . Owing to the superficial nature of the mms migration these neurons serve as paradigm to study neuronal migration and axonal pathfinding processes . The highly complex neuronal circuits of the vertebrate nervous system are established during development when growing axons travel considerable distances towards their targets to generate the appropriate connections . This wiring process depends on attractive and repulsive factors which emanate from final or intermediate cellular targets and which are interpreted by cell surface receptors located on axonal growth cones [26] , [27] . Although the general principles were uncovered during the past years our understanding of axonal pathfinding processes is far from being complete . Current methods to analyze candidates regulating neuronal migration and axonal navigation processes are laborious and often involve the generation of transgenic animals for each gene analyzed . Non-transgenic methods , as DiI labeling of neurons or vector-driven mis-expression of gene constructs , are suitable for use with certain applications , however , they are neither cell specific nor can they be targeted to distinct neuronal subpopulations . Here we describe a novel transgenic mouse model , which allows the specific and exclusive visualization and manipulation of subsets of neurons in the developing brain . To demonstrate the power of this system we have analyzed the role of Pax6 in migrating neurons of the brainstem . In Pax6 mutant mice migration of these neurons is distorted and some neurons differentiate at ectopic positions . Using transplantation , knock-down and over-expression experiments we show that distinct migratory features are controlled by discrete sets of Pax6 downstream genes . These results demonstrate the potential of our transgenic mouse model as a tool to study the role of Pax6 in individual neurons . Moreover , our system should be widely applicable to study virtually any gene that acts during cell determination , axonal pathfinding and/ or cell migration processes .
The functional analysis of genes in restricted tissues often involves the generation of inducible knockout mice or mice over expressing transgenic constructs . To simplify this time-consuming process we developed an in vitro model that enables the visualization and manipulation of defined populations of neurons . To label neurons in a largely unlabelled background we searched for genes that were expressed in only a subset of neuronal precursors and in migrating neurons . Pax6 meets these criteria ideally . Pax6 is expressed in several groups of tangentially migrating neurons and their precursors as well as in a small population of radially migrating neurons and their precursors ( Figure 1A–1C ) [5] , [10] , [12] , [14] , [16] , [28] . We adopted the Tet binary system [29] and generated YAC ( yeast artificial chromosome ) transgenic mice which expressed the tetracycline dependent transactivator ( tTA ) in all Pax6 positive cells . A 420 kb YAC spanning the human PAX6 locus ( Y593 ) [30] was modified such that the PAX6 coding region was replaced with a cassette containing an IRES ( internal ribosomal entry site ) and the tTA ( Figure 1D ) . Previously , we and others had shown that the unmodified YAC Y593 contains all elements driving full functional PAX6 expression [30]–[32] and , in agreement with this , Tg ( PAX6-tTA ) mice showed a wide overlap of tTA and endogenous murine Pax6 expression ( Figure S1 ) . Tg ( PAX6-tTA ) mice were entirely normal and control experiments insured that neuronal patterning and migration was unaltered . tTA is a transcriptional activator that at moderate levels of expression is completely inert in vertebrates , yet , enables the activation of artificial constructs containing a tTA-DNA-binding element ( TRE = tetracyline responsive element ) . To examine whether our transgenic model specifically allows the labeling of only Pax6 positive cells we introduced by electroporation reporter constructs driving the green fluorescent protein into transgenic embryos . In all instances only Pax6 positive cells , e . g . retinal precursor cells , cortical precursors , or cerebellar granule cells , expressed the reporter genes ( Figure S1 ) . Non-transgenic embryos or Pax6 negative tissues did not induce reporter gene expression ( Figure 1H , Figure S1 ) . Together these results demonstrate that Tg ( PAX6-tTA ) mice enable the targeting of reporter gene constructs specifically to Pax6 positive cells and tissues during development . As Tg ( PAX6-tTA ) mice allow any gene to be targeted to Pax6 expressing cells , they are of potential value to study neuronal migration and axonal pathfinding processes and for the analysis of Pax6 downstream effects . As a proof of principle , we chose to focus on the marginal migration stream ( mms ) . Like other tangentially migrating neurons , mms neurons use the same or similar navigational cues as do growing axons , and migration of mms neurons is severely disturbed in Pax6 mutant Pax6Sey/Sey mice ( see below ) . Neurons of the mms are generated at the rhombic lip and migrate circumferentially around the embryonic brainstem to generate the contralateral lateral reticular ( LRN ) and the external cuneate ( ECN ) nuclei ( Figure 1A ) [16] , [21]–[23] , [25] . Migration starts at E13 . 0 and is completed by E16 . 5 . Pax6 is expressed in precursors at the rhombic lip , in all migrating neurons of the mms and during the initial period of settling in the target nuclei ( Figure 1B , 1C and data not shown ) . Antibody staining and in situ hybridization ( not shown ) of Tg ( PAX6-tTA ) mice confirmed a complete overlap of Pax6 and tTA expression in these neurons ( Figure 1I ) . To visualize migrating mms neurons in Tg ( PAX6-tTA ) mice , reporter constructs were introduced into neuronal precursor cells in the left rhombic lip by electroporation at E12 . 5 before migration had begun ( Figure 1E ) . Whole brainstems including the cerebellar primordium were then sustained in organotypic filter cultures for up to 14 days as an open book preparation which allowed the observation of migrating neurons with a fluorescence microscope from above ( Figure 1F , 1G ) . Our approach to use a binary system ensured that only Pax6 positive neurons containing tTA and a TRE reporter construct expressed the desired reporter genes . This procedure resulted in the specific labeling of mms neurons originating only from one rhombic lip . Pax6 positive neurons originating from the opposite rhombic lip remained unlabelled as were Pax6 negative ( and therefore tTA negative ) neurons originating from regions close to the rhombic lip . Unlabelled neurons included neurons of the submarginal migration stream ( sms ) which generate the inferior olive ( IO ) thus demonstrating the specificity of our model . To allow the simultaneous visualization and manipulation of neurons we designed reporter constructs containing two TRE elements ( Figure 1J ) . Control constructs co-expressed a cytoplasmic green fluorescent protein ( EGFPm ) and a nuclear red fluorescent protein ( DsRed2nls ) in 99% ( ±1; n = 10 ) of labeled neurons demonstrating that our reporter constructs enable the co-expression of two genes in the same neurons ( Figure 1J ) . To enable statistical analysis of the cultures , the territories of the LRN and the ECN were delineated using visible landmarks ( Figure 1K , 1K′ , 1L; see also the Materials and Methods section ) . Immunolabeling of cultures expressing a HA-tagged Pax6 construct demonstrate that over-expression of TRE constructs in Tg ( PAX6-tTA ) transgenic cultures result in moderate levels of protein expression that are in the range of physiological Pax6 concentrations ( Figure 1M , 1M′ , 1M'' ) . Pax6 mutant Pax6Sey/Sey mice display multiple neuronal patterning and migration defects . We therefore wished to determine whether Pax6 also regulates the mms . At the anatomical level , several features of the mms are severely disturbed in Pax6 mutant Pax6Sey/Sey embryos . Most noticeable , the initiation of migration and the midline crossing was delayed by 0 . 5 days ( asterisks in Figure 2A , 2A′ and data not shown; see also Figure S2 and Figure 4A , 4A′ The expression patterns of Pax2 , Dcx , NK1R , and DopH was unaltered indicating that there is no general developmental delay in the mutant brainstem ( data not shown ) . In Pax6Sey/Sey embryos some migrating mms neurons used a sub-marginal instead of a marginal migration path ( black arrowhead in Figure 2A′; see also Figure S2 and Figure 4A′ ) and at E14 . 5 a large number of mutant neurons accumulated around the midline suggesting a reduced pace in midline crossing ( black arrow in Figure 2B , 2B′ ) . Furthermore , a subset of Pax6 positive neurons migrated along the midline into the parenchyma of the hindbrain ( white arrowheads in Figure 2C , 2C′ ) . We used a variety of markers , e . g . antibodies against the potassium channel Kcnj6 , and DiI tracing of mossy fiber projections to discriminate between mms neurons ( generating the ECN and LRN ) and neurons of the sms ( generating the IO ) . These experiments all indicated a complete loss of neurons in the ECN and a disorganized settling of neurons in the LRN ( Figure 2D , 2D′ , 2E , 2E′; Figure S2 , and data not shown ) . Many Kcnj6 positive neurons were even observed within the inferior olivary territory ( black arrows in Figure 2E′ ) and dorsally to the IO at the midline ( open arrow in Figure 2E′ ) . In agreement with previous reports [16] , we found a slight enlargement of the IO at E14 . 5 when we used the Ets transcription factor Etv1 as a IO specific marker [33] ( Figure S3 and data not shown ) . However , by labeling for the axon-guidance-molecule B ( RgmB ) no alterations in the general architecture of the IO were seen [34] ( Figure S3 ) . This is consistent with our observation that misguided Pax6Sey/Sey mms neurons make only a negligible contribution to the IO or settle in the periphery of the IO . In summary , these data demonstrate that migration of Pax6Sey/Sey mms neurons is severely disrupted . Mutant neurons of the mms are initially delayed . Later , a number of neurons use a sub-marginal migration path , migration is disturbed at the midline and several neurons migrate to ectopic positions along the midline . Lastly , the normal structure of the LRN is lost , the ECN is completely missing , and the IO is enlarged . In order to dissect the complex neuronal cell migration defects observed in Pax6Sey/Sey mice , Pax6Sey/Sey mice were crossed to the Tg ( PAX6-tTA ) transgenic line . Comparison of cultures obtained from wt and from Pax6Sey/Sey embryos confirmed the anatomical observations described above . Pax6Sey/Sey mms cells showed an initial delay in the onset of the migration ( Figure 3A , 3B ) and a disturbance at midline crossing later on . After 5 DIV ( 5 days in vitro ) Pax6Sey/Sey mms neurons settled randomly in the LRN ( Figure 3B′ ) but failed to form any ECN structures ( Figure 3B'' ) . In contrast , wt cultures formed a well organized LRN in which cells settled in a dorsal and ventral sub-nucleus of the LRN ( Figure 3A′ ) and in a distinguished ECN ( Figure 3A'' ) . To quantify the effect we counted labeled cells in cultures from wt ( LRNi = ipsilateral LRN = 112±7; LRNc = contralateral LRN = 266±16; ECN = 86±14; total number of cells = 610±34; n = 50 ) and Pax6Sey/Sey embryos ( LRNi = 313±23; LRNc = 426±46; ECN = 8±2; total number of cells = 670±71; n = 8 ) ( Figure 3K ) . These data suggested that there were no alterations in the gross number of migrating neurons between wt and Pax6Sey/Sey embryos and confirmed the complete absence of an ECN in Pax6Sey/Sey embryos . In both , mutant and wt tissues , a proportion of LRN neurons settled ipsilaterally ( Figure 3K ) . All LRN neurons , however , projected to the contralateral cerebellum in respect to their origin from one rhombic lip , explaining the observations by Bourrat and Sotelo of an ipsilateral and contralateral contribution of mossy fibers [17] . The Pax6Sey/Sey migration defect could be caused either by a direct cell autonomous action of Pax6 in migrating RL precursors or via an indirect non-autonomous effect , for example in the ventral domain of Pax6expression ( e . g . by altering migration cues at the midline ) . We performed three types of experiments to discriminate between these alternatives . First , we transplanted transfected Pax6Sey/Sey rhombic lips onto wt brainstems and vice versa ( Figure 3C , 3D ) . Unexpectedly , migrating Pax6Sey/Sey neurons ( in a wt host ) formed a well organized LRN ( Figure 3C′ ) , but no ECN ( Figure 3C'' ) . In contrast , wt neurons ( in a mutant host ) failed to form a correctly organized LRN ( Figure 3D′ ) , but were able to generate a normal ECN ( Figure 3D'' ) . These data suggested , that Pax6 may act cell autonomously in generating ECN neurons , but non-autonomously in specifying the correct sub-organization of LRN neurons . To further validate this assumption we rescued the Pax6Sey/Sey migration defect by re-expression of Pax6 . We tested the two major splice variants and of these , the expression of the Pax6 ( -5a ) isoform in PaxSey/Sey6 rhombic lips resulted in a full recovery of the ECN ( Figure 3E'' ) , but a disorganized LRN ( Figure 3E′ ) , whereas , the Pax6 ( +5a ) variant was ineffective ( not shown ) . Thus , in the RL Pax6 splice variants differ in their biological activity , similar to the embryonic cortex [35] . Lastly , we diminished the endogenous Pax6 mRNA by using siRNAs . Transfection of siRNAs or over expression of shRNA constructs directed against Pax6 ( Figure S4 ) resulted in a massive reduction of ECN cells in wt explants ( Figure 3F'' ) , whereas , control constructs had no effect ( not shown ) . The effect was quantified by counting labeled cells that had settled in the ECN ( Figure 3L ) . Taken together , these experiments demonstrate that our model system enables the simultaneous visualization and manipulation of tangentially migrating cells in the mouse brainstem . In addition , we have shown that Pax6 plays numerous distinct roles in the formation and migration of mossy fiber producing neurons . Moreover , the combination of a binary model and organotypic culture assays facilitates a quick discrimination between cell-autonomous and non-autonomous effects . We identified several genes whose expression was altered in Pax6Sey/Sey mms neurons ( Figure S5 . ; see also Materials and Methods ) . To gain more insights into the function of these putative Pax6 downstream targets all genes were over-expressed or their expression level was diminished with shRNAs . Those genes which showed the most noticeable effects are summarized in Table 1 . The altered migration and settling behavior of Pax6Sey/Sey ECN/LRN neurons suggested that migration cues were changed in Pax6Sey/Sey embryos . The most prominent candidates are ligand/ receptor couples of the Slit/ Robo- and Netrin/ Dcc- pathways [36] , [37] . Expression of Netrin1 , Dcc , and Robo1 , 2 and 3 was unaltered in migrating Pax6Sey/Sey mms neurons ( Figure 4A , 4A′ , Figure S2 , and data not shown ) . However , Slit1 and Slit2 which were expressed in the hypoglossal nuclei were both lost in Pax6Sey/Sey embryos ( Figure 4B , 4B′ and Figure S2 ) [5] , [10] . Motorneurons of the hypoglossal nuclei are in close proximity to the LRN settling territories suggesting that Slit1 and Slit2 expression provided from these neurons may determine the place of LRN settlement . To test this hypothesis we performed transplantation experiments and shRNA driven knock-down of the Slit-receptor Robo3 in migrating mms neurons . Both types of experiments resulted in a disorganized LRN similar to the phenotype observed in Pax6Sey/Sey mice ( Figure 4C-4H ) . The above results indicate that factors provided from the hypoglossal nucleus , ( most likely Slit1 and Slit2 ) determine the place of LRN settlement . These data also explain the cell non-autonomous role of Pax6 during this process . Hypoglossal neurons are Pax6 negative , but are completely lost in Pax6Sey/Sey embryos ( Figure S2 ) [5] , [10]; hence , Slit1 and Slit2 are most likely not direct targets of Pax6 . Additional experiments suggest that Slit1 and Slit2 may also act as repellent to push mms neurons to the marginal migration route during the initial phase of migration ( data not shown ) . Two POU transcription factors were among the genes whose expression pattern was altered in the mms of Pax6Sey/Sey embryos . Pou4f2 ( also: Brn3b ) was strongly expressed in about 18 . 6% ( ±4 . 6% , n = 3 ) of E14 . 5 and 23 . 3% ( ±6 . 2% , n = 3 ) of E15 . 5 wt mms neurons but was completely lost in the Pax6Sey/Sey mms ( Figure 5A , 5A′ , 5B , 5B′ ) . Pou4f1 ( also: Brn3a ) was expressed between E13 . 5 and E15 . 5 in a subset of mms neurons , but was up-regulated in the E14 . 5 and E15 . 5 Pax6Sey/Sey mms ( Figure 5C , 5C′ ) . Expression of Pou4f1 and Pou4f2 in Pax6Sey/Sey IO neurons was unaltered ( Figure 5A , 5A′ , 5C , 5C′ ) . Pou4f2 plays several roles in specifying and guiding retinal ganglion cells and their axons . We therefore asked whether Pou4f2 may accomplish similar tasks in rhombic lip derived neurons . Pou4f2 was only expressed in a subset of wt mms neurons . We therefore over-expressed Pou4f2 in all migrating mms neurons . Remarkably , growth cones of all Pou4f2 over-expressing neurons were arrested at the midline for about 1 . 5 days ( ±0 . 5 days , n = 17 ) , whereas the majority axons in control cultures crossed the midline instantly ( Figure 5D , 5F ) . Interestingly , in control cultures the growth cones of some neurons also appeared to be arrested at the midline: 5% ( ±3% ) at 1DIV , 15% ( ±6% ) at 2DIV , 25% ( ±5% ) at 3DIV , and 6% ( ±3% ) at 4DIV ( n = 11 ) . This correlates well to the peak of Pou4f2 expression at E14 . 4 and E15 . 5 ( in cultures: 2DIV and 3DIV ) . Over-expression of Pou4f2 had also a noticeable effect on the settling behavior of LRN neurons . Quantification of LRN neurons at 5DIV revealed that Pou4f2 expressing LRN neurons preferably settled at the ispilateral side ( LRNc/LRNi = 0 . 8 ± 0 . 1 , n = 17; Figure 5G , 5M ) compared to control cultures in which the majority of LRN neurons settled at the contralateral side ( LRNc/LRNi = 2 . 5 ±0 . 1 , n = 50; Figure 5E , 5M ) . Similar relations were obtained at 6DIV and 8DIV suggesting that Pou4f2 over-expression altered the migration behavior of mms neurons and did not cause a delayed settlement of these neurons . The effect was specific to Pou4f2 and could not be mimicked by over-expression of Pou4f1 , Pou4f3 or Pou6f1 ( Figure 5M and data not shown ) . Together these data suggest that Pou4f2 acts through a novel mechanism which induces an arrest of growth cones at the midline to regulate the ratio of ipsilaterally versus contralaterally settling neurons . We altered expression levels of about 25 potential Pou4f2 retinal target genes [38]–[40] and of these two showed an effect on the migration behavior of mms neurons . Over-expression of Gfi1 , a zinc finger transcription factor , reduced the contra-/ipsi-lateral ratio of LRN neurons ( Figure 5M ) . In contrast , the down-regulation of Gap43 by shRNA constructs caused a higher contra-/ipsi-lateral ratio of LRN neurons ( Figure 5M and Figure S4 ) . Gap43 is slightly reduced in the Pax6Sey/Sey mms ( Figure S5 ) . In addition , mis-expression of Pou4f2 resulted in a massive down-regulation of Pou4f1 in transfected , but not in control , rhombic lips ( Figure 5N ) , suggesting that the loss of Pou4f2 in Pax6Sey/Sey mms neurons leads to an up-regulation of Pou4f1 ( Figure 5C , 5C′ ) . Pou4f1 over-expression or down-regulation , however , did not alter migration behavior of mms neurons ( Figure 5M ) . Pou4f2 is expressed only in a subset of Pax6 positive mms neurons suggesting that other factors together with Pax6 may co-regulate Pou4f2 . In the developing retina Pou4f2 expression depends on two transcription factors: the bHLH protein Math5 and the zinc finger gene Wt1 [41]–[44] . Wt1 was found to be expressed in the rhombic lip , though , in a region just dorsally to the Pax6 positive domain ( Figure 5J ) . Math5 was neither expressed in the rhombic lip nor in migrating mms neurons , however , a close homologue , Math1 , was expressed in neuronal precursors at the rhombic lip and in a subset of early migrating mms neurons [22] , [23] ( Figure 5K ) . Thus , Math1 , but neither Math5 nor Wt1 , was the most likely candidate to regulate Pou4f2 or Pou4f1 expression in mms neurons . Consistent with this , mis-expression of Math1 , but not of Wt1 ( + and – KTS splice variants ) or Math5 , led to a midline arrest of migrating mms neurons and a reversed settling behavior of LRN neurons ( Figure 5H , 5I , 5M ) . Mis-expression of Math1 resulted in an up-regulation of Pou4f2 and a down-regulation of Pou4f1 ( Figure 5O , 5P ) . Together , these data suggest , that Pou4f2 expression in rhombic lip derived mms neurons depends on Pax6 and Math1 and that Pou4f2 may regulate Pou4f1 and Gap43 in mms neurons . In summary , our work led to the identification of a gene cascade acting in tangentially migrating neurons of the brainstem , in which Pou4f2 plays a central role to induce a previously unknown mechanism that controls midline crossing behavior . Furthermore , our results imply that our model system is applicable to quickly analyze genetic hierarchies in Pax6 positive cells and may therefore serve as a general tool .
The extraordinary complexity of cell determination , migration and wiring processes in the developing mammalian brain creates a major challenge for developmental neurobiologists . Here , we introduced a simple yet powerful technology to quickly analyze any gene potentially involved in these processes . Our model is of threefold use: first to study the function of Pax6 and of Pax6 downstream genes in their genuine environment , second to investigate genes involved in general patterning , axonal pathfinding and cell migration processes , and third to enable the analysis of tissue-specific gene functions . The Tg ( PAX6-tTA ) model complements and improves existing approaches and has certain benefits: it combines cell specific transfection protocols and organotypic culture assays , thus , facilitating the quick analysis of genes in a natural tissue environment . The experimental design and the binary nature of the Tg ( PAX6-tTA ) model is fundamentally simple and has several advantages over systems that are based purely on transgenic animals . First , the electroporation and subsequent culture of embryonic tissues allows the screening of large number of genes without the need of generating new transgenic animals for each construct . In fact , less than 10% of the constructs we have tested revealed phenotypes . Thus , only those genes showing positive results in culture assays may be used subsequently to generate transgenic lines . Of note however , some of the phenotypes reported here , for example , the midline arrest or the altered ipsi- to contra-lateral ratio , would have been missed in purely transgenic systems . Second , variation of the electroporation protocol allows transfections ranging from just a few cells to a complete Pax6 expression domain with thousands of cells . Hence , our approach allows adjustment according to the needs: either to monitor single migrating cells or to determine global patterning effects . In addition , neighboring cells and non-electroporated contra-lateral sides serve as internal controls . The usefulness of our system critically depends on the tightness of the TRE based promoter and on the ability of the constructs to express two genes simultaneously . To ascertain the tightness of our system we used repeated electroporations and high DNA concentrations ( up to 5 µg/µl ) . Even under these extreme conditions we were never able to detect any reporter gene expression in Pax6 negative cells at any developmental stage . Thus , under the conditions used in this report the combination of Tg ( PAX6-tTA ) mice and TRE based promoters allow expression of reporter gene constructs only in Pax6 positive cells . It is also important to note , that our strategy to use a YAC based technology combined with an internal ribosomal entry site ( IRES ) resulted in moderate levels of reporter gene expression which were in the range of physiological concentrations . To ensure the simultaneous expression of two reporter genes we tested several types of TRE constructs . Only our approach , to use two consecutive TRE based promoters led to the activation of nearly equal amounts of two genes at the same time in the same cell . A bidirectional TRE element that previously had been shown to work in transgenic animals failed in our system [45] . One obvious difference is that in transgenic animals typically multiple copies of constructs are stably integrated into the genome , whereas , in our assay transfections were transient . Pax6 loss of function phenotypes are often highly complex involving massive malformations in the affected organs . Pax6 is expressed in neuronal precursors of the telencephalon , commissural neurons in the dorsal spinal cord , in adult neuronal stem cells , the early eye cup , in the pancreas , in precursors and in migrating cells of several tangential and radial migration streams of the rhombencephalon and of the forebrain [5] , [6] , [10] , [11] , [14] , [28] , [46] , [47] . In addition to its technical advances , the Tg ( PAX6-tTA ) model represents a novel , highly versatile technology to study the function of Pax6 or any other gene in these tissues . As a paradigm , we have dissected the role of Pax6 in tangentially migrating cells of the brainstem . In principle , however , this system shall be applicable to any Pax6 positive tissue and we have initial evidence that our model allows to specifically target Pax6 positive telencephalic precursor cells , cerebellar granule cells , the developing retina , the rostral migratory stream , the pontine migration and ventral precursor cells of the brainstem and spinal cord ( Figure S1 and data not shown ) . With the help of this model it should therefore be possible to systematically analyze cell fate decisions and the migratory behavior of Pax6 expressing cells at any developmental stage . Several studies have revealed that Pax6 is required for hindbrain and spinal cord development [5] , [7] , [10] , [14] , [15] . Our work adds that Pax6 also controls the determination and migration of rhombic lip derived neurons ( for a summary see Table 1 and Figure 6 ) . Pax6 functions twofold: first , Pax6 controls guidance cues which push migrating mms neurons to the marginal path and which control the settling pattern of LRN neurons . The most likely sources of these cues are the hypoglossal nuclei which are located close to the midline and in proximity to the LRN . Slit1 and Slit2 are expressed in the hypoglossal nuclei and the Slit receptor Robo3 is expressed in migrating mms neurons [48] . Slit expression provided by the hypoglossal nuclei may therefore act as repellent to push mms neurons to a marginal migration route and may also specify the settlement of neurons in the LRN . The loss of Slit-expressing hypoglossal nuclei in Pax6Sey/Sey embryos [5] , [10] causes a major reduction of the repellent ( a minor source of Slit is still present in midline cells ) . Consequently , migrating mms neurons would use a more sub-marginal migration route and settle less organized in Pax6Sey/Sey embryos . Furthermore , Slit expression at the RL may be involved during the initial phase of mms migration . Secondly , Pax6 functions cell-autonomously in migrating mms neurons to control the determination , the timing of migration , and midline crossing . Several genes show altered expression in Pax6Sey/Sey mms neurons ( Table 1: Pou4f1 , Pou4f2 , Unc5h1 , Mafb , Chordin ) and may convey individual aspects of migration . We and others find that several transcription factors relay Pax6 downstream effects in dorsal brainstem neurons: Ngn1 in precursor cells ventral to the RL [16] , and Pou4f1 , and Pou4f2 in migrating neurons ( this report ) . Mis-expression of Ngn1 or Ngn2 in Pax6Sey/Sey embryos failed to rescue the migration defects observed in the Pax6 mutant ( Table 1 ) . Neither did the mis-expression or down-regulation of these genes generate small eye - like migration defects in wt embryos ( Table 1 ) . On the other hand , Pou4f2 , which is lost in the mms of Pax6Sey/Sey embryos ( Figure 5B and also in the pontine migration and in the cerebellum , data not shown ) , alters migration behavior of mms neurons . Together these data suggest that Pou4f2 may regulate genes involved in pathfinding processes , whereas , Ngn1 acts earlier in the cell determination process . There are striking similarities in gene expression pattern between sensory neurons and RL derived neurons . We found that at least two thirds of the genes which are co-expressed with Pax6 and Pou4f2 in retinal ganglion cells are also co-expressed with these genes in mms neurons . Furthermore , genetic hierarchies seem to be analogous: in the retina Math5 controls Pou4f2 , which then acts upstream of Pou4f1 [38] , [41]–[43] , whereas , in RL derived neurons Math1 , a close homologue of Math5 , initiates related pathways . General genetic pathways are conserved between retinal and RL derived neurons and our model may therefore help to elucidate some of the phenotypes observed in Pou4f1-/- and Pou4f2-/- mice . Both mouse models have revealed distinct axonal pathfinding errors [39] , [49]–[52] . Mis-expression of Pou4f2 ( or Math1 ) in RL derived neurons stalls growth cones at the midline for several hours . To our knowledge , this is the first report of such a midline arrest and it may thereby be a paradigm for a novel mechanism controlling midline crossing . The arrest does neither induce a growth cone collapse nor does it inhibit midline crossing per se as all neurons generate axons that cross the midline after a “waiting period” . These axons all migrated into the cerebellum like those of control cultures . Gfi1 mis-expression and Gap43 knockdown were able to partially mimic the Pou4f2 induced phenotype , however , additional unknown targets or a combinatory code may be needed to elicit the full phenotype . As Pou4f2 was only expressed in about 1/4 of wt mms neurons , the loss of Pou4f2 in Pax6Sey/Sey embryos mimics only minor aspects of the Pax6Sey/Sey phenotype . The down regulation of Pou4f2 by shRNA constructs resulted in a severe midline disturbance of neuronal processes at similar to the phenotype observed in Pax6Sey/Sey cultures , whereas , in control cultures neuronal processes crossed the midline instantly ( data not shown ) . Comparable phenotypes were also observed in Pax6Sey/Sey cultures , in cultures transfected with Pax6 shRNA constructs , and in Pax6Sey/Sey brainstem sections . Tangentially migrating neurons follow similar navigational cues as developing axons [14] , [19] , [48] , [53]–[58] . Hence , tangentially migrating neurons of the mms provide an excellent system to study axonal pathfinding and neuronal cell migration processes . Migrating mms neurons are easily accessible as they navigate along the pial surface . Our model takes advantage of the superficial migration of these neurons and provides a straightforward assay to specifically label and manipulate these cells without affecting their surroundings . Members of most families of guidance receptors ( Netrin receptors , Slit receptors , Semaphorins , Eph receptors , and Ephrins ) are expressed in migrating mms neurons [48] , [53] , [54] ( Engelkamp , unpublished ) and at least two of these pathways are essential for the correct guidance of mms neurons: the Slit/ Robo - [48] and the Netrin/ Dcc- pathways [53] , [56] , [57] ( see also Table 1 ) . Our system should therefore also have important implications for the study of the signal cascades entailed in these pathways . In summary , we have established a novel model system which allows the simultaneous visualization and manipulation of neuronal subpopulations . As a prototypical model we have focused on the role of Pax6 in migrating brainstem neurons . Yet , our results imply that our model system is applicable to a range of other cells in the developing brain and may therefore serve as a general tool to quickly study axonal pathfinding , neuronal cell migration or patterning processes .
The Small Eye allele [4] was maintained on a CD1 background . Embryos were obtained from matings of heterozygote ( Pax6Sey/+ ) mice . 0 . 5 denotes the morning when the vaginal plug was found . Experiments were always performed on matching pairs of control ( wt ) and Pax6Sey/Sey embryos that were carefully staged . All phenotypes described were confirmed on at least six individual Pax6Sey/Sey embryos obtained from different crossings . There was no noticeable phenotypic difference between Pax6Sey/+ and wt embryos and therefore , in our experiments wt designates wt and Pax6Sey/+ embryos . For brainstem cultures , matings between heterozygote Tg ( PAX6-tTA ) and wt CD1 mice or between heterozygote Pax6Sey/+/ Tg ( PAX6-tTA ) and heterozygote Pax6Sey/+ mice ( to generate Pax6Sey/Sey cultures ) were used . Genotyping was performed by PCR with primers directed against the Tet repressor ( upper: GCGCTGTGGGGCATTTTACTTTAG; lower: CCGCCAGCCCCGCCTCTTC ) . All animal procedures were carried out in accordance to the guideline approved by institutional protocols . YAC Y593 [30] was modified such that exons 8 to 11 of the Pax6 gene were replaced by homologous recombination with a construct containing the following elements in 5′ to 3′ order: Pax6k30 – IRES – tTA – loxP – LYS2 – loxP – Pax6k32 . Pax6k30 and Pax6k32 corresponded to the sequences 29 . 792 to 30 . 296 and 31 . 587 to 32 . 095 of the Pax6 cosmid cFAT5 ( NCBI accession no . Z95332 ) , respectively , and were generated via PCR . The IRES ( internal ribosomal entry site ) was derived from pIRES-EGFP ( Invitrogen ) , however , the original ATG-11 start codon was reconstituted to enhance translational initiation . tTA ( Tet-On-system ) , a fusion of the tetracycline repressor and the activation domain of VP16 , was derived from pUHD15-1neo ( Clontech ) . The LYS2 gene from S . cerevisiae was derived from pAF107 , which was obtained from B . Dujon , Institute Pasteur , Paris , France [59] . LoxP sequences were generated via PCR . All constructs were sequence verified . Homologous recombination in yeast was performed using standard techniques . The integrity of the recombined YAC was then verified by PCR and southern blotting . Preparation of the YAC DNA and the generation of transgenic mice were as described [30] . In situ hybridization was performed on free floating vibratome sections as previously described [14] . Probes for Math1 [60] , Neurod1 and Neurod2 [61] , Pax6 [47] , Unc5h3 [62] and Slit1 , Slit2 [63] were obtained from H . Zoghbi , A . Bartholomä , R . Hill , S . Ackerman , and M . Little , respectively . Probes for Dcc and RgmB were as published [34] , [64]; other probes were obtained by RT-PCR . The PCR products were subcloned and their identities were confirmed by sequencing . The general staining patterns of all probes matched published expression patterns . Probes were as follows: Etv1 ( bp 853–1820 of NM_007960 ) ; Fgfr2 ( bp 343–1192 of NM_201601 ) ; Pou4f1 ( bp 1321–2199 of NM_011143 ) , Pou4f2 ( bp 216 – 1762 of S68377 ) ; Robo3 ( bp 3648–4673 of NM_011248 ) . Several genes which are down- or up-regulated in Pax6Sey/Sey embryos were identified with the help of a large scale in situ screen using >300 putative candidates . Individual probes are available on request . The vector for the co-expression of two constructs in Tg ( PAX6-tTA ) mice contained the following elements in 5′ to 3′ order: MCSI – TRE – PminCMV – IntronA - BGHPolyA – MCSII – TRE - SV40PolyA; MCS = multiple cloning sites; TRE = 7 repeats of the tetracycline responsive element , PminCMV = minimal CMV promoter , and IntronA were from ptetOi-MCS ( obtained from Martin Spiegel , Tübingen ) ; SV40polyA and BGHPolyA = polyadenylation signals ( derived from pTetOi-MCS and pRc/CMV , Invitrogen , respectively ) . Fluorescent markers to label migrating cells were a modified EGFP or DsRed2 ( Clontech ) . Full length clones for Gfi1 , Math1 , Math5 , Ngn1 , and Ngn2 were obtained from the German Resource Center for Genome Research ( RZPD ) and sequence verified; clones for all other genes were obtained by RT-PCR and confirmed by sequencing . Fusions with a triple HA-tag or the engrailed repressor domain ( EnR ) were generated by PCR . shRNA constructs were generated in the psiSTRIKE vector ( Promega ) using the Promega Web tool for designing the hairpin oligonucleotides . In the psiSTRIKE vector shRNAs are expressed under control of the U6 RNA polymerase promoter . Efficiency of shRNA knockdown was demonstrated in HEK293 cells using the psiCHECK/ Dual Luciferase system according to the manufacturers protocol ( Promega ) . All constructs were sequence verified . Responder constructs ( 2-4 µl at 0 . 5 µg/µl in GBSS/ 0 . 01% Methyl Fast , Sigma ) were injected into the fourth ventricle of E12 . 5 wt and Tg ( PAX6-tTA ) mouse embryos by using glass needles . Electroporation was then performed with forceps-like electrodes with platinum ending ( Ø = 0 . 5 mm ) ( one Electrode above the right RL and the other under the left jaw ) . Conditions were 8 pulses at 50V , 50msec with a pulse interval of 1sec . We used the square pulse generator EPI2500 ( L . Fischer , Heidelberg ) . After electroporation , the hindbrain ( rhombomeres 1–8 including the cerebellar anlage ) was dissected , opened at the roof plate and cultivated with the ventricular site onto MillicellCM filters ( Millipore ) in culture medium ( DMEM/F12 ( 1∶1 ) ; 0 . 6% Glucose; 0 . 02 mM Glutamine; 5 mM HEPES; 5% Fetal Calf Serum; 5% Horse Serum ) at 37°C and 5%CO2 . Depending on the antibodies used , brainstem preparations were fixed with 4% or 0 . 2% PFA in PBS for 12 hours at 4°C . Cryosections were cut at 14 µm . Primary and secondary antibodies used for staining were as follows: mouse monoclonal ( mAb ) α-Pax6 ( [10] , 1∶1000 , DSHB ) ; rabbit pAb α-Pou4f2 ( also Brn3b , 1∶300 , Covance ) ; rabbit pAb α-Kcnj6 ( also Girk2 , 1∶300 , Chemicon ) ; rabbit α-Wt1 ( Santa Cruz ) ; rabbit α-VP16 ( Clontech ) ; α-HA-tag ( 1∶100 , Roche ) and α-mouse and α-rabbit secondary antibodies conjugated with Alexa488 or Alexa596 ( Molecular Probes ) . Quantification of Pou4f2 positive mms neurons was done on every 3rd of serial sections double stained for Pax6 and Pou4f2 . Images were taken at a Zeiss Axiophot microscope equipped with a Spot camera , at a confocal Zeiss LSM microscope , or at a Leica MZ12 equipped with a camera device . Images were processed using the MetaView software ( Universal Imaging Corporation ) and Adobe Photoshop . To perform statistical analysis the position of the ECN and the LRN were determined in wt un-manipulated cultures by in situ RNA staining of Pax6 and Kcnj6 . The resulting territories were then overlaid onto the electroporated cultures with the help of three landmarks: a ) the position of the rhombic lip; b ) the position of the floor plate; and c ) the position of the superior and inferior olivary complexes , which both are visible in phase contrast images of the cultures . This procedure allowed classifying 97% of labeled neurons on the contralateral side and 90% on the ipsilateral side as either ECN or LRN neurons . The remaining 3% ( or 10% for the ipsilateral side ) of labeled cells were scattered neurons mainly in between the ECN and the LRN . Quantification of growth cones arrested at the midline in wt cultures was done by counting all growth cones in a 25 µm wide territory at the midline . Continuous observations of cultures implied that mms growth cones traveled at an average speed of at least 500 µm/day , suggesting that within any 25 µm interval only 5% of growth cones should be detected if migration would not pause . Quantification of the volume of the inferior olive was done with AxioVision ( Zeiss ) . | In mammals , many genes execute a unique set of distinctive and common functions in different cell types . Strategies to address these individual roles often involve the generation of series of transgenic animals . Here , we present a novel approach that combines a single transgenic mouse line with tissue-specific transfection protocols and organotypic cultures to enable the quick analysis of numerous genes in a cell-specific manner . As a proof of principle , we analyzed the function of transcription factors in tangentially migrating neurons of the developing vertebrate hindbrain . We identified a temporary halt in migration as a novel mechanism for neurons to decide whether to cross or not cross the midline . Our model may serve as a general tool to quickly study axonal pathfinding , neuronal cell migration , or patterning processes in a well-defined population of neurons . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"developmental",
"biology",
"embryology",
"developmental",
"neuroscience",
"neural",
"circuit",
"formation",
"axon",
"guidance",
"molecular",
"development",
"biology",
"neuroscience",
"pattern",
"formation",
"cell",
"fate",
"determination"
] | 2011 | A Novel Approach to Selectively Target Neuronal Subpopulations Reveals Genetic Pathways That Regulate Tangential Migration in the Vertebrate Hindbrain |
Kaposi’s sarcoma associated herpesvirus ( KSHV ) infection stabilizes hypoxia inducible factors ( HIFs ) . The interaction between KSHV encoded factors and HIFs plays a critical role in KSHV latency , reactivation and associated disease phenotypes . Besides modulation of large-scale signaling , KSHV infection also reprograms the metabolic activity of infected cells . However , the mechanism and cellular pathways modulated during these changes are poorly understood . We performed comparative RNA sequencing analysis on cells with stabilized hypoxia inducible factor 1 alpha ( HIF1α ) of KSHV negative or positive background to identify changes in global and metabolic gene expression . Our results show that hypoxia induces glucose dependency of KSHV positive cells with high glucose uptake and high lactate release . We identified the KSHV-encoded vGPCR , as a novel target of HIF1α and one of the main viral antigens of this metabolic reprogramming . Bioinformatics analysis of vGPCR promoter identified 9 distinct hypoxia responsive elements which were activated by HIF1α in-vitro . Expression of vGPCR alone was sufficient for induction of changes in the metabolic phenotype similar to those induced by KSHV under hypoxic conditions . Silencing of HIF1α rescued the hypoxia associated phenotype of KSHV positive cells . Analysis of the host transcriptome identified several common targets of hypoxia as well as KSHV encoded factors and other synergistically activated genes belonging to cellular pathways . These include those involved in carbohydrate , lipid and amino acids metabolism . Further DNA methyltranferases , DNMT3A and DNMT3B were found to be regulated by either KSHV , hypoxia , or both synergistically at the transcript and protein levels . This study showed distinct and common , as well as synergistic effects of HIF1α and KSHV-encoded proteins on metabolic reprogramming of KSHV-infected cells in the hypoxia .
Kaposi’s sarcoma associated herpesvirus ( KSHV ) is the etiological agent of Kaposi sarcoma , primary effusion lymphoma and multicentric Castleman disease [1–3] . By altering the expression of core metabolic enzymes , KSHV infected cells acquire a metabolic strategy of aerobic glycolysis generally referred as to the Warburg effect where these cells drive a high rate of glycolysis even in the presence of molecular oxygen [4–8] . This alteration of host metabolism mimicking Warburg effect by KSHV is believed to be necessary for the maintenance of latently infected cells [4] . Similar to most cancer cells , the mitochondria is also an organelle targeted by KSHV in viral infected cells altering apoptotic pathways and metabolism , so necessitating up-regulation of glycolysis to compensate for the energy demands of rapidly growing cells [9–12] . Metabolite profiling of KSHV infected cells suggest a wide difference between metabolite pools of KSHV infected cells when compared to control cells , including those which are common to anabolic pathways of most cancer cells [5] . Interestingly , the metabolite changes are not limited to only carbohydrates , but also included fatty acids and amino acids where inhibition of key enzymes in this pathway led to apoptosis of infected cells [5 , 13] . KSHV infection-mediated elevation of metabolites pools are due to enhanced anabolic activity rather than degradation from respective macromolecules [5] . Previous attempts to identify the mechanism of such reprogramming confirm increased expression of host factors such as glucose transporters , as well as hypoxia inducible factor ( HIF1α ) which are prerequisites for such changes in KSHV infected cells . In addition , decreased mitochondrial copy number and down regulated EGLN2 and HSPA9 have been reported upon over-expression of KSHV coded microRNAs , and are believed to be among the many KSHV factors involved in metabolic changes [14] . Nevertheless , previous observations either do not support or was unable to determine the involvement of other KSHV-encoded factors involved in metabolic differences caused by KSHV infection . Hypoxia and HIF1α play critical roles in pathogenesis of KSHV by modulating expression of KSHV genes as well as stabilizing several KSHV-encoded proteins [15 , 16] . KSHV infection alone can mimic several physiological and metabolic changes due to hypoxia and those common to cancer cells . Hypoxia on the other hand plays an important role in KSHV reactivation biology where HIF1α facilitates KSHV-encoded RTA-mediated reactivation by binding with LANA and up-regulating RTA expression [16 , 17] . Hypoxia is also reported to enhance viral reactivation potential associated with 12-O-tetradecanoylphorbol-13-acetate [18] . The role of hypoxia in maintenance of latency and KSHV associated pathogenesis is also crucial , where the promoter of the key latent gene cluster coding for LANA , vFLIP and vCyclin harbors hypoxia responsive elements and can be activated by HIF1α [15] . Among other KSHV factors affecting the HIF1α axis , is the constitutively active G protein-coupled receptor ( vGPCR ) encoded by KSHV [19 , 20] . vGPCR is a bonafide oncogenic protein and stimulates angiogenesis by increasing the secretion of vascular endothelial growth factor ( VEGF ) , which is a key angiogenic stimulator and a critical mitogen for the development of Kaposi’s sarcoma [21 , 22] . KSHV-encoded vGPCR enhances the expression of VEGF by stimulating the activity of the transcription factor HIF1α , which activates transcription from a HRE within the 5'-flanking region of the VEGF promoter [23] . Stimulation of HIF1α by KSHV encoded vGPCR involves phosphorylation of its regulatory/inhibitory domain by p38 and mitogen-activated protein kinase ( MAPK ) signaling pathways , thereby enhancing its transcriptional activity [24] . Specific inhibitors of the p38 / MAPK pathways are able to inhibit the transactivating activity of HIF1α induced by the KSHV-encoded vGPCR , as well as the VEGF expression and secretion from cells expressing this receptor [24] . These findings suggest that the KSHV-encoded vGPCR oncogene subverts convergent physiological pathways leading to angiogenesis and provides the first insight into a mechanism whereby growth factors and oncogenes acting upstream of MAPK , as well as inflammatory cytokines and cellular stresses that activate p38 , can interact with the hypoxia-dependent machinery of angiogenesis [24] . However , the role of vGPCR in modulating other physiological pathways is poorly explored . In the present study , we investigated the role of stabilized HIF1α on the metabolic status of KSHV positive cells and compared the results with KSHV negative cells with the same genetic background under normoxic or hypoxic conditions . We present data for differentially expressed KSHV-encoded genes when HIF1α is stabilized . We then showed the changes in global transcription of cells growing in normoxia or hypoxia with HIF1α to identify the common targets of HIF1α and KSHV infection . Our results showed enhanced induction of a tumorigenic metabolic phenotype in KSHV-positive cells growing in hypoxia compared to KSHV-negative cells growing under the same condition . Further , we now identify a comprehensive list of metabolic genes differentially expressed on KSHV infection in the hypoxic environment . These results provide new insights into the role of KSHV factors , in cooperation with hypoxia on the global metabolic status of KSHV positive cells .
KSHV infection is known to stabilize HIFs and this stabilization provides the cells a mechanism to survive in a hypoxic environment by up-regulating several cellular pathways involved in metabolism , survival and angiogenesis . We wanted to determine how KSHV infected cells respond compared to their isogenic KSHV-negative counterparts in hypoxic environments , and their metabolic requirements . There is no available control cell line with the same isogenic background for comparative studies in B-cells . Therefore , we selected KSHV-negative BJAB cells and KSHV positive BJAB-KSHV stably infected with KSHV [25] . We first characterized and confirmed the presence of full length KSHV in BJAB-KSHV cells at the level of the viral genome and transcriptome to determine if gross genomic alterations had occurred . Amplification of 10 different KSHV genomic regions with KSHV specific primers confirmed the presence of a KSHV genome most likely intact in BJAB-KSHV cells ( S1A and S1B Fig ) . The sequences of the primers used to characterize BJAB-KSHV cells are also provided in S1 Table . The BJAB-KSHV cells were further characterized at the level of transcripts by amplifying the KSHV-encoded latent gene vCyclin from the cDNA made from BJAB-KSHV cells ( S1C Fig ) . The isogenic background and authenticity of these two cell lines were also examined by short tandem repeat ( STR ) profiling . The STR profiling results for these two cell lines were compared with each other as well as with BJAB cells STR profile obtained from ExPASy Bioinformatics Resource portal database . The STR profile results confirmed the same origin and isogenic background of these cells ( S2 Table ) . To study the effects of hypoxia , we proceeded with two different approaches to induce hypoxia in cell culture . In the first approach , we treated the cells with CoCl2 , a chemical inducer of hypoxia to induce stabilization of HIF1α with minimal effects on the growth rate of the cells [26] . In the second approach we grew cells in 1% O2 hypoxic condition . Puromycin was omitted from the media of BJAB-KSHV and control cells for the entire treatment period . HIF1α stabilization was confirmed by western blot using HIF1α specific antibody ( Fig 1A and 1B ) . An estimation of glucose consumed by BJAB or BJAB-KSHV cells suggested that the bulk of glucose from medium was being consumed during the initial period of 24–48 hours , in which cells growing under normoxic conditions showed an exponential growth pattern ( Fig 1C & 1E ) . The growth patterns of cells growing in either CoCl2 or 1% O2 were quite different and showed diminished proliferation rates . Growing the cells in the same partially depleted medium showed a retarded growth in both normoxic as well as hypoxic environments , though hypoxic induction due to low oxygen showed a more drastic adverse effects on cell survival ( Fig 1C & 1E ) . The estimation of glucose consumed by BJAB and BJAB-KSHV cells grown in normoxia and hypoxia showed a large difference in the consumption of glucose between these cells . Within the initial 24 hours , BJAB-KSHV cells showed an almost 18% higher glucose consumption as compared to BJAB cells during the same time period ( 940 . 7mg compared to 773 . 8mg glucose per million cells ) ( Fig 1C & 1E ) . The BJAB-KSHV cells showed a similar increase in uptake of glucose throughout the time points of 48 , 72 and 96 hours compared to BJAB cells . A time dependent enhancement in the glucose uptake was observed for BJAB-KSHV cells when compared to BJAB cells growing in normoxic condition . However , the diminished medium condition and hypoxia due to 1% oxygen led to a drastic reduction in cell survival and growth past 72 hours . To rule out the possibility of an effect of puromycin pretreatment on glucose uptake , glucose consumption was also measured in cells growing either in the presence of puromycin , or its absence for 48 hours . The results showed no significant difference in glucose consumption due to presence or absence of puromycin in culture of BJAB-KSHV cells ( S1D Fig ) . The effect of puromycin due to hypoxic induction or its downstream target was also determined by measuring real time expression of HIF1α and VEGFA . The results showed no effects of puromycin on expression of HIF1α nor VEGFA ( S1E Fig ) . To estimate lactate released in medium by these cells a standard curve of lactate ranging from 0 to 10 nmol/μl was prepared followed by a pilot experiment to determine the range of lactate in the medium . Here , different volumes of fresh culture medium ( 1μl and 10 μl ) and 1μl medium from growing cultures were used ( S1F Fig ) . Based on the pilot experiment , 10 μl of a 10X diluted culture medium was used to estimate lactate released in medium by BJAB and BJAB-KSHV cells growing under normoxia or CoCl2/1%O2-induced hypoxia ( Fig 1D and 1F ) . A pattern similar to glucose uptake was observed for lactate release in these cells under similar growth conditions suggesting a directly proportional relationship between glucose uptake and lactate release . We also investigated whether this metabolic phenotype was mimicked in primary infection to peripheral blood mononuclear cells , KSHV infection of PBMCs was monitored growing them in the presence of CoCl2 or 1%O2 . The infection of PBMCs with KSHV was confirmed by immuno-staining for KSHV latent protein LANA and the induction of hypoxia was confirmed by western blot to detect HIF1α ( Fig 1G and 1H ) . The percentage of cells infected with KSHV was empirically calculated by LANA immune-staining . The infection efficiency of PBMCs with KSHV was approximately 50% . Estimation of glucose uptake and lactate release by infected PBMCs grown under conditions of normoxia or in CoCl2 or 1%O2 at 48 hours post-infection showed an enhanced glucose dependency and lactate release similar to BJAB and BJAB-KSHV cells ( Fig 1I and 1J ) . As the 1% oxygen for induction of hypoxia showed highly adverse effects on cell survival , we performed RNA sequencing experiments on BJAB and BJAB-KSHV cells growing in normoxia or CoCl2-induced hypoxia showing a more relevant physiological response of HIF1α stabilization due to KSHV infection . Analysis of RNA sequencing data for differential gene expression of KSHV encoded genes identified 42 transcripts coded by KSHV ( Fig 2A–2D ) . A histogram for genes across the KSHV genome is provided in Fig 2A . Statistical analysis revealed that expressions of 11 KSHV-encoded genes were significantly changed when grown under CoCl2–induced hypoxia compared to their normoxic counterpart . Among these 11 genes , the viral G-protein coupled receptor ( vGPCR ) , which is a constitutively active homolog of human G-protein coupled receptor [24] , was found to be up-regulated by 3 . 62 fold ( Fig 2B ) . Three other genes up-regulated due to hypoxia were K1 ( Immunoreceptor tyrosine-based activation motif containing signal transducing membrane protein ) , ORF2 ( homolog of cellular Dihydrofolate reductase ) , and ORF4 ( Complement binding protein ) with a fold change of 1 . 58 , 1 . 26 and 1 . 38 , respectively ( Fig 2B–2D ) . Among the down-regulated genes , K12 , ORF 40 , and vFLIP were heavily down-regulated with a fold change of -4 . 1 , -3 . 58 and -2 . 68 , respectively ( Fig 2B–2D ) . The levels of LANA and vCyclin transcripts were induced but not statistically significant due to possible differential efficiency of sequencing through these templates ( Fig 2B ) . However , they were clearly induced as shown by RT-PCR of cells grown in CoCl2 and 1% O2 ( Fig 2B & 2C ) . RTA transcripts were moderately increased as detected by sequencing , but was clearly increased when validated by RT-PCR in CoCl2 and 1% O2 ( Fig 2B & 2C ) . Interestingly , all the four KSHV encoding interferon regulatory factors ( vIRFs ) were down-regulated with a fold change of -3 . 19 , -2 . 19 , -1 . 7 and -2 . 69 for vIRF1 , vIRF2 , vIRF3 and vIRF4 , respectively ( Fig 2B–2D ) . To validate the results obtained from differential gene expression seen for KSHV-encoded genes by RNA-sequencing , real-time PCR was also performed for the individual genes using gene specific primers . The primers used for real-time PCR are provided in S3 Table . Similar results were obtained by real-time PCR assays where vGPCR showed the highest up-regulation and K12 as a greatest down-regulated gene ( Fig 2B & 2C ) . Similarly , RTA was shown to be up-regulated by RT-PCR in CoCl2 and 1% O2 , as expected ( Fig 2B & 2C ) . To further corroborate the differential gene expression of KSHV-encoded genes , we wanted to determine if a similar pattern was observed in low oxygen environment . BJAB-KSHV cells grown in a hypoxic chamber with 1% oxygen were collected and real-time PCR analysis was performed on the KSHV-encoded genes . The results showed a similar pattern of expression for the genes analyzed . However , the magnitude of change was slightly lower for vGPCR while it was slightly higher for K1 as compared to their expression in CoCl2-induced hypoxia ( Fig 2C & 2D ) . The expression of ORF2 , ORF4 , vFLIP , vCyclin , LANA and RTA was also observed with the same pattern as it was seen in CoCl2-induced hypoxia ( Fig 2C & 2D ) . Interestingly , the expression of some of the vIRFs were slightly less than that observed in CoCl2-induced hypoxia suggesting that the expression of vIRFs are also dependent on the overall ATP pool ( Fig 2D ) . To determine the physiological relevance of the differentially expressed KSHV-encoded genes in response to hypoxia , real time expression of vGPCR , K1 , vFLIP , vCyclin , LANA and RTA were also analyzed in the primary effusion lymphoma ( PEL ) cell line BC3 , grown in both CoCl2 as well as 1% O2 induced hypoxia . The results strongly supported a universal effect of hypoxia on the expression of these KSHV-encoded genes ( Fig 2E & 2F ) . These results led to further analysis of other critical KSHV-encoded genes when HIF1α was stabilized in the naturally infected KSHV positive cell line , BC3 . We analyzed expression of 27 candidate KSHV-genes from BC3 cells grown under normoxic and CoCl2 induced hypoxic condition . The primer sets used are included in S3 Table . The resulting data showed that ORF9 ( DNA polymerase ) , ORF18 ( involved in late gene regulation ) , ORF25 and ORF26 ( major capsid protein ) , ORF27 ( Glycoprotein ) , ORF28 ( BDLF3 EBV homolog ) , ORF34 , ORF40 ( Helicase-primase ) , ORF57 ( mRNA export/splicing ) , and ORFK14 . 1 were significantly up-regulated in BC3 cells grown under hypoxic conditions ( S2A–S2C Fig ) . Similarly , ORF11 ( predicted dUTPase ) , ORF31 ( nuclear ad cytoplasmic protein ) , ORF32 , ORF33 ( tegument proteins ) , ORF44 ( Helicase ) , ORF64 ( Deubiquitinase ) and ORFK14 ( vOX2 ) were significantly down-regulated in BC3 cells grown under hypoxic condition ( S2A–S2C Fig ) . The expression of ORF6 ( ssDNA binding protein ) , ORF7 ( virion protein ) , ORF8 ( Glycoprotein B ) , ORF36 ( serine protein kinase ) , ORF54 ( dUTPase/Immunmodulator ) , ORF56 ( involve in DNA replication ) , ORF69 ( BRLF2 nuclear egress ) and ORFK8 . 1 ( Glycoprotein ) showed little or no significant change ( S2A–S2C Fig ) . Based on the results showing HIF1α stabilization and up-regulated expression of vGPCR , we performed a bioinformatics analysis of the vGPCR promoter region for identification of possible hypoxia responsive elements ( HREs ) [16] . A search for HREs consensus ( ASGT; where S = C/G ) within the vGPCR promoter identified 9 different HREs ( Fig 3A ) . To determine the role of these HREs in directly regulating transcription of vGPCR in a HIF1α dependent manner , luciferase based reporter assays were performed . In brief , 10 different clones from the promoter region of vGPCR were generated and the results from the luciferase activity showed that the HREs at the 3rd , 4th , 5th , 6th , and 7th positions were significantly responsive to HIF1α ( although not equally responsive ) . The promoter region containing all 9 HREs showed the strongest response , followed by clone C6 containing the initial 5 HREs ( Fig 3C ) . The primers used to generate the clones are provided in S4 Table . Next , we wanted to determine if HIF1α knockdown in KSHV-positive cells can rescue the hypoxia associated expression of KSHV-encoded genes . The ShControl and ShHIF1α-BC3 cells were generated by lentivirus based transduction . Knock-down of HIF1α transcripts was confirmed at the transcript levels by real time PCR ( Fig 3D ) . We confirmed the expression of HIF1α at the protein level by HIF1α western blot of lysates from ShControl and ShHIF1α BC3 cells grown under CoCl2 or 1% O2 induced hypoxia ( Fig 3E ) . Real-time PCR analysis to determine the vGPCR and vFLIP expression in CoCl2 treated cells . HIF1α knockdown cells showed a reversal of expression as treatment with CoCl2 did not show the effect in HIF1α competent cells ( Fig 3F and S2D Fig ) . As vGPCR is a potent candidate for the activation of several proliferation pathways , we wanted to determine whether the metabolic phenotype observed for KSHV positive cells was only due to elevated HIF1α levels , or if vGPCR expression was sufficient to induce the metabolic changes . We transfected HEK293T cells with an expression plasmid coding for KSHV-encoded vGPCR and compared it to that of cells transfected with empty vector . We also compared the results with cells expressing HIF1α . The results suggest that hypoxia or vGPCR can modulate the metabolic phenotype ( Fig 3G and 3H ) . RNA sequencing on total RNA from BJAB and BJAB-KSHV cells grown under normoxic condition or CoCl2 induced hypoxic condition was analyzed to determine the differential gene expression profiles . To increase our confidence in the differential gene expression data for host genes , the fold-change difference for VEGFA ( a known target of HIF1α ) was first analyzed ( Fig 4A ) . Real time PCR validation was also performed for VEGFA transcripts from CoCl2 treated cells ( Fig 4B ) . A comparative analysis was performed between BJAB vs BJAB-CoCl2 , BJAB vs BJAB-KSHV and BJAB vs BJAB-KSHV-CoCl2 cells . The comparative analysis between BJAB cells grown under normoxia and CoCl2 induced hypoxia revealed major transcriptional changes between the two conditions ( Fig 4C ) . CoCl2 treatment resulted in up-regulation of 2 , 182 transcripts ( p≤0 . 001; FC ≤2 or ≥ 2 ) . Similarly 1882 transcripts were observed down-regulated due to CoCl2 treatment ( p≤0 . 001; FC ≤2 or ≥ 2 ) . A volcano plot for the differentially expressed genes in BJAB-CoCl2 cells compared to BJAB cells is also presented showing that expression of a large number of genes was clearly modulated ( Fig 4C ) . The top 10 up-regulated and top 10 down-regulated genes from the comparative group are provided ( Fig 4D ) . Analysis of the RNA sequencing data for the differential gene expression in BJAB-KSHV cells compared to BJAB cells resulted in detection of 357 up-regulated transcripts ( p≤0 . 001; FC ≤2 or ≥ 2 ) . Similarly , 233 transcripts were observed down-regulated in BJAB-KSHV cells compared to BJAB cells ( p≤0 . 001; FC ≤2 or ≥ 2 ) . A volcano plot for the differentially expressed genes in BJAB-KSHV cells compared to BJAB cells is shown ( Fig 5A ) . The top 10 up-regulated and top 10 down-regulated genes from the comparative group are provided ( Fig 5B ) . To further corroborate the transcriptional profiles of CoCl2 induced hypoxia or the combinatorial effects of hypoxia and KSHV infection , a comparative analysis of RNA sequencing data between BJAB and BJAB-KSHV-CoCl2 was performed . The results revealed a more enhanced effect of CoCl2 on transcription of host genes . Compared to 2182 transcripts up-regulated in BJAB-CoCl2 cells , 2560 transcripts were up-regulated in BJAB-KSHV-CoCl2 cells ( p≤0 . 001; FC ≤2 or ≥ 2 ) . Similarly , an enhanced effect on down-regulation of transcripts was also observed in BJAB-KSHV-CoCl2 cells where a total of 2 , 143 transcripts were observed down-regulated ( p≤0 . 001; FC ≤2 or ≥ 2 ) , compared to only 1 , 882 genes in BJAB-CoCl2 cells . A volcano plot for the differential gene expression of BJAB vs . BJAB-KSHV-CoCl2 cells is shown in S3A Fig . The top 10 up-regulated and top 10 down-regulated genes from this comparative group is also provided in S3B Fig . KSHV infection is known to stabilize HIF1α [16 , 20] . We wanted to determine which genes are common targets of KSHV infection , and CoCl2-induced hypoxia . We also analysed the data to identify synergistic activation or suppression activities linked to the combination of KSHV and hypoxia . A Venn diagram was prepared ( using Partek software ) for the differentially expressed genes ( p ≤0 . 001; FC ≤2 or ≥ 2 ) in BJAB-CoCl2 , BJAB-KSHV and BJAB-KSHV-CoCl2 cells compared to BJAB cells ( Fig 6A ) . Among the 357 transcripts observed up-regulated in BJAB-KSHV cells and 2 , 182 transcripts up-regulated in BJAB-CoCl2 cells , 160 transcripts were common ( Fig 6A ) . Similarly , among 233 down- regulated transcripts in BJAB-KSHV cell and 1 , 882 down-regulated transcripts in BJAB- CoCl2 cells , 60 transcripts were common ( Fig 6A ) . Interestingly , 105 transcripts out of 357 up-regulated in BJAB-KSHV cells were specific for KSHV . These transcripts were also up-regulated in BJAB-KSHV-CoCl2 cells ( Fig 6A ) . Among the 233 down-regulated genes in BJAB-KSHV cells , 59 were observed to be specific for KSHV . These genes were also down-regulated in BJAB-KSHV-CoCl2 cells ( Fig 6A ) . Intensity maps of common up-regulated and down-regulated genes are provided in S4 Fig . We wanted to know if transcription regulatory genes were common targets of CoCl2-induced hypoxia , and KSHV infection . Our analysis showed that the DNMTs , mainly DNMT3A and DNMT3B were two common targets for down-regulation induced by both hypoxia and KSHV infection . Real-time PCR analysis for these DNMTs followed by western blot analysis for protein levels showed similar results as observed from our RNA sequencing , although DNMT3A was clearly more dramatic in its suppression ( Fig 6C , 6D and 6E ) . 310 genes involved in glucose , fatty acids and amino acids metabolism were identified by reviewing genes involved in these processes ( S6 Table ) . This list was used to identify differentially expressed genes in each group when compared to BJAB cells . To optimize the number of genes differentially expressed in BJAB-KSHV , BJAB–CoCl2 or BJAB-KSHV-CoCl2 , the analysis stringency was maintained to allow for statistical significance ( p<0 . 05 ) . A Venn diagram was created for the common genes from the 310 metabolic regulated genes which were differentially expressed ( up-regulated or down-regulated ) in each group above . Compared to BJAB cells , a total of 16 metabolic genes were up-regulated in BJAB-KSHV cells ( Fig 7A ) . These up-regulated genes predominantly belonged to either glycolysis or the pentose phosphate pathway ( ALDOA , ENO1 , ENO2 , HK2 , PDK3 , PDP2 , PFKL and PGK1 , PRPS1 , PRPS2 and RPE ) , which supports a direct role for KSHV-infection in elevation of glycolysis . In addition , a subset of TCA cycle regulated genes ACLY , IDH3B , MDH1 and PCK1 were also up-regulated ( Fig 7C ) . Interestingly , these genes were exclusively restricted to the KSHV-positive background , and were not up-regulated in cells grown under CoCl2-induced hypoxic condition ( Fig 7A & 7C ) . Interestingly , in cells with elevated HIF1α due to CoCl2 treatment , activation of a different subset of metabolic genes from the glycolysis and TCA cycle pathways ( 15 genes for BJAB-CoCl2 and 16 genes for BJAB-KSHV-CoCl2 ) were identified . The up-regulated genes due to CoCl2 treatment were ALDOA , ALDOB , BPGM , PDPR and PGM3 ( glycolysis ) and DLST and IDH1 ( TCA cycle ) ( Fig 7A & 7C ) . In addition , a set of glycogen synthesis genes ( GSK3A , GSK3B , PHKG1 , PHKG2 and PYGM ) were also observed to be induced in CoCl2 treated cells ( Fig 7A & 7C ) . Interestingly , HIF1α stabilization due to CoCl2 treatment appeared to be dominant over KSHV infection . The expressions of these up-regulated genes were similar in both BJAB and BJAB-KSHV cells treated with CoCl2 ( Fig 7A & 7C ) . A second set of evidence showing glycolytic up-regulation by KSHV infection , or CoCl2-induced HIF1α was visible from the set of down-regulated genes in the TCA cycle ( IDH2 , PDHB , SDHA and SUCLG2 ) , and glycogen metabolism ( AGL , GB1 , PCK2 and PGM1 ) ( Fig 7B ) . These genes were down-regulated in both KSHV alone , and CoCl2 alone , in addition to their combination . CoCl2 treatment in fact showed an additional set of genes which were down-regulated compared to BJAB or BJAB-KSHV cells ( Fig 7B & 7C ) . The differentially expressed genes observed by RNA sequencing were further validated by real time PCR ( Fig 8A ) , using gene specific primers ( S5 Table ) . Among the metabolic genes , Transketolase ( TKT ) and Succinate dehydrogenase subunit A ( SDHA ) were down-regulated by either KSHV-infection or CoCl2 treatment . Interestingly , the expression of both TKT and SDHA were suppressed by KSHV infection and HIF1α stabilization ( Fig 8A and 8B ) . As vGPCR over-expression was associated with global transcriptional regulation and generation of reactive oxygen species ( ROS ) [27] , we hypothesized that expression of these genes can be a consequence of induced vGPCR in the hypoxic environment . To confirm the role of vGPCR in the down-regulation of TKT and SDHA , vGPCR knock down BJAB-KSHV cells were generated by lentivirus based transduction ( Fig 8C ) . Real-time expression of TKT and SDHA was analyzed in Sh-vGPCR BJAB-KSHV cells grown under normoxia or CoCl2 induced hypoxic conditions . The results showed clear involvement of vGPCR in regulating expression of these genes . ShControl cells showed a significant down-regulation of both TKT and SDHA expression in the hypoxic environment , however , Sh-vGPCR cells did not show any significant down-regulation ( Fig 8D & 8E ) . Importantly , we did not observe any strong up-regulation of TKT and SDHA in hypoxia . However , in Sh-vGPCR cells their expression was definitely increased compared to wild type KSHV indicating a role in transcription regulation under hypoxic conditions . To further corroborate a role for KSHV-encoded vGPCR in metabolic changes observed in the hypoxic environment , we investigated whether vGPCR knockout cells showed a reversal of the metabolic phenotype observed under hypoxia . The KSHV-bacmid clone containing vGPCR-Frame shift knock out mutant ( KSHV-vGPCR-FS-KO ) or vGPCR-Frame shift mutant reversed ( KSHV-vGPCR-FS-R ) KSHV [28] were transfected into HEK293T cells to generate stable lines . The transfected GFP positive cells were selected using hygromycin ( Fig 9A ) . To confirm that the frame shift mutation and revertant was maintained , genomic DNA from stable cells were isolated and the KSHV region encompassing the insertion site , PCR amplified and sequenced . The electropherogram showing the sequencing results confirmed the frame shift insertion and reversion to wild type ( Fig 9B ) . KSHV reactivation of the vGPCR-Frame shift mutant or revertant was induced by treatment with TPA and Butyric acid followed by standard virus purification . As expected , the vGPCR-Frame shift mutant showed a substantial decrease in lytic replication with significantly less yield in genome copies . To determine if vGPCR was a critical factor for the observed metabolic changes in hypoxia , PBMCs were infected with the vGPCR knockout and revertant KSHV virus and cells were subjected to hypoxic induction by treating with CoCl2 . KSHV infection was monitored by GFP signal and the induction of hypoxia was confirmed by western blot to detect HIF1α ( Fig 9D ) . The percentage of cells infected with KSHV was empirically calculated by counts of GFP signals in the cell population . The infection efficiency of PBMCs with KSHV was approximately 50% , although we observed slightly weaker GFP signal intensity from cells infected with the KSHV-vGPCR-FS-KO ( Fig 9C ) . Cells were grown in normoxic , or CoCl2 induced hypoxia for 48 hours followed by media collection and measurement of glucose uptake . The results suggest a clear role for KSHV-encoded vGPCR in contributing to the metabolic changes . The PBMCs infected with the vGPCR kockout KSHV showed a significantly lower glucose uptake compared to the revertant ( Wild type ) KSHV infected cells ( Fig 9E ) . As vGPCR has been shown to modulate transcriptional changes through the global signaling molecule i . e reactive oxygen species ( ROS ) , we wanted to determine if vGPCR mediated ROS had any role in the transcriptional regulation of genes which were differentially expressed in our study . We first determined the levels of reactive oxygen species in PBMCs infected with revertant ( wild type ) KSHV treated with or without ROS scavenger , superoxide dismutase ( SOD ) by DCFH-DA staining ( Fig 9E and 9F ) . As expected , the results showed a high level of reactive oxygen species in PBMCs infected by KSHV . The level was significantly lower in infected cells treated with SOD ( Fig 9F ) . RNA was isolated from these cells and reversed transcribed . The expression of TKT was determined by real-time PCR of cDNA . The results showed a reversal in the expression pattern of TKT upon SOD treatment , suggesting a possible role of vGPCR mediated ROS in transcriptional regulation of cellular genes .
Similar to infection with most of oncogenic viruses , KSHV infection leads to stabilization of HIFs in host cells by either preventing its degradation or by up-regulating its expression at the transcription level [24 , 29–33] . The stabilized HIF1α alone or in conjugation with host and viral factors modulates several physiologic pathways supporting survival and growth of the infected cells [8 , 24] . Further , stabilization of hypoxia inducible factors due to viral infection only partially mimic the in vitro experimental methods of inducing hypoxia in cell culture by growing the cells either under low oxygen or chemical induction by Cobalt Chloride ( CoCl2 ) /Deferoxamine mesylate ( DFO ) [34 , 35] . The stabilization of hypoxia inducible factor due to viral infection activated the HIF1α dependent pathways , whereas hypoxia due to low oxygen led to activation of several other energy associated pathways such as the AMPK dependent pathways [36–38] . Independent of the HIF1α stabilization mechanism , the interaction of stabilized hypoxia inducible factors with KSHV factors resulted in modulation of several pathways with impact on the host cell as well as the virus biology [16 , 18 , 39] . Hypoxia induces expression of the latency associated nuclear antigen ( LANA ) , the key viral factor responsible for attachment of viral episomal DNA to the host chromosome [15 , 40] . On the other hand hypoxia is known to induce lytic replication of KSHV as well as enhancing the reactivation potential of known chemical inducers TPA and Butyric acid [16 , 18] . LANA can be described as a bonafide oncogenic protein with the ability to degrade cellular tumor suppressors as well as activate oncogenes [41–43] . Therefore , we would predict an enhanced tumorigenic state of KSHV infected cells in the hypoxic environment . However , KSHV-encoded reactivation and transcriptional activator ( RTA ) is also a downstream target of HIF1α and so allows for the possibility that KSHV-positive cells grown in a hypoxic environment will be more susceptible to lytic reactivation . Independent of cell destiny , both the latent or lytic pathways require enhanced metabolic activities for generating their required macromolecular components . However , the exact contributors to induction of the hypoxic phenotype to KSHV-positive cells have not been fully explored . What are the differences in utilization of physiological pathways in KSHV-negative and KSHV-positive cells grown in hypoxia has not been completely explored , particularly in B-cell lineages . A huge challenge to investigating these questions is the limitation of available KSHV negative cell line controls . The comparative studies done previously for KSHV modulated pathways were performed by infecting cells of epithelial or endothelial origin while taking the parental cells as control[44] . However , for studies in B-cells , peripheral blood mononuclear cells were used [45] . Furthermore , the efficiency of KSHV infection remains a critical determinant in these studies [46] . In our current study , we used BJAB and BJAB-KSHV cells with the same genetic background [25] , for comparative analysis of differential metabolic signatures and associated mechanisms due to the change in transcription profiles . Characterization of BJAB-KSHV cells showed that it can be used as a model B-cell line for our comparative analysis . We first investigated the metabolic behavior of BJAB and BJAB-KSHV cells grown in 1% oxygen or CoCl2-induced hypoxia . The results suggested an enhanced glucose dependency and a cancer cell metabolic phenotype of high lactate release in BJAB-KSHV cells growing in hypoxia due to both low oxygen , as well as CoCl2 treatment as compared to their counter KSHV negative BJAB cells . The observed changes were not exclusive to long-term infected cells . The initial infection of PBMCs with KSHV can also induce a similar pattern of changes when grown under hypoxic conditions . Interestingly , hypoxia induction due to low oxygen concentration induced a much higher glucose dependency and lactate release compared to CoCl2-induced hypoxia . Also , hypoxia due to low oxygen did not allow long-term survival of cells compared to hypoxia induced due to CoCl2 . A higher rate of cell death after 48 hours was observed in culture . As hypoxia induction due to CoCl2 treatment shared more physiological relevance with stabilized HIF1α in KSHV positive cells at least in terms of cell survival , the RNA sequencing experiment performed on BJAB-KSHV cells growing in normoxia or CoCl2 induced hypoxia led to identification of novel HIF1α targets encoded by KSHV . One such target , vGPCR , is a constitutive homolog of cellular GPCR [24] and has been implicated in up-regulation of pathways common to most cancer cells including MAP-kinase , and the angiogenic pathways [24] . Although , vGPCR is considered a lytic gene for KSHV reactivation [47] , its role in tumorigenesis is also reported due to its activation of on MEK/ERK [24 , 47] . In addition , vGPCR is also known to inhibit transcription of other KSHV-encoded lytic genes [48 , 49] . It can modulate global expression of host genes as its induced expression in cells is associated with global changes in signaling due to increased levels of ROS [27] . The slightly higher expression of KSHV-encoded ORF2 ( homolog of cellular Dihydrofolate reductase ) suggest a shift towards the anabolic pathway for nitrogenous base synthesis , which is a pre-requisite for cell transformation and viral reactivation . The induced expression of vGPCR and its correlation with enhanced glucose uptake , as well as its suppression in ShHIF1α cells provided hints towards its role in metabolic changes in KSHV infected cells in hypoxia . The results showing differential expression of KSHV-encoded transcripts also provided information about the half life and stability of these KSHV-encoded transcripts . For example , despite being transcribed from the same regulatory region , the transcript levels of LANA , vCyclin and vFLIP showed differential abundance , as well as a different pattern of expression under varying conditions . Though , it is now a well known fact that the transcript abundance not only depends on its transcriptional generation but also on the rate of its decay as a consequence of either its half life or regulation by non-coding RNAs , it would be interesting to explore the mechanism behind the differential abundance of vCyclin and vFLIP specifically under hypoxic conditions . The analysis of global transcription changes in KSHV positive background and its similarity to those in KSHV negative cells grown in hypoxia suggested that HIF1α plays a role in KSHV infection in its associated pathology . Recently , a study designed to explore common transcriptional signatures for hypoxia and KSHV infection was performed using cells of different origins under low oxygen conditions . A convergence of targets due to KSHV infection and hypoxia was observed [50] . However , this study identified a limited number of genes differentially expressed under these conditions [50] . In combination with our data , we now provide a global picture of the role of hypoxia , KSHV infection , and their combinatorial effect on transcription of cellular genes regulated in the background of KSHV-infected cells . Further , it is possible that the role of HIF1α in B-cells , and other cell lineages including epithelial or endothelial cells may have differences in terms of transcripts or translated products . In B-cells , the oxygen partial pressure resembles that of blood supply . This is likely very different in solid tumors originated from epithelial or endothelial cells . In KSHV infected B-cells , induced HIF1α levels is mainly due to signaling modulated by KSHV-encoded antigens , whereas in KSHV-infected endothelial cells , the stabilization of HIF1α is likely due to the combined effects of signaling modulated by KSHV-encoded antigens in addition to low oxygen . In KSHV-infected endothelial cells , the hypoxic conditions may lead to activation of AMPK pathways due to low oxygen supply , which eventually affects ATP production through mitochondrial respiration and accumulation of AMP . The effect of induced HIF1α in B-cells may not activate the AMPK pathways . This study also provided information on the differences due to hypoxia induced due to low oxygen compared to CoCl2 , especially in terms of cell survival and growth . An analysis of the effects on transcription of metabolic genes shows the up-regulation of the glycolytic pathway due to KSHV infection , and that induced hypoxia can affect the same pathways but though different targets . However , shut-down of specific cellular genes was similar in hypoxia and in KSHV infection . While investigating possible cellular factor ( s ) involved in transcriptional regulation , analysis of global modulators such as DNMTs showed a significant down-regulation in their expression . This was mainly for DNMT3A and 3B due to CoCl2 induced hypoxia ( for both BJAB and BJAB-KSHV cells ) as seen in our RNA sequencing data sets . The expression of both DNMT3A and 3B was low in KSHV infected cells , however , the differences between them were not significant . Validation of the RNA sequencing results of DNMTs expression by real-time PCR confirmed suppression by both hypoxia and KSHV infection on DNMT3A , and 3B expression . Further greater suppression of at least DNMT3A was seen when hypoxia and KSHV infection were combined . The analysis of DNMTs expression is likely not the only explanation for the changes in expression of these large set of genes . However , it would be interesting to investigate the role of other global modulators including small non-coding RNAs on the expression of large set of genes in response to hypoxia or/and KSHV . These studies are ongoing with potential for elucidating additional mechanistic insights into the hypoxia-KSHV axis .
Peripheral blood mononuclear cells ( PBMCs ) from undefined and healthy donors were obtained from the Human Immunology Core ( HIC ) , University of Pennsylvania . The Human Immunology Core maintains approved protocols of Institutional Review Board ( IRB ) in which a Declaration of Helsinki protocols were followed and each donor/patient gave written , informed consent . BJAB ( KSHV-negative B-cell ) cells [51] were obtained from Elliot Kieff ( Harvard Medical School , Boston , MA ) originally purchased from American Type Culture Collection ( ATCC ) . KSHV-positive BJAB-KSHV cells [25] were obtained from Michael Lagunoff ( University of Washington , Seattle , WA ) . The KSHV-positive lymphoma-derived BC3 cell line was obtained from the ATCC . BJAB , BJAB-KSHV and BC3 cells were grown at 37°C/5% CO2 in RPMI medium containing 7% bovine growth serum ( BGS ) and penicillin ( 100 units/ml ) / streptomycin ( 0 . 1mg/ml ) . BJAB-KSHV cells were maintained with additional selection using puromycin ( 2μg/ml ) . Human Embryonic Kidney cell line ( HEK 293T ) was obtained from Jon Aster ( Brigham and Womens Hospital , Boston , MA ) and grown in DMEM medium containing 5% BGS with antibiotics at the above concentration . Cobalt chloride stock ( 100mM ) was prepared in water . Hypoxia was induced by adding CoCl2 at a final concentration of 100 μM or by growing the cells in a hypoxia chamber with 1% oxygen . ShControl and ShHIF1α lentiviruses were generated by transfection of HEK293T cells with transfer plasmids and third generation packaging and envelop plasmids as descried earlier [52] . In brief , HEK293T cells were grown in 100 mm cell culture dish at 40–60% confluency . 10 μg of transfer plasmids in combination with packaging and envelop plasmids were transfected by calcium phosphate method . After initial discard of culture medium containing transfection mix , the supernatants were collected between 24–96 hours at 12 hours intervals . The supernatants were filtered through a 0 . 45 μm syringe filter and lentiviruses were concentrated by ultracentrifugation at 23 , 500 rpm for 2 hours [45 , 52] . The pelleted lentiviruses were resuspended in 1ml of complete medium and frozen until used for transduction . Transduction was performed by mixing cells with resuspended lentiviral stock in the presence of 8μg/ml polybrene . 48 hours post transduction , cells were selected in the presence of 2μg/ml puromycin . The pCEFL-vGPCR [53] construct was a kind gift from Enrique A . Mesri ( University of Miami Miller School of Medicine , Miami , FL ) . The vGPCR-Knock Out ( vGPCR-KO , Frame shift Mutant ) , vGPCR-Knock Out Reversed Bacmid clones , Sh-vGPCR lentiviral and control plasmids [28] were provided by John Nicholas ( John Hopkins Bloomberg School of Public Health , Baltimore , MD ) . Large scale maxiprep for vGPCR-Knock Out ( Frame shift Mutant ) and vGPCR-Knock Out Reversed Bacmids were prepared using Luria broath culture and Qiagen large construct kit ( Qiagen Inc . , Hilden , Germany ) . Electroporation of HEK-293T cells were done in 4mm cuvette on Biorad Gene Pulser Xcell electroporation system . The positive clones of HEK293T-BAC-KSHV-vGPCR-KO and HEK293T-BAC-KSHV-vGPCR-KO reversed were selected in 100 μg/ml hygromycin . KSHV virion stocks were prepared from the KSHV positive BC3 cells as previously described [45] . In brief , KSHV reactivation was mediated by adding TPA ( to a final concentration of 20ng/ml ) and Butyric acid ( final concentration 3mM ) and the cells were placed in an incubator at 37°C/5% CO2 for 5 days . The cells and supernatant were pelleted down by centrifugation for 30 minutes at 3 , 000 rpm . The supernatant was filtered through a 0 . 45 micron syringe filter . The cell pellets were resuspended in 10 ml of 1X PBS followed by 3X freeze/thaw cycle . The lysed cells were again collected by centrifugation for 30 minutes at 3000 rpm and the supernatants were filtered through 0 . 45 micron syringe filter and pooled . The filtrates were subjected to ultracentrifugation at 23 , 500 rpm for 2 hours to collect the KSHV virions . Infection of PBMCs was carried out in the presence of 8 μg/ml polybrene as described earlier [45] . RNA isolation was performed according to standard method of phenol chloroform extraction using TRizol reagent ( Ambion , Grand Island , NY ) . 2μg of total RNA was used to synthesize cDNA by random priming method using Superscript cDNA synthesis kit ( Applied Biosystems Inc . , Foster City , CA ) . 1 μl of 10 times diluted cDNA was used for real-time PCR using Power SYBR green PCR reagent ( Applied Biosystems Inc . , Carlsbad , CA ) using a Step One Plus or Quant Studio system ( Applied Biosystems Inc . , Carlsbad , CA ) . All real-time PCR assays were performed in duplicates , with one experimental repeats for each gene . Real-time PCR for a select set of genes were performed at least in triplicate . The differences in fold change were calculated by delta delta CT method using default parameter settings . DNA from cells were isolated using Blood & cell culture DNA isolation mini kit ( Qiagen Inc . , Hilden , Germany ) . Gel eluted PCR product was used for DNA sequencing ( DNA sequencing facility , Department of Genetics , University of Pennsylvania ) using both forward and reverse primers . Short tandem repeat ( STR ) profiling for BJAB and BJAB-KSHV cells was performed using GenePrint 10 kit ( Promega Inc , Madison , WI ) at the genomics core facility , Department of Genetics , University of Pennsylvania . Glucose concentration available in cell culture medium was estimated using the hexokinase measurement kit ( Sigma Inc . , St . Louis , MO ) . The amount of glucose uptake by cells was measured by subtracting the amount of available glucose from the total amount of glucose available in fresh medium . The value of glucose uptake was finally normalized per million of cells . The amount of lactate present in the medium was estimated using the lactate estimation kit ( BioVision Inc . , Milpitas , CA ) . In brief , a standard curve for the known amount of lactate was created . A pilot experiment was performed using 1 μl and 10 μl of 1:10 times diluted cell culture medium from control or treated cells to estimate the range of lactate released . Intracellular reactive oxygen species was determined by fluorescence of the cell permeable dye DCFH-DA . DCFH-DA stock solution at a concentration of 5 mM was prepared in DMSO . In brief , cells were stained with 5μM DCFH-DA for 30 minutes in complete media at 37°C in cell culture incubator followed by collection of cells at 1500 rpm for 5 minutes . The cells were counted and equal numbers were used to determine fluorescence on a micro plate reader ( Molecular Devices , Sunnyvale , CA ) . Cell lysates were separated on SDS-polyacrylamide gels followed by wet transfer to nitrocellulose membrane . 5% skimmed milk was used for blocking at room temperature for 1 hour with gentle shaking . Primary antibody against HIF1α ( Santa Cruz Biotechnology , Dallas , TX and Novus Biosciences , Littleton , CO ) , GAPDH ( Novus Biosciences , Littleton , CO ) , DNMT3A and DNMT3B ( Abcam , Cambridge , MA ) were incubated overnight at 4°C with gentle shaking followed by washing with TBST 3-times ( 5 minutes intervals ) . Probing with IR conjugated secondary antibodies was performed at room temperature for 1 hour followed by washing with TBST . Membranes were scanned using an Odyssey scanner ( LI-COR Inc . , Lincoln , NE ) for detection of bands . For confocal microscopy , 25 , 000 cells were semi-dried on 8-well glass slides followed by fixation in 4% paraformaldehyde . Combined permealization and blocking was performed in 1XPBS containing 0 . 3% Triton X-100 and 5% goat serum followed by washing with 1X PBS . Anti-LANA antibody ( purified ascites ) was diluted in PBS containing 1% BSA and 0 . 3% triton X-100 and was incubated overnight at 4°C . Slides were washed with PBS followed by incubation with Alexa448 conjugated anti-mouse secondary antibody . DAPI staining was performed for 15 minutes at room temperature followed by washing and mounting . Images were captured by confocal microscope ( Olympus , Lambertville , NJ ) . Total RNA was isolated by standard phase extraction using TRizol reagent ( Ambion Inc . , Grand Island , NY ) . Isolated RNA was analyzed for quantity and quality using a Bio-photometer ( Eppendorf Inc . , Hamburg , Germany ) . Samples for RNA sequencing were prepared using Illumina RNA sequencing sample prep kit . The indexed ready to run samples were run on Illumina platform at the University of Washington ( Core services ) . The reads for sequencing data were aligned with KSHV and the human genome . All the RNA sequencing experiments were performed in duplicates . The fold change expression and statistical relevance of the differential gene expression was calculated by CLC bio software ( Qiagen Inc . , Hilden , Germany ) . The differential gene expression was represented by volcano plot using R-software . Intensity plots for the fold-change expression of a selected set of genes were created using Partek software ( Partek Inc . , St . Louis , MO ) . | Hypoxia inducible factors ( HIFs ) play a critical role in survival and growth of cancerous cells , in addition to modulating cellular metabolism . Kaposi’s sarcoma associated herpesvirus ( KSHV ) infection stabilizes HIFs . Several factors encoded by KSHV are known to interact with up or downstream targets of HIFs . However , the process by which KSHV infection leads to stabilized HIF1α and modulation of the cellular metabolism is not understood . Comparative RNA sequencing analysis on cells with stabilized hypoxia inducible factor 1 alpha ( HIF1α ) , of KSHV negative or positive cells led to identification of changes in global and metabolic gene expression . Our results show that hypoxia induces glucose dependency of KSHV positive cells with high glucose uptake and high lactate release . KSHV-encoded vGPCR was identified as a novel target of HIF1α regulation and a major viral antigen involved in metabolic reprogramming . Silencing of HIF1α rescued the hypoxia associated phenotype of KSHV positive cells . Analysis of the host transcriptome identified several common targets of hypoxia and KSHV-encoded factors , as well as other synergistically activated genes belonging to cellular metabolic pathways . This study showed unique , common and the synergistic effects of both HIF1α and KSHV-encoded proteins on metabolic reprogramming of KSHV-infected cells in hypoxia . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] | [
"sequencing",
"techniques",
"cell",
"physiology",
"carbohydrate",
"metabolism",
"medicine",
"and",
"health",
"sciences",
"pathology",
"and",
"laboratory",
"medicine",
"oxygen",
"pathogens",
"microbiology",
"cell",
"metabolism",
"glucose",
"metabolism",
"viruses",
"dna",
"viruses",
"hypoxia",
"molecular",
"biology",
"techniques",
"rna",
"sequencing",
"herpesviruses",
"research",
"and",
"analysis",
"methods",
"artificial",
"gene",
"amplification",
"and",
"extension",
"medical",
"microbiology",
"gene",
"expression",
"microbial",
"pathogens",
"chemistry",
"kaposi's",
"sarcoma-associated",
"herpesvirus",
"molecular",
"biology",
"biochemistry",
"chemical",
"elements",
"cell",
"biology",
"polymerase",
"chain",
"reaction",
"viral",
"pathogens",
"genetics",
"biology",
"and",
"life",
"sciences",
"physical",
"sciences",
"metabolism",
"organisms"
] | 2018 | Metabolic reprogramming of Kaposi’s sarcoma associated herpes virus infected B-cells in hypoxia |
Instrumental responses are hypothesized to be of two kinds: habitual and goal-directed , mediated by the sensorimotor and the associative cortico-basal ganglia circuits , respectively . The existence of the two heterogeneous associative learning mechanisms can be hypothesized to arise from the comparative advantages that they have at different stages of learning . In this paper , we assume that the goal-directed system is behaviourally flexible , but slow in choice selection . The habitual system , in contrast , is fast in responding , but inflexible in adapting its behavioural strategy to new conditions . Based on these assumptions and using the computational theory of reinforcement learning , we propose a normative model for arbitration between the two processes that makes an approximately optimal balance between search-time and accuracy in decision making . Behaviourally , the model can explain experimental evidence on behavioural sensitivity to outcome at the early stages of learning , but insensitivity at the later stages . It also explains that when two choices with equal incentive values are available concurrently , the behaviour remains outcome-sensitive , even after extensive training . Moreover , the model can explain choice reaction time variations during the course of learning , as well as the experimental observation that as the number of choices increases , the reaction time also increases . Neurobiologically , by assuming that phasic and tonic activities of midbrain dopamine neurons carry the reward prediction error and the average reward signals used by the model , respectively , the model predicts that whereas phasic dopamine indirectly affects behaviour through reinforcing stimulus-response associations , tonic dopamine can directly affect behaviour through manipulating the competition between the habitual and the goal-directed systems and thus , affect reaction time .
A very basic assumption in theories of animal decision making is that animals possess a complicated learning machinery that aims for maximizing rewards and minimizing threats to homeostasis [1] . The primary question within this framework is then how the brain , constrained by computational limitations , uses past experiences to predict rewarding and punishing consequences of possible responses . The dual-process theory of decision making proposes that two distinct brain mechanisms are involved in instrumental responding: the “habitual” , and the “goal-directed” systems [2] . The habitual system is behaviourally defined as being insensitive to outcome-devaluation , as well as contingency-degradation . For example , in the experimental paradigm of outcome-devaluation , the animal is first trained for an extensive period to perform a sequence of actions for gaining access to a particular outcome . The outcome is then devaluated by being paired with an aversive stimuli ( conditioned taste-aversion ) , or by over-consumption of that outcome ( sensory-specific satiety ) . The critical observation is that in the test phase , which is performed in extinction , the animal continues responding for the outcome , even though it is devaluated . The goal-directed process , on the other hand , is defined as being sensitive to outcome-devaluation and contingency-degradation . This behavioural sensitivity is shown to emerge when the pre-devaluation training phase is limited , rather than extensive Adams [3] . Based on these behavioural patterns , two different types of associative memory structures are proposed for the two systems . The behavioural autonomy demonstrated by the habitual system is hypothesized to be based on the establishment of associations between contextual stimuli and responses ( S-R ) , whereas representational flexibility of the goal-directed system is suggested to rely on associations between actions and outcomes ( A-O ) . A wide range of electrophysiological , brain imaging , and lesion studies indicate that different , and topographically segregated cortico-striato-pallido-thalamo-cortical loops underlie the two learning mechanisms discussed above ( see [4] for review ) . The sensorimotor loop , comprising of glutamatergic projections from infralimbic cortices to dorsolateral striatum , is shown to be involved in habitual responding . In addition , phasic activity of dopamine ( DA ) neurons , originating from midbrain and projecting to different areas of the striatum is hypothesized to carry a reinforcement signal , that is shown to play an essential role in the formation of S-R associations . The associative loop , on the other hand , is proposed to underlie goal-directed responding . Some critical components of this loop include dorsomedial striatum and paralimbic cortex . The existence of two parallel neuronal circuits involved in decision making arises the question of how the two systems compete for taking control over behaviour . Daw and colleagues , proposed a reinforcement learning model in which , the competition between the two systems is based on the relative uncertainty of the systems in estimating the value of different actions [5] . Their model can explain some behavioural aspects of interaction between the two systems . A critical analysis of their model is provided in the discussion section . In this paper , based on the model proposed in [5] , and using the idea that reward maximization is the performance measure of the decision making system of animals , we propose a novel , normative arbitration mechanism between the two systems that can explain a wider range of behavioural data . The basic assumption of the model is that the habitual system is fast in responding , but inflexible in adapting its behavioural strategy to new conditions . The goal-directed system , in contrast , can rapidly adapt its instrumental knowledge , but is considerably slower than the habitual system in making decisions . In the proposed model , not only the two systems seek to maximize the accrual of reward -by different algorithms- , but the arbitration mechanism between them is also designed in a way to exploit the comparative advantages of the two systems in value estimation . As a direct experimental observation for supporting the assumptions of the model , it has been reported classically that when rats traverse a T-maze to obtain access to an outcome , at the choice points , they pause and vicariously sample the alternative choices before committing to a decision [6]–[8] . This behaviour , called “vicarious trial-and-error” ( VTE ) , is defined by head movements from one stimulus to another at a choice point , during simultaneous discrimination learning [9] . This hesitation- and conflict-like behaviour is suggested to be indicative of deliberation or active processing by a planning system [6] , [7] , [10] , [11] . Important for our discussion , it has been shown that after extensive learning , VTE frequency declines significantly [6] , [12] , [13] . This observation is interpreted as a transition of behavioural control from the planning system to the habitual one , and shows difference in the decision-time between habitual and goal-directed responding [14] . Beside being supported by the VTE behaviour , the assumption about the relative speed and flexibility of the two systems allows the model to explain some behavioural data on choice reaction time . The model also predicts that whereas phasic activity of DA neurons indirectly affects the arbitration through intervening in habit formation , tonic activity of DA neurons can directly influence the competition by modulating the cost of goal-directed deliberation .
Reinforcement learning ( RL ) is learning how to establish different types if instrumental associations for the purpose of maximizing the accrual of rewards [15] . In the RL framework , stimuli and responses are referred to as states and actions , respectively . An RL agent perceives its surrounding environment in the form of a finite set of states , , in each of which , one action among a finite set of actions , , can be taken . The dynamics of the environment can be formulated by a transition function and a reward function . The transition function , denoted by , represents the probability of reaching state after taking action at state . The reward function , , indicates the probability of receiving reward , by executing action at state . This structure , known as the Markov Decision Process ( MDP ) , can be demonstrated by a 4-tuple , . At each time-step , , the agent is in a certain state , say , and makes a choice , say , from several alternatives on the basis of subjective values that it has assigned to them through its past experiences in the environment . This value , denoted by , is aimed to be proportional to the sum of discounted rewards that are expected to be received after taking action onward: ( 1 ) is the discount factor , which indicates the relative incentive value of delayed rewards compared to immediate ones . Model-free and model-based RL , are two variants of reinforcement learning with behavioural characteristics similar to the habitual and goal-directed systems , respectively [5] . These two variants are in fact two different mechanisms for estimating the -function of equation 1 , based on the feedbacks , , that the animal receives from the environment through learning . In temporal difference RL ( TDRL ) , which is an implementation of model-free RL , a prediction error signal , , is calculated each time the agent takes an action and receives a reward from the environment . This prediction error is calculated by comparing the prior expected value of taking that action , , with its realized value after receiving reward , : ( 2 ) is the maximum value of all feasible actions available at . The prediction error signal is hypothesized to be carried by the burst firing of midbrain dopamine neurons . This signal can be used to update the estimated value of actions: ( 3 ) is the learning rate , representing the degree to which the prediction error adjusts the -values of the habitual system . Assuming that the reward and transition functions of the environment are stationary , equations 2 and 3 will lead the -values to eventually converge through learning to the expected sum of discounted rewards . Therefore , after a sufficiently long learning period , the habitual system will be equipped with the instrumental knowledge required for taking the optimal behavioural strategy . This optimal decision making is achievable without the agent knowing the dynamics of the environment . This is why this mechanism is known as model-free reinforcement learning . The gradual convergence of -values to their steady levels , leads the habitual system toward being insensitive to sudden changes in the environment's dynamics , such as outcome-devaluation and contingency degradation . Instead , as all the information required for making a choice between several alternatives is cached in S-R associations through the course of learning , the habitual responses can be made within a short interval after the stimulus is presented . Instead of keeping and updating point estimations , by using Kalman reinforcement learning [16] , the habitual system in our model keeps probability distributions for the -values of each state-action pair ( See Methods for mathematical details ) . These probability distributions contain substantial information that will be later used for arbitration between the habitual and the goal-directed systems . In contrast to the habitual process , the value estimation mechanism in a model-based RL is based on the transition and reward functions that the agent has learned through past experiences [5] , [15] . In fact , through the course of learning , the animal is hypothesized to learn the causal relationship between various actions and their outcomes , as well as the incentive value of different outcomes . Based on the former component of the environment's dynamics , the goal-directed system can deliberate the short-term and long-term consequences of each sequence of actions . Then by using the learned reward function , calculating the expected value for each action sequence will be possible . Letting denote the value of each action calculated by this method , the recursive value-iteration algorithm below can compute it ( See Methods for algorithmic details ) : ( 4 ) Due to employing the estimated model of the environment for value estimation , the goal-directed system can rapidly revise the estimated values after an environmental change , as soon as the transition and reward functions are adapted to the new conditions . This can explain why the goal-directed system is sensitive to outcome-devaluation and contingency-degradation [5] . But according to this computational mechanism , one would expect the value estimation by the goal-directed system to take a considerable amount of time , as compared to the habit-based decision time . The difference in speed and accuracy of value estimation by the habitual and goal-directed processes is the core assumption of the arbitration mechanism proposed in this paper , that allows the model to explain a set of behavioural and neurobiological data . If we assume for simplicity that the goal-directed system is always perfectly aware of the environment's dynamics , then it can be concluded that this system has perfect information about the value of different choices at each state . This is a valid assumption in most of the experimental paradigms considered in this paper . For example , in outcome-devaluation experiments , due to the existence of a re-exposure phase between training and test phases , the subjects have the opportunity to learn new incentive values for the outcomes . Although the goal-directed system , due to its flexible nature , will always have “more accurate” value estimations compared to the habitual system , the assumption of having “perfect” information might be violated under some conditions ( like reversal learning tasks ) . This violation will naturally lead to some irrational arbitrations between the systems . Thus , the advantage of using the goad-directed system can be approximated by the advantage of having perfect information about the value of actions . But this perfect information can be extracted from transition and reward functions at the cost of losing time; a time which could be instead used for taking rapid habitual actions and thus , receiving less rewards in magnitude , but more in frequency . This trade-off is the essence of the arbitration rule between the two systems that we propose here . In other words , we hypothesize that animals balance the benefits of deliberations against their cost . Its benefit is proportional to the value of having perfect information , and its cost is equal to the potential reward that could be acquired during the time that the organism is waiting for the goal-directed system to deliberate . As illustrated schematically in Figure 1 , at each time-step , the habitual system has an imperfect estimate for the value of each action in the form of a distribution function . Using these distribution functions , the expected benefit of estimating the value of each action by the goal-directed system is computed ( see below ) . This benefit , called “value of perfect information” , can be denoted by . The cost of deliberation , denoted by , is also computed separately ( See below ) . Having the cost and benefit of deliberation for each action , if the benefit is greater than the cost , i . e . , the arbitrator will decide to run the goal-directed system for estimating the value of action ; otherwise , the value of action that will be used for action selection will be equal to the mean of the distribution function cached in the habitual system for that action . Finally , based on the estimated values of different actions that have been derived from either of the two instrumental systems , a softmax action selection rule , in which the probability of choosing each action increases exponentially with its estimated value , can be used ( See Methods ) . Upon executing the selected action and consequently receiving a reward and entering a new state , both the habitual and goal-directed systems will update their instrumental knowledge for future exploitations . Based on the decision theoretic ideas of “value of information” [17] , a measure has been proposed in [18] for information value in the form of expected gains in performance , resulted from improved policies if perfect information was available . This measure , which is computed from probability distributions over the -value of choices , is used in the original paper for proposing an optimal solution for the exploration/exploitation trade-off . Here , we use the same measure for estimating the benefit of goal-directed search . To see how this measure can be computed , assume that the animal is in the state , and one of the available actions is , with the estimated value assigned to it by the habitual system . At this stage , we are interested to know how much the animal will benefit if it understands that the true value of actions is equal to , rather than . Obviously , any new information about the exact value of an action is valuable only if it improves the previous policy of the animal that was based on . This can happen in two scenarios: ( a ) when knowing the exact value signifies that an action previously considered to be sub-optimal is revealed to be the best choice , and ( b ) when the new knowledge shows that the action which was considered to be the best , is actually inferior to some other actions . Therefore , the gain of knowing that the true value of is can be defined as [18]: ( 5 ) and are the actions with the best and second best expected values , respectively . In the definition of the gain function , the first and the second rules correspond to the second and the first scenarios discussed above , respectively . According to this definition , calculating the gain function for each choice requires knowing the true value of that state-action pair , , which is unavailable . But , as the habitual system is assumed to keep a probability distribution function for the value of actions , the agent has access to the probability of possible values of . Using this probability distribution of , the animal can take expectation over the gain function to estimate the value of perfect information ( ) : ( 6 ) Intuitively , and crudely speaking , the value of perfect information for an action is somehow proportional to the overlap between the distribution function of that action and the distribution function of the expectedly best action . Exceptionally , for the case of the expectedly best action , the signal is proportional to the overlap between its distribution function and the distribution function of the expectedly second best action . It is worth to emphasize that for the calculation of signals , the goal-directed system has in no way been involved and instead , all the necessary information has been provided by the habitual process . The signal for an action expresses the degree to which having perfect information about that action , i . e . knowing its true value , results in policy improvement and thus , is indicative of the benefit of deliberation . It is worth mentioning that computing the integral proposed in equation 6 is shown to have a closed form equation [18] and thus , the integral doesn't need to be actually taken . Therefore , assuming that the time needed for evaluating is considerably less than that of running the goal-directed system is plausible . For computing the cost of deliberation , on the other hand , assuming that deliberation about the value of each action takes a fixed time , , the cost of deliberation can be quantified as ; where is the average rate of reward per time unit . Average reward can be interpreted as the opportunity cost of latency in responding to the environmental stimuli [19] . It means that when the average reward has a high value , every second in which a reward is not obtained is costly . Average reward can be computed as an exponentially-weighted moving average of obtained rewards: ( 7 ) The arbitration mechanism proposed above , is an approximately optimal trade-off between speed and accuracy of responding . This means that given that the assumptions are true , the arbitration mechanism calls or doesn't call the goal-directed system , based on the criterion that sum of discounted rewards , as defined in equation 1 , should be maximized [See Methods for optimality proof] . The most challenging assumption , as mentioned before , is that the goal-directed system is assumed to have perfect information on the value of choices . As some cases that challenge the validity of this assumption one could mention the cases where only the goal-directed system is affected ( for example after receiving some verbal instructions by the subject ) . Clearly , the cached values in the habitual system and thus the signal will not be affected under such treatments , though the real accuracy that the goal-directed system has in estimating values has changed .
First discovered by Adams [3] and later replicated in a lengthy series of studies [20]–[23] , it has been shown that the effect that the devaluation of outcome exerts on the animal's responses depends upon the extent of pre-devaluation training; i . e . responses are sensitive to outcome devaluation after moderate training , whereas overtraining makes responding insensitive to devaluation . To check the validity of the proposed model , the model has been simulated in a schedule analogous to those used in the above mentioned experiments . The formal representation of the task , which was first suggested in [5] , is illustrated in Figure 2 . As the figure shows , the procedure is composed of 3 phases . The agent is first placed in an environment where pressing the lever ( ) followed by entering the food magazine ( ) results in obtaining a reward with the magnitude of one; but magazine entry before lever press , or pressing the lever and not entering the magazine leads to no reward . As the task is supposed to be cyclic , after performing each chain of actions , the agent goes to the initial state and will start afresh ( Figure 2:A ) . After a certain amount of training in this phase , the food outcome is devalued by being paired with poison , which is aversive with magnitude of one ( equivalently , its reward is equal to -1 ) ( Figure 2:B ) . Finally , to assess the effect of devaluation , the performance of the agent is measured in extinction , i . e . in the absence of any outcome ( neither appetitive , nor aversive ) , in order to avoid the instrumental associations acquired during training from being affected in the test phase ( Figure 2:C ) . The behavioural results , as illustrated in Figure2:D , show that behavioural sensitivity to goal-devaluation depends on the extent of pre-devaluation training . In the moderate training case , the rate of responding has significantly decreased after devaluation , which is an indicator of goal-directed responding . However , after extensive training , no significant sensitivity to devaluation of the outcome is observed , implying that responding has become habitual . Through numerical simulation , homogeneous agents , i . e . agents with equal free parameters of the model , have carried out the experimental procedure under two scenarios: moderate vs . extensive pre-devaluation training . The only difference between the two scenarios is in the number of training trials in the first phase of the schedule: 40 trials for the moderate , and 240 trials for the extensive training scenario . The results are illustrated separately for these two scenarios in Figure 3 . It must be noted that since neither the “lever-press” nor the “enter-magazine” actions are performed by the animal during the devaluation phase , the habitual knowledge remains intact in this period; i . e . the habitual system is not simulated during the devaluation period . Devaluation is assumed to only affect the reward function , used by the goal-directed system . Figure 3:A and G show that at the early stages of learning , the signal has a high value for both of the actions , and , at the initial state , . This indicates that due to initial ignorance of the habitual system , knowing the exact value of both of the actions will greatly improve the agent's behavioural strategy . Hence , the benefit of deliberation is more than its cost , . By obtaining a reward , the signal elevates gradually . Concurrently , as the -values estimated by the habitual process for the two actions converge to their real values through learning , the difference between them increases ( Figure 3:D and J ) . This increase leads to the overlap between the distribution functions over the two actions becoming less and less ( Figure 3:E and K ) and consequently , the signal decreasing gradually . Now by focusing on the moderate training scenario , it is clear that when devaluation has occurred at the trial number 40 , the signals have not yet become less than ( Figure 3:A ) . Thus , the actions have been goal-directed at the time of devaluation and hence , the agent's responses have shown a great sensitivity to devaluation at the very early stages after devaluation; i . e . the probability of choosing action has sharply decreased to 50% , which is equal to that of action ( Figure 3:B and F ) . Figure 3:C also shows that in the moderate training scenario , deliberation time has always been high; indicating that actions have always been deliberated using the goal-directed system . In contrast to the moderate training scenario , the signal is below at the time of devaluation in the extensive training scenario ( Figure 3:G ) . This means that at this point of time , the cost of devaluation has exceeded its benefit and hence , actions are chosen habitually . This can be seen in Figure 3:I , where deliberation time has reached zero after almost 100 training trials . As a consequence , the agent's responses have not sharply changed after devaluation ( Figure 3:H and L ) . Because the test has been performed in extinction , the average reward signal has gradually decreased to zero after devaluation and concurrently , the signal has slowly raised again , due to the reduction of the difference between the -values of the two choices ( Figure 3:J ) and so , the augmentation of the overlap between their distribution functions . At the point that has exceeded , the agent's responses have become goal-directed again and so , deliberation time has boosted ( Figure 3:I ) . Consistently , the rate of selection of each of the two choices has been adapted to the post-devaluation conditions ( Figure 3:H ) . In a nutshell , the simulation of the model in these two scenarios is consistent with the behavioural observation that moderately trained behaviours are sensitive to outcome devaluation , but extensively trained behaviours are not . Moreover , the model predicts that after extensive training , deliberation time declines; a prediction that is consistent with the VTE behaviour observed in rats [6] . Furthermore , the model predicts that deliberation time increases with a lag after devaluation in the extensive training scenario , whereas it remains unchanged before and after devaluation in the moderate training scenario . Just for the sake of more clarification , the reason that the mean value of in Figures 3:E and K is above zero is because of the cyclic nature of the task , i . e . by taking action at state , the agent goes back to the same state , which might have a positive value . The focus of the previous section was on simple tasks with only one response for each outcome . In another class of experiments , the development of behavioural autonomy has been assessed in more complex tasks where two different responses produce two different outcomes [21] , [24]–[26] . Among those experiments , to the best of our knowledge , it is only in the experiment in [26] that the two different choices ( and ) are concurrently available and hence , the animal is given a choice between the two responses ( Figure 4:A ) . In the others , the two different responses are trained and also tested in separate sessions and so , their schedules are not compatible with the requirements of the reinforcement learning framework that is used in our model . In [26] , rats received extensive concurrent instrumental training in a task where pressing the two different levers produces different types of outcomes: food pellets and sucrose solution . Although the outcomes are different , they have equal reinforcing strength , in terms of the response rates supported by them . A task similar to that used in their experiment is formally depicted in Figure 4 . After extensively reinforcing the two responses , one of the outcomes was devalued by flavour aversion conditioning , as illustrated in Figure 4:B . Subsequently , given a choice between the two responses , the sensitivity of instrumental performance to this devaluation was assessed in extinction tests . The results of their experiment showed that devaluation reduced the relative performance of the response associated with the devalued outcome at the very early stage of the test phase , even after extensive training . Thus , it can be concluded that whatever the amount of instrumental training , S-R habits do not overcome goal-directed decision making when two responses with equal affective values are concurrently available . Simulating the proposed model in the task of Figure 4 has replicated this behavioural observation . As illustrated in Figure 5:A , initially , the signal for the two responses has a high value which gradually decreases over time as the variance of the distribution functions over the estimated values of the two responses decreases; meaning that the habitual process becomes more and more certain about the estimated values . However , due to the forgetting effect , i . e . the habitual system forgets very old samples and does not use them in approximating the distribution function , the variance of the distribution functions over the values of actions doesn't converge to zero , but instead , converges to a level higher than zero . Moreover , because the strength of the two reinforcers is equal , as revealed in Figure 5:D , the distribution functions do not get divorced ( Figure 5:E ) . As a result of these two facts , the signal has converged at a level higher than ( Figure 5:A ) . This has led to the performance remaining goal-directed ( Figure 5:C ) and sensitive to devaluation of one of the outcomes; i . e . after devaluing the outcome of the action , its rate of selection has sharply decreased and instead , the probability of selecting has increased ( Figure 5:B and F ) . As it is clear from the above discussion , the relative strength of the reinforcers critically affects the arbitration mechanism in our model . In fact , the model predicts that when the affective values of the two outcomes are close enough to each other , the signal will not decline and hence , the behaviour will remain goal-directed and sensitive to devaluation , even after extensive training . But if the two outcomes have different reinforcing strength , then their corresponding distribution functions will gradually get divorced and thus , the signal will converge to zero . This leads to the habitual process taking control of behaviour and the performance becoming insensitive to outcome devaluation . This prediction is in contrast to the model proposed in [5] , in which the arbitration between the two systems is independent of the relative incentive values of the two outcomes . In fact , in that model , whether the value of an action comes from the habitual or the goal-directed system , only depends on the uncertainty of the two systems about their estimated values and thus , the arbitration between the two systems is independent of the estimated value for other actions . Using a classical reversal learning task , Pessiglione and colleagues have measured human subjects' reaction time by temporal decoupling of deliberation and execution processes [27] . Reaction time , in their experiment , is defined as the interval between stimulus presentation and the subsequent response initiation . Subjects are required to choose between two alternative responses ( “go” and “no-go” ) , as soon as one of the two stimuli ( “” and “” ) appear on the screen . As shown in Figure 6:A , at each trial , one of the two stimuli and will appear in random , and after the presentation of each stimuli , only one of the two actions results in a gain , whereas the other action results in a loss ( ) . The rule governing the appropriate response must be learned by the subject through trial and error . After several learning trials , the reward function changes without warning ( ) . This second phase is called the reversal phase . Finally , during the extinction phase , the “go” action never leads to a gain , and the appropriate action is to always choose the “no-go” response ( ) . To analyse the results of the experiments , the authors have divided each phase into two sequential periods: a “searching” period during which the subjects learn the reward function by trial and error , and an “applying” period during which the learned rule is applied . The results show that in the searching period of each phase , the subjects might choose either the right or the wrong choice , whereas during the applying period , they almost always choose the appropriate action . Moreover , as shown in Figure 6:B , the subjects' reaction time is significantly lower during the applying period , compared to the searching period . Figure 7 shows that our model captures the essence of experimental results reported in [27] . In fact , the model predicts that during the searching period , the goal-directed process is involved in decision making , whereas during the applying period , the arbitration mechanism doesn't ask for its help in value estimation . It should be noticed that the reaction time reported in [27] , is presumably the sum of stimulus-recognition time , deliberation time , etc . Thus , a fixed value , which is the sum of all the other processes involved in choice selection , must be added to the deliberation time computed by our model . One might argue that variations in reaction time in the mentioned experiment could also be explained by a single habitual system , by assuming that lack of sufficient learning induces a hesitation-like behaviour . For example , high uncertainty in the habitual system at the early stages of learning a task , or after a change is recognized , can result in a higher-than-normal rate of exploration [18] . Thus , assuming that exploration takes more time than exploitation , reaction time will be higher when the uncertainty of -values is high . However , as emphasized by the authors in [27] , uncertainty doesn't have any effect on the subject's movement time , but only on the reaction time . In fact , movement time remains constant through the course of the experiment . Movement time is defined as the interval between response initiation and submission of the choice . Since movement time is unaffected by the extent of learning , it is unlikely that variations in reaction time be due to a hesitation-like effect and thus , as an alternative , it can be attributed to involvement of deliberative processes . Moreover , such an explanation lacks a normative rationale for the assumption that exploration takes more time than exploitation . According to a classical literature in behavioural psychology , choice reaction time ( CRT ) is fastest when only one possible response is available , and as the number of alternatives increases , so does the response latency . Originally , Hick [28] found that in choice reaction time experiments , CRT increases in proportion to the logarithm of the number of alternatives . Later on , a wealth of evidence validated his finding ( e . g . , [29]–[35] ) , such that it became known as “Hick's law” . Other researchers [36] , [37] found that Hick's law holds only for unpracticed subjects , and that training shortens CRT . They also found that in well-trained subjects , there is no difference in CRT as the number of choices varies . In a typical CRT experiments , a certain number of stimuli and the same number of responses are used in each session of the experiment . Figure 8 shows the tree representation of an example task with four stimuli and four alternatives . In each trial , one of the four alternatives appears at random , and only one of the four responses results in a reward . As in the CRT experiments the subjects are provided with a prior knowledge about the appropriate response after the presentation of each stimuli , we assume that this declarative knowledge can be fed into and used by the goal-directed system in the form of transition and reward functions . Furthermore , subjects are asked to make true responses , and at the same time as fast as possible . Hence , since subjects know the structure of the task in advance , they show very high performance ( as defined by the rate of correct responses ) in the task . As demonstrated in Figure 9 , the behaviour of the model has replicated the results of CRT experiments: at the early stages of learning , the deliberation time increases as the number of choices increases , whereas after sufficient training , no difference in deliberation time can be seen . It must be mentioned that in contrast to behavioural data , our model predicts a linear correlation between the CRT and the number of alternatives , rather than a logarithmic function . Again , a fixed value characterizing stimulus-identification time must be added to the deliberation time computed by our model in order to reach the reaction time reported in the CRT literature . Since in CRT experiments a declarative knowledge about appropriate responses is provided to the subjects , they have a relatively high performance from the very beginning of the experiment . The proposed model can explain this behavioural characteristic due to the fact that at the early stages of the experiment , when the habitual system is totally ignorant about the task structure , the goal-directed system controls the behaviour and exploits the prior knowledge fed into it . Thus , a single habitual system cannot explain the performance profile of subjects , even though it might be able to replicate the reaction-time profile . For example , a habitual system that uses a winner-take-all neural mechanism for the -values of different choices to compete [38] , [39] also predicts that at the early stages of learning where the -values are close to each other , reaching a state that one action overcomes the others takes longer , compared to the later stages where the best choice has a markedly higher -value than other actions . Such a mechanism also predicts that at the early stages , if the number of choices increases , the reaction time will also increase . However , since feeding the subject's declarative knowledge into the habitual system is not consistent with the nature of this system , a single habitual system cannot explain the performance of subjects in Hick's experiment .
As mentioned , training-induced neuroplasticity in cortico-basal ganglia circuits is suggested to be mediated by dopamine ( DA ) , a key neuromodulater in the brain reward circuitry . Whereas phasic activity of midbrain DA neurons is hypothesized to carry the prediction error signal [40] , [41] , and thus imposes an indirect effect on behaviour through its role in learning the value of actions , the tonic activity of DA has shown to have a direct effect on behaviour . For example , DA agonists have been demonstrated to have an invigorating effect on a range of behaviours [42]–[46] . It is also shown that higher levels of intrastriatal DA concentration is correlated with higher rates of responding [47] , [48] , whereas DA antagonist or DA depletion results in reduced responsivity [49]–[53] . Based on these evidence , it has been suggested in previous RL models that tonic DA might report the average reward signal ( ) [19] . By adopting the same assumption , our model also provides a normative explanation for those mentioned experimental results , in terms of tonic DA-based variations in deliberation time . In the economic literature of decision theory , rational individuals make optimal choices based on their desires and goals [54] , without taking into account the time needed to find the optimal action . In contrast , models of bounded rationality are concerned with information and computational limitations imposed on individuals when they are encountered with alternative choices . Normative models of rational choice that take into account the time and effort required for decision making are known as rationality of type II . This notion emphasizes that computing the optimal answer is feasible , but not economical in complex domains . First introduced by Herbert Simon , it was argued that agents have limited computational power and that they must react within a reasonable amount of time [55] , [56] . To capture this concept , [57] used the Scottish word “satisficing” which means satisfying , to refer to a decision making mechanism that searches until an alternative that meets the agent's aspiration level criterion is found . In other words , the search process is continued until a satisfactory solution is found . Borrowed from psychology , aspiration level denotes a solution evaluation criterion that can be either static or context-dependent and acquired by experience . A similar idea has been taken by neuroscientists to explain the speed/accuracy trade-off , using signal detection theory ( see [58] for review ) . In this framework , the accumulated information gathered from a sequence of observations from a noisy evidence must reach a certain threshold , in order for the animals to convert the accumulated information into a categorical choice . If the threshold goes up , the accuracy increases . As in this case more information must be gathered to satisfy that increased level of accuracy , response latency will decrease . Simon's initial proposal has launched much attempt in both social science and computer science to develop models that sacrifice optimality in favor of fast-responding . The focus has been on complex uncertain environments , where the agent must respond in a limited amount of time . The answer given to this dilemma in social science is often based on a variety of domain-specific heuristic methods [59] , [60] in which , rather than employing a general-purpose optimizer , animals use a set of simple and hard-coded rules to make their decisions in each particular situation . In the artificial intelligence literature , on the other hand , the answer is often based on approximate reasoning . In this approach , details of a complex problem are ignored in order to build a simpler representation of the original problem . Finding the optimal solution of this simple problem will be feasible in an admissible amount of time [61] . To capture the concept of time limitation and to incorporate it into models of decision making , we have used the dual-process theory of decision making . The model we have proposed is based on the assumption that the habitual process is fast in responding to environmental stimuli , but is slow in adapting its behavioural strategies , particularly in environments with low stability . The goal-directed system , in contrast , needs time for deliberating the value of different alternatives by tracing down the decision tree , but , is flexible in behavioural adaptation . The rule for arbitrating between these two systems assumes that animals balance decision quality against the computational requirements of decision-making . However , the optimality of the arbitration rule is based on the strong assumption that the goal-directed decision process has perfectly learned the environmental contingencies . This assumption might be violated at some points , particularly at the very early stages of learning a new task . When both systems are totally ignorant of the task structure , although the habitual system is in desperate need of having perfect information ( high signal ) , the goal-directed system doesn't have any information to provide . Thus , deliberation not only doesn't improve animal's strategy , but leads to a waste of the time that could be used for blind exploration . Though , since the goal-directed system is very efficient in terms of exploiting the experienced contingencies , this sub-optimal behaviour of the model doesn't last long . More importantly , in real world situations , the goal-directed process seems to always have considerably more accurate information than the habitual system , even in environments that have never been explored before . This is because many environmental contingencies can be discovered by mere visual observation ( e . g . searching for food in an open field ) or verbal instruction ( as in the Hick's task discussed before ) , without any experience being required . Our model is in fact based on the previous computational model of the dual-process theory , proposed by Daw and colleagues [5] . After assigning model-free and model-based RL models to habitual and goal-directed systems , respectively , they suggest an uncertainty-based arbitration mechanism between the two systems . In their model , each of the two systems not only separately estimate a value for each certain action , but their uncertainties about that value-estimations are also computed . As in our model , lack of enough experiences in the environment results in uncertainty in the habitual system . The source of uncertainty in the goal-directed system , on the other hand , is ( 1 ) uncertainty in transition and reward functions , due to the lack of enough experiences and ( 2 ) “pruning” , which refers to incomplete consideration of the all parts of the decision tree when considering the consequences of alternative choices . The latter source of uncertainty is not explicitly modeled and instead , is captured by adding a noise to the estimated values . At any given point of time , both systems get involved in value and uncertainty estimation for all the available choices and when they have both finished , the system that is more certain about its estimation of the value of each action will determine the value of that action for action-selection . As a result of this arbitration rule , the goal-directed system is dominant at the early stages of learning; but after extensive learning , the habitual process will take control over behaviour . This happens because uncertainty of the habitual system decreases through the course of learning , whereas the goal-directed process remains uncertain due to the incomplete search of the decision tree ( the added noise ) . Thus , their model can explain the canonical observation in the experimental paradigm of outcome-devaluation ( Outcome-sensitivity after moderate , but not extensive training ) . The added noise to the goal-directed system in that model actually characterizes , in an adhoc way , all the computational constraints that the goal-directed system is confronted with; e . g . time constraint , working memory constraint , caloric needs , etc . It has also been pointed out in [5] , that the trade-off between behavioural flexibility and computational costs can be captured in a cost-benefit fashion . In this respect , the arbitration mechanism we have proposed in this paper is a variant of the model proposed in [5] , where only one of the computational constraint , i . e . deliberation time , is modeled in an explicit , cost-benefit account . Beside this noticeable behavioural harmony of that model with the current dual-process literature , it suffers from some deficiencies . These deficiencies arise from the fact that in that model , the goal-directed system ceaselessly searches for the optimal policy , regardless of the system that is controlling the behaviour . In contrast to this assumption , overtraining of a behaviour is shown to causes a transition in neural activity from the associative to the sensorimotor network; i . e . , whereas PFC and caudate nucleus are activated at the early stages of learning a new motor response , this activity shifts to motor cortices and putamen as the response becomes well-trained [62] , [63] . As a result , response latency in that model doesn't vary through learning . Of course , it should be mentioned that by adding the noise to the goal-directed system in order to model pruning , time-limitations have been implicitly incorporated into the model; but as this noise level remains fixed through learning , the involvement of the goal-directed system , and so the deliberation time , doesn't change even after extensive training . As mentioned before , the core idea that we have proposed here for arbitration between the two systems is that there should be a balance between speed and accuracy in responding . A similar idea has been previously used by Shah and Barto [64] , but in an evolving sensory representation framework . In the task that they have simulated , subjects must choose among the potential goals in each trial . However , the sensory representation of the true goal of each trial is weak at the beginning of the trial , and resolves gradually during the course of the trial [65] . The basic assumption of their model is that the planning system can select actions only when goal representation is fully resolved , but the habitual system can also use “uncertain” accumulated sensory information . At the early trials of learning the task , since the value of different choices is not learned by the habitual system yet , this system cannot choose among the choices within a considerable period of time . This is due to using a winner-take-all competition mechanism for action selection [38] , [39] . Thus , at the early trials , the sensory representation has enough time to be fully resolved and as a consequence of this , the planning system controls behaviour . However , after extended training , the habitual system can make a decision before the goal is fully identified , based on uncertain sensory information . Although both the model we proposed here and the model proposed in [64] use speed-accuracy trade-off for arbitration between the two systems , there is fundamental differences between them . Whereas the extra time needed by the planning system in is used for state recognition [64] , this time is used for deliberating the consequences of choices in our model . In fact , it is the process of state recognition that is time-consuming in their model , and not the process of deliberation . Due to this difference , the model of [64] can only be applied in cases where stimulus identification takes non-negligible time , which doesn't seem to be the case of the experiments addressed by our model . Changes in the animals' response rate has been previously explained in the reinforcement learning literature [19] , [66] . Importantly , in the model proposed by Niv et al . [19] , as in our model , animals make a balance between the cost and benefit of acting quickly . is the cost of responding after an interval . Thus , in their model , as in our model , the animal benefits from responding fast , because it loses less potential rewards . But as they do not model the goal-directed system , the cost of acting quickly in their model is due to an extra fatigue-like cost induced by responding fast , whereas this cost in our model is due to inaccurate and inflexible value estimations . We believe that both factors , influence the animals' response rate . But as a result of this fundamental difference , the two models have different behavioural predictions . In fact , the term in the model proposed in [19] refers to “execution time” , whereas in our model it refers to “reaction time” . Notice that reaction time is , by definition , the interval between stimulus presentation and performance initiation , whereas execution time ( movement time ) refers to the interval between response initiation and its finalization . Due to this difference , their model cannot explain any of the three experiments on reaction time that our model can: ( 1 ) VTE behaviour , ( 2 ) increase in reaction time as the number of choices increases , ( 3 ) decrease in reaction time after reversals , in the go/no-go task . Interestingly , by temporal decopulation of deliberation and execution , it has been shown in [27] that whereas reaction time has significantly decreased after reversal in a go/no-go task , the execution time has remained intact . As mentioned previously , one prediction of the competition mechanism proposed in this paper is that outcome sensitivity is dependent on the relative value of the choices that are concurrently available . That is , if the value of choices are sufficiently close together , the habitual system will remain uncertain about what the best choice is ( equivalent to high ) , even after extensive training . This will result in the informational gain of knowing the exact value of choices remaining high and thus , the goal-directed system staying dominant . Such a mechanism can explain the behavioural data reported in [26] . By contrast , the model predicts that in a concurrent schedule where the value of the two choices are sufficiently different , responding will eventually become habitual . This is because after extensive training , the habitual system will have sufficient information for choosing the better choice among the two , without needing the exact value of them; i . e . , without needing the goal-directed system . To our knowledge , this prediction is not tested yet . In this respect , the model has a different prediction from what the model proposed in [5] predicts . According to that model , the goal-directedness of responding doesn't depend on the relative value of choices and thus , it predicts that responding will remain goal-directed in concurrent schedules , whether the values of choices are equal or not . Another prediction of our model is that if the two choices in a concurrent schedule lead to a unique outcome , responses will remain sensitive to devaluation , regardless of the amount of instrumental training . This is because when the outcomes are identical , the values of the two choices that lead to it will be exactly the same . In fact , when the values of the two choices are equal , our model predicts that responding will remain goal-directed , whether the identity of the outcomes of choices are the same or not . However , in the model proposed in [5] , if the two outcomes are identical , it can be said that since fewer outcome values must be learned , the asymptotic uncertainties of the habitual system will decrease . Thus , according to that model , responding might become habitual or remain goal-directed after extensive training , depending on the parameters of the model . It should be mentioned that in an experiment by Holland [21] , sensitivity to devaluation is tested where two different choices result in an identical outcome . However , since in that experiment responding for the two choices is trained and tested in separate sessions , rather than the choices being available concurrently , the reinforcement learning framework cannot see it as if the values of the choices could be compared together . Therefore , in order to test the above prediction of our model , it is necessary to use a concurrent schedule . Another theoretical account for competition between the S-R and the A-O systems proposed by Dickinson [67] predicts that competition between the systems depends on the relative value of choices . In this account , responding is goal-directed if , and only if , the animal experiences instrumental contingency between responses and outcomes . Experienced contingency is defined as the correlation between a change in response rate and a change in reward rate . Consistent with behavioural data , this theory predicts that in one-choice tasks where a ratio schedule is used , the response rate and thus the reward rate increase during the initial acquisition period . Hence , due to the positive experienced correlation between the changes in these two variables , responding will be goal-directed . However , after extended training , response rate , as well as reward rate , converge to a high rate . This will remove any experienced contingency perceived by the animal and thus , the habitual system becomes dominant . For the case of concurrent schedules where the two outcomes are different but have equal values , this account predicts that even after extensive training , the animal might choose either of the two responses from time to time . Thus , every time that the animal performs one of the two responses , it experiences a loss of the outcome that could be acquired by performing the other response . In this respect , the animal always experiences a local correlation between response and outcome rates and thus , remains goal-directed even after extensive training . This prediction is also consistent with behavioural data [26] . However , if the identity of the two outcomes are the same , this theory will have a different prediction . In such a case , since the outcomes are identical , the rate of outcome will be fixed after extensive training regardless of which of the two responses is performed . Thus , in this case , the local A-O rate correlation dies out and responding becomes habitual . Moreover , this account predicts that if the two choices result in different outcomes that have markedly different values , responding will become habitual after extensive training . This is because after extensive training , the high-value choice will become stereotyped and the other response will be chosen rarely . Thus , since only one of the two outcomes is often experienced with a consistently high rate , the locally experienced A-O rate correlation decreases . In fact , the experienced A-O rate correlation is negatively correlated with the difference between the values of the two outcomes: the higher the difference between the values , the lower the experienced instrumental contingency . As a result , if the values of the two outcomes are sufficiently different , responding will become habitual eventually . In this respect , both the theoretical account of [67] and our model predict that arbitration depends on the relative value of the two choices . A summary of the predictions of the reviewed dual-process accounts are provided in Table 1 . The experimental schedules of the first and the third rows of the table , as discussed before , are used in [3] and [26] , respectively . As shown , the prediction of all three arbitration mechanisms for these two cases are the same , and supported by behavioural data . However , the theories have differential predictions in the other two cases that are not tested yet . One critical assumption of our model that is worth being tested is the assumption that arbitration between the systems is independent of any knowledge that is acquired by the goal-directed system . This assumption is in contrast to the model proposed in [5] , where the uncertainty of the goal-directed system also plays role in competition among the systems . One way to test this assumption of our model is to manipulate the knowledge of the goal-directed system , while other variables are remained intact , and to test the impact on the goal-directedness of animal's behaviour . For this purpose , a place/response task similar to what is suggested in Figure 10 can be used . In the first phase , the animal is moderately trained to retrieve food from one arm of a T-maze . Since the training period is moderate , we expect that at the end of this phase , the animal will use a place strategy ( goal-directed system ) at the choice point , rather than a response strategy ( habitual system ) . Thus , if the animal is then directly tested in the third phase , e . g . , the starting arm is placed at the opposite end of the maze , it is expected to still turn toward the window . Now , the critical prediction of our model is that if any manipulation is applied only to the goal-directed system during a new phase between training and test , it should not change the animal's strategy . In fact , our model will be falsified if after such manipulations , the animal chooses the “turn right” response at the choice point ( going in the opposite direction of the window ) , which indicates that it is using the response strategy , rather than the place strategy . One manipulation is to put the animal inside the right arm for some very few trials , while the food reward comes at random or is totally removed . This will increase the uncertainty of the goal-directed system about the outcome of the strategy “running toward the window” . Note that the number of trials should be sufficiently small such that the animal is not able to learn the new conditions , but only to increase its uncertainty . Among the variables of our model that influence arbitration ( i . e . , , , and ) , the only variable that is affected due to this manipulation is the average reward variable ( ) . However , since this variable is decreased , the model predicts that such a manipulation will make responding even more goal-directed than before . As the animal has not experienced being at the choice point during the second phase , the habitual system will remain intact in this phase . In sum , our model predicts that whatever the number of trials in the second phase is , the animal must still respond goal-directedly ( turn toward the window ) in the test phase , even though the second phase has increased the uncertainty of the goal-directed system . The above experiment is in fact a way to test the hypothesis of the model that outcome-sensitivity after re-exposure ( in devaluation experiments ) is not the result of shift in control from the habitual to the goal-directed system ( through manipulating the goal-directed knowledge during the incentive learning period , as suggested in [5] ) , but instead , it is because the goal-directed system has been dominant even before devaluation , and the only effect of the re-exposure phase is learning the new incentive value of outcomes ( updating the reward function of the goal-directed system ) . This explanation is the dominant explanation for incentive learning [68] . However , if the rats in the above experiment show response strategy in the third phase ( in contrast to what our model predicts ) , it will support the hypothesis that manipulating the goal-directed system can affect arbitration , and that outcome-sensitivity after devaluation might be due to such a manipulation [5] . Another assumption of our model is that when the animal is at the choice point , the time needed for computing the , which is in fact the time needed for arbitration , is trivial , compared to the time needed for goal-directed search . As mentioned before , this is a plausible assumption since the signal can be computed by a closed form equation [18] . However , it might be argued that goal-directed responding can also be achieved within a trivial period of time . This is possible , for example , by assuming that the goal-directed system is capable of evaluating the value of choices in an off-line mode ( when the animal is not necessarily performing the task ) and caching them for future exploitations . Similarly , the goal-directed system might be argued to be neurally implemented by an attractor equation for value iteration ( e . g . [69] ) . Fortunately , the assumption of our model that goal-directed search requires a considerable time is experimentally testable by measuring the animal's reaction time at the choice points , and comparing them when responding is habitual vs . when it is goal-directed ( see Figure 3:I ) . One limitation of the proposed model is that the computation of the average reward signal , which is assumed to be encoded by tonic dopamine , requires the simulated task to be cyclic and highly repetitive . For example , since shifts in the animal's motivational states don't have an immediate impact on the average reward signal , they cannot have a direct effect on the arbitration mechanism . This is despite the fact that motivational states , like hunger and thirst , are demonstrated to modulate the tonic firing activity of dopamine neurons [70] , even before new training under the new motivational state being provided to the animal . It is also analytically more reasonable that the opportunity cost be a function of motivational states; e . g . a hungry animal has a higher opportunity cost , compared to a sated one . One way to resolve this limitation is to develop a more realistic formulation for opportunity cost , rather than the simple average reward formulation . A similar limitation of the model concerns the necessity of experiencing rule changes by the subject , for the arbitration mechanism to be affected . In fact , the model is silent about how an unexperienced , but verbally communicated , environmental change can affect the competition between the two systems . At least in some cases for humans , it seems that a communicated change in the context makes the goal-directed system able to override the habitual response . Modeling such a phenomenon requires a normative way for the arbitration mechanism to be directly influenced by verbal instructions . Although in our model verbal instructions are supposed to affect the subjects' goal-directed knowledge , they don't contribute to the arbitration mechanism . A critical question that must be answered in any dual-process account of decision making is why animals need two systems . In fact , if the goal-directed system makes more rational decisions , then why the habitual system should have survived ? One raw answer to this question could be that animals' brains were not redesigned anew through the course of evolution , but new capabilities were added to the underlying , evolutionarily old brain structures . A more sophisticated answer is that deliberation is subject to some constraints in a way that making habitual responses is more optimal at many choice points . The constraint that our model relies on is the slowness of deliberation . But it can be argued that an increase in response latency is only one of the costs that the animals' decision making machinery must pay for flexibility in sensorimotor coordination; and some other advantages can be counted for the habitual process , each of which is potentially the basis of another normative computational model . Working memory limitations is another constraint imposed on the goal-directed process . The previously acquired information that the goal-directed system requires for its analysis must first be loaded to working memory . Hence , subject to working memory limitations , the goal-directed system might not be provided with enough materials for an accurate deliberation and so , its response might be less optimal than the corresponding habitual response . One more comparative advantage of the habitual system is that it seems impossible , or at least very costly to deliberate about more than one issue at a time , whereas the habitual responses involve massively parallel processing [71] . For example , so many habitual responses are made by a taxi driver while he/she is driving , but the deliberative system is involved in only one issue , e . g . finding the shortest path to reach the destination . Another influential factor that seems to favour habitual decisions despite their non-optimality is that goal-directed deliberation consumes more energy than habitual action selection . For example , low availability of blood glucose , which is the main fuel supporting brain function , results in impairments in cognitive tasks [72] . This factor can be captured by adding an energy cost term , ( ) , to the cost of deliberation , and hence , for arbitration between the two processes , the signal must be compared with . In both dual-process models proposed in [5] and in this paper , the only type of interaction between the two systems is “competition” . However , collaborative interaction between different associative structures can also facilitate optimal action selection . Among different anatomy-based proposals offered for how segregated cortico-basal ganglia loops might be integrated , the spiral organization of DA neurons have proved compatible with the RL framework . Through these spiral connections between the striatum and the Ventral Tegmental Area/Sabstantia Nigra , the output of more ventral areas of the striatum can affect the functioning of more dorsal regions [73] , [74] . Accordingly , it has been hypothesized that by propagating the teaching signal from associative to motor areas of the basal ganglia , more abstract policy representations can facilitate learning habitual motor-level actions [75]–[77] . Based on these evidence , the goal-directed system can be assumed to affect the computation of the prediction error signal , in order to accelerate consolidating the optimal responses in the habitual system . This can substantially resolve the curse of dimensionality in model-free RL , which refers to the exponential growth of learning required for the habitual system when the complexity of the environment increases [78] . | When confronted with different alternatives , animals can respond either based on their pre-established habits , or by considering the short- and long-term consequences of each option . Whereas habitual decision making is fast , goal-directed thinking is a time-consuming task . Instead , habits are inflexible after being consolidated , but goal-directed decision making can rapidly adapt the animal's strategy after a change in environmental conditions . Based on these features of the two decision making systems , we suggest a computational model using the reinforcement learning framework , that makes a balance between the speed of decision making and behavioural flexibility . The behaviour of the model is consistent with the observation that at the early stages of learning , animals behave in a goal-directed way ( flexible , but slow ) , but after extensive learning , their responses become habitual ( inflexible , but fast ) . Moreover , the model explains that the animal's reaction time must decrease through the course of learning , as the habitual system takes control over behaviour . The model also attributes a functional role to the tonic activity of dopamine neurons in balancing the competition between the habitual and the goal-directed systems . | [
"Abstract",
"Introduction",
"Model",
"Results",
"Discussion"
] | [
"control",
"engineering",
"computer",
"science",
"behavioral",
"neuroscience",
"control",
"systems",
"computational",
"neuroscience",
"biology",
"neuroscience"
] | 2011 | Speed/Accuracy Trade-Off between the Habitual and the Goal-Directed Processes |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.