text
stringlengths 4.16k
144k
|
|---|
performed qPCR based copy number measurements, as described in the Supplementary Data, for raf1Δ, flpΔ and raf1ΔflpΔ mutants. As expected, there was a substantial decrease in the PCN in the flpΔ mutant. However, to our surprise PCN almost doubles in a raf1Δ strain ( Figure 1D). This result apparently contradicts the hitherto believed model of the 2-m plasmid maintenance (15,16,27), where Raf1 is described as an acti- (C) Plasmid stability was measured as the percentage of cells growing on the selective media (i.e. the cells retaining the plasmid) following prolonged growth under non-selective condition and instability (i%) was calculated based on the initial and final stabilities as mentioned earlier (39) (no. of generations ≥ 10). Cells were either grown in YPD or SC-galactose. Paired t-test P-values are shown between the pairs of instability values for statistical significance. (D) Copy number assay: qPCR based copy number measurements to study the effect of Raf1 and Flp on the amplification system. SGY2023, SGY2033 and SGY2029 were used to measure the PCN in raf1Δ, flpΔ and rafΔflpΔ strains, respectively. Raf1 was either overexpressed or deleted in W303 [Cir + ] to measure the PCN. Both the deletion and over-expression of Raf1 showed similar effect on the PCN. Average PCN was increased to around 100 copies per cell when Raf1 was either deleted or over-expressed. Apart from increased average copy number, the variance was also observed to increase. (E) mRNA quantification by RT-qPCR: Quantification of the mRNA level of the three plasmid encoded genes (REP1, REP2 and FLP) showed more than 2-fold increase in the FLP mRNA level in raf1Δ as compared to the wild-type strain while no change in the mRNA level of REP1 or REP2 was observed. vator of FLP and therefore it is expected that the removal of RAF1 should result in attenuated FLP activity and reduced PCN. To investigate whether the increase in PCN in raf1Δ strain is due to increased Flp activity, the expression levels of FLP, REP1 and REP2 genes were measured by assaying the mRNA levels using RT-qPCR in the raf1Δ strain and were compared to the expression level in the wild-type strain ( Figure 1E). FLP was found to be expressed at more than 2-fold higher level in the raf1Δ strain as compared to the wild-type strain, while no difference was observed for either REP1 or REP2. In an earlier study (16,27). Rep1or Rep2 over-expression was shown to reduce FLP mRNA level, however, a prolonged over-expression of Rep1 but not Rep2 caused only a marginal increase in the FLP mRNA level. This observation suggests that the effect of concentration of Rep1 on the FLP transcription is not linear, possibly due to the presence of a feedback transcriptional repression of REP1 itself by the Rep1-Rep2 complex (15,16,27). Moreover, the mRNA level was assayed by northern analysis while in this study we have used RT-qPCR which can detect much lesser variations in the mRNA level. The reason for a high expression of Flp and consequently a higher PCN in raf1Δ strain could be due to a change in stoichiometry of Rep1-Rep2 repressor complex. Intriguingly, as Rep1 is attenuated by a negative feedback loop by the Rep1-Rep2 complex which can be antagonized by Raf1, the regulatory circuit controlling the expression of the plasmid encoded genes is such that both the deletion and the over-expression of Raf1 has similar effect on the stoichiometry of the Rep1-Rep2 complex (see Figures 1C, D and 4A). Moreover, the increase in the PCN, with a concomitant decrease in the stability ( Figure 1) in raf1Δ cells further supports the proposition that the reduced plasmid stability of the 2-m derived plasmid is not an effect of reduced PCN in these cells, suggesting that Raf1 might have a direct effect on the Rep1-Rep2-mediated partitioning pathway. Since it was observed that the loss rate of the 2-m derived plasmid increases due to both Raf1 over-expression and deletion (albeit to different extent) ( Figure 1) and the PCN increases due to RAF1 deletion, we investigated if the copy number also shows the same trend as the plasmid instability. The PCN was therefore measured in a wild-type [Cir + ] cell with over-expressed Raf1 (see Supplementary Data for the description of PCN measurement) and was found that the average PCN also increases as it happens in the RAF1 deleted cells ( Figure 1D). Thus the PCN follows the same trend as the plasmid loss rate and hence this result supports the hypothesis that both deletion and over-expression of Raf1 has identical effect on the plasmid maintenance. Interestingly, Figure 1D shows another intriguing effect of the perturbation of Raf1 on the copy number control mechanism. In both the cases (deletion and over-expression) while the PCN increases, an increase in the overall spread of the data points can also be clearly seen. This observation suggests that the copy number control mechanism loses its stringency and the PCN varies over a larger range of values. This increase in the noise could be due to the loss of a second layer of control that Raf1 provides in addition to the negative feedback control of REP1 expression by the Rep1-Rep2 repressor complex. RAF1 binds to the partitioning locus STB We found that the plasmids show an increased mother bias due to lack of Raf1 ( Figure 1B). The most obvious reason for this increased mother bias can be attributed to a direct influence of Raf1 on the partitioning system similar to what has been reported earlier (28). However, whether this effect of Raf1 on the partitioning system is indirect due to a fluctuation in the PCN in raf1Δ was not addressed. We observed an increased plasmid loss rate when FLP was deleted along with RAF1 ( Figure 1C) confirming a Flp (hence PCN) independent effect of Raf1 on the plasmid stability. To validate our hypothesis that Raf1 may be directly involved in partitioning, we investigated the association of Raf1 with the partitioning locus STB in vivo using ChIP (Chromatin Immunoprecipitation) and monohybrid assays. Both the assays showed that Raf1 binds to the partitioning locus STB in a [Cir + ] strain but not in a [Cir 0 ] strain (Figure 2A-C) which is in contrast to the earlier finding that demonstrated the interaction of Raf1 with STB in vitro in the absence of the Rep proteins (29). The result was further verified by rescuing the Raf1-STB interaction in [Cir 0 ] strain by providing 3HA fused Rep1 or Rep2 driven by an inducible GAL1 promoter. FLAG tagged Raf1 was expressed in the same strain driven by the GAL10 promoter from a CEN plasmid, and STB was provided with an episomal plasmid (YEp). ChIP was performed, and STB was pulled down by both anti-HA and anti-FLAG antibodies. In all the ChIP experiments Raf1 consistently showed pull down at STB when either Rep1 or Rep2 was present ( Figure 2D). Since all the tagged proteins were driven by the GAL promoter, dextrose was taken as a negative control. To validate the ChIP assay data, monohybrid assay was performed ( Figure 2B). STB was cloned upstream of the HIS3 reporter in the plasmid pHISi-1 (Clonetech). Raf1 fused to Gal4 activation domain (AD) was expressed through pGAD424 vector. Rep1 fused to AD was used as a positive control for the monohybrid experiment. All the monohybrid assays were done in [Cir + ] strains. RAF1 might not act as a transcription factor This study and other studies demonstrate that Raf1 affects both PCN and plasmid stability at the same time ((28,29) see Figure 1). What could be the underlying mechanism through which Raf1 affects both the partitioning and the amplification systems together? From the studies it can be proposed that Raf1 does so by altering the activity of Flp and the Rep1-Rep2 complex. There are two hypotheses--Raf1 may alter the activity of the proteins by controlling the expression of the corresponding genes by occupying their promoters as a transcription factor or it may directly interact with these proteins to bring about the changes. To address the former hypothesis, we looked for the localization of Raf1 at the promoters (P REP1 , P REP2 , P FLP , P RAF1 ) of the 2-m encoded genes. The promoters were validated and their strength was measured by cloning a 200-bp upstream region of each of the four ORFs upstream of a LacZ reporter. The expression of the reporter was assayed qualitatively by a colony lift filter assay and quantitatively by liquid -galactosidase assay. Under all the four (E) All the four promoters (P REP1 , P REP2 , P FLP and P RAF1 ) were immuno-precipitated using anti-FLAG antibodies against the FLAG tagged Raf1. Raf1 showed no pull down at any of the promoters. STB was used as the positive control. The occupancy of the HA fused Rep proteins at the promoters was also verified by ChIP using anti-HA antibodies. All the promoters along with STB showed IP with the Rep proteins. While Rep1 showed equal pull down at all the promoters and at STB, Rep2 showed less pull down at STB as compared to the promoters. A linear form of the 2-m circle is represented at the bottom with the sequence of the amplicon targeted for the qPCR. The values for the enrichment per input is one order higher for Rep2 as compared to that of Rep1; the scale is, therefore, different for the 3HA-Rep1 and 3HA-Rep2 ChIP results. The comparison among different targets (P REP1 , P REP2 , P FLP , P RAF1 and STB) is performed from the same ChIP experiment. promoters, LacZ expression was observed, albeit with different levels of expression (Supplementary Figure S2) indicating that the regions of the plasmid cloned upstream of the LacZ reporter can indeed fire LacZ expression and hence contain promoter activity. However, we found that Raf1 does not bind to any of the promoters ( Figure 2E, top panel). This suggests that Raf1 might not be involved in the regulation of the activities of the 2-m proteins by occupying their gene promoters. The Rep proteins however, showed pull down at all the four promoters. This observation is in agreement with the model proposed in the earlier studies (15,16,27) that suggest the binding of the Rep1-Rep2 complex at the promoters of the 2-m plasmid encoded genes. Surprisingly, Rep2 showed at least 10-fold more enrichment per input at all the promoters and STB as compared to Rep1. It is difficult to infer if this difference in the pull down efficiency is physiologically relevant, since all ChIP assays were performed with overexpressed Rep1 and Rep2. Nevertheless, assuming an equal level of expression of both Rep1 and Rep2, it can be said that under identical conditions, Rep2 has a higher affinity for the promoters and STB. This is consistent with the fact that Rep2 is predominantly a nuclear protein and is sufficient in single copy to maintain the plasmid stability (26). Furthermore it has previously been demonstrated by in vitro assays that Rep2 has a higher affinity towards STB (29). Raf1 physically interacts with the Rep proteins Earlier studies (15,16) and the data presented above support the notion that Raf1 can function as an antagonist to Rep1-Rep2 repressor complex and can alter the Flp activity (Supplementary Figure S6 and Figure 1E, respectively). Since in the above section we failed to observe the interaction of Raf1 with any of the promoters of the plasmid borne ORFs, we hypothesize that Raf1 does its function through physically interacting with Rep1, Rep2 or Flp. We tested this hypothesis by performing yeast two-hybrid assay (37) and found that Rep1 and Rep2 interact with Raf1 independently in both the [Cir + ] (Supplementary Figure S1) and [Cir 0 ] strains, however, no such interaction was observed between Raf1 and Flp (Supplementary Figure S6) or between Raf1 itself (Supplementary Figure S1). These results suggest that the observed effect of Raf1 on Rep1-Rep2 complex may be due to a physical blockage of the formation of this complex and the effect of Raf1 on Flp may be mediated indirectly via Rep1-Rep2 complex. Although this result supports the hypothesis that Raf1 acts as an anti-repressor by blocking the Rep1-Rep2 interaction, more direct evidence was required to establish this hypothesis. A competitive twohybrid assay and BiFC assay (Figure 3) was performed to validate this |
hypothesis. Raf1 blocks the formation of the Rep1-Rep2 repressor complex From above and the previous studies it can be gleaned that Raf1 blocks the formation of the Rep1-Rep2 repressor complex through physical interactions. We wished to demonstrate the blockage of Rep1-Rep2 by Raf1 with a more direct experiment. We hypothesized that if Raf1 indeed blocks the Rep1-Rep2 interaction, its overexpression should result in a reduced Rep1-Rep2 interaction leading to the reduced expression of the HIS3 reporter in a yeast two-hybrid assay using AD-Rep1 and BD-Rep2. To test this hypothesis three strains were constructed to perform competitive two-hybrid assay--(i) AD-fused Rep1 and BDfused Rep2 with Cu(II) inducible 3XFLAG-Raf1, (ii) ADfused Rep1 and BD-fused Rep2 with galactose-inducible 3XFLAG-Raf1 and (iii) AD-fused Rep1 and pGBD vector (negative control) with Cu(II) inducible 3XFLAG-Raf1. The over-expression of 3XFLAG-Raf1 upon induction with CuSO 4 has been verified in the first and third strains ( Figure 3B). It was observed that in the presence of Cu(II) (100 and 200 M CuSO 4 ) the interaction of Rep1 and Rep2 was slightly but reproducibly reduced due to the over-expression of Raf1 (see the first row in each panel of Figure 3A) but no effect on the interaction of Rep1 and Rep2 was observed in the absence of Raf1 (see the second row in each panel of Figure 3A). No growth was observed in the negative control irrespective of Raf1 over-expression (see the last row in each panel of Figure 3A). However, a similar assay using monohybrid strain harboring either AD-Rep1 or AD-Rep2 shows that Raf1 does not block the interaction of the Rep proteins with the STB (Supplementary Figure S3). These results indicate that Raf1 attenuates Rep1-Rep2 complex through protein-protein interactions. However, Raf1 cannot negate Rep1-STB, or Rep2-STB interactions suggestive of different binding surfaces of these proteins are involved in protein-protein or DNA-protein interactions. To further confirm our hypothesis that Raf1 physically blocks Rep1-Rep2 interaction, we performed a BiFC (Bimolecular Fluorescent Complementation) assay used to study the physical interaction between two proteins (38). For this, the interaction between VN173-Rep1 and VC155-Rep2 was investigated in the presence of either Raf1 or one of the un-tagged Rep1 or Rep2 (driven by GAL promoter) as competing partner. It was expected that 3XFLAG-Raf1, 3XHA-Rep1 or 3XHA-Rep2 would compete for interaction with VN173-Rep1 and VC155-Rep2 and will lead to depleted fluorescence. The strains for the BiFC assay was constructed as explained in the Supplementary Table S2. While VN173-Rep1 was driven by the leaky CUP1 promoter, VC155-Rep2 or VC155-Flp was driven by the inducible GAL1 promoter. BiFC was observed between Rep1 and Rep2 but not between Rep1 and Flp, (Figure 3C, panels I and II and Supplementary Figure S4) thus confirming that the signal observed due to Rep1-Rep2 interaction is a true signal and not an artifact. Importantly, the BiFC signal was found to be both nuclear and as a localized dot within the nucleus. The dot-like signal segregated between the mother and the bud during cell division in a fashion similar to the chromosome segregation suggesting that the localized BiFC designates the STB bound Rep1-Rep2 complex (Supplementary Figure S4). The strains showing BiFC were then transformed with integrative plasmids harboring galactose-inducible Raf1 or Rep1 or Rep2 to study the effect of the over-expression of these proteins on the Rep1-Rep2 interaction (see Supplementary Data for strain construction). The effect of over-expression on the BiFC signal was quantified using FACS analysis as the change in the signal intensity after induction with galactose. Strain (SGY2058) not containing the YFP (or BiFC) fragments (see Supple- Table S2) was used as reference for FACS analysis. Samples were prepared for FACS analysis as described in the Supplementary Data. An increase in the fluorescence, as a readout for the physical interaction, was observed in all the samples except for the negative control (bearing VN173-Rep1 and VC155-Flp) ( Figure 3C, panel I and Figure 3D, red curve) which was quantified as the shift in the median of the histogram of each sample of one time-point from the reference sample of the same time-point, and was plotted on the Y-axis as (median fluorescence). Highest depletion of the BiFC signal was observed when Raf1 was overexpressed ( Figure 3C, panel III and Figure 3D, green curve). However, contrary to our expectation, over-expression of 3HA fused Rep1 or Rep2 did not cause any depletion in the BiFC signal, and rather enhanced the signal ( Figure 3D). This enhancement of the signal suggests that the overexpression of 3HA-Rep1 or 3HA-Rep2 helped in the formation of the more VN173-Rep1-VC155-Rep2 complex. Overall, competitive two-hybrid and BiFC assays suggest that Raf1 blocks the Rep1-Rep2 interaction and Rep2 helps in the formation of this complex. DISCUSSION The maintenance of the endogenous 2-m plasmid in high copies in the nucleus of various Saccharomyces strains is believed to be a function of the interplay between the partitioning and the amplification systems borne by the plasmid. A genetic circuit that regulates the expression of the plasmid-encoded genes is at the centre of this interplay. From the earlier experimental evidence, it appears that Rep1-Rep2 complex plays a pivotal role in controlling this circuit. Till date it is believed that this complex facilitates both (i) equal partitioning of the plasmid through its role at STB which has been well demonstrated and (ii) maintenance of a steady state PCN through its function as a repressor at the 2-m gene promoters which has been proposed based on transcript analysis. Interestingly, it was suggested that Raf1 might modulate the Rep1-Rep2 complex and thus may have a role both in the partitioning and copy number maintenance. In this study we provide biochemical evidence that Rep1-Rep2 complex indeed binds to the 2-m gene promoters to regulate the expression. However, we cannot conclude whether this binding is direct or indirect through other mediators. Biochemical assay using purified Rep proteins and the DNA targets might reveal this. Importantly, we have characterized the Raf1 function and provide evidence that this protein can physically associate and negate the function of Rep1-Rep2 complex. This way Raf1 can serve as a fine tuner between the partitioning and the amplification systems. This work reveals another layer of molecular mechanism to explain the remarkable stability of the 2-m plasmids at a high copy number. Raf1 plays a role in both the partitioning and the amplification of the 2-m plasmid and promotes faithful plasmid segregation Raf1 has been considered a minor player in the entire scheme of the 2-m plasmid maintenance mechanism and its role in the stable maintenance of the 2-m plasmid has been previously demonstrated (28) but with little mechanistic details. In this study, we verified the effects of both RAF1 deletion and over-expression on both plasmid segregation and copy number. Deletion of Raf1 caused increased plasmid loss rate of the 2-m derived plasmid but so did the Raf1 over-expression (Figure 1). The increased plasmid loss rate due to deletion was less severe than the overexpression. Since the plasmid stability is directly influenced by the Rep1-Rep2 complex and Raf1 is known to affect the plasmid maintenance only when both Rep1 and Rep2 are present (15), we speculated that the increased loss rate in both the deletion and over-expression strains is due to a decrease in the concentration of the Rep1-Rep2 complex. Following the same rationale, we speculated that the PCN should show a similar trend as the plasmid loss rate since the former is directly influenced by the Flp activity which is attenuated by the Rep1-Rep2 complex (16). As expected, the PCN followed the same trend as the plasmid loss rate ( Figure 1D) along with an increase in the FLP mRNA (Figure 1E). Moreover, the PCN showed more variation among different samples when Raf1 was either over-expressed or deleted, suggesting that not only there is a decrease in the FLP attenuation, the control on the expression of FLP is also weakened. This increase in the noise might have resulted from the loss of the fine control that Raf1 provides on PCN under its native expression regime. As expected, the deletion of FLP resulted in a slight increase in the plasmid loss rate ( Figure 1C), possibly due to the reduced PCN ( Figure 1D). However, the plasmid loss rate in RAF1 deleted strain was found significantly higher than the strain where FLP was deleted ( Figure 1C, P = 0.04) and remained more or less same in the strain where both RAF1 and FLP were deleted ( Figure 1C, P = 0.14). These results suggest that the plasmid loss rate is primarily affected by the RAF1 deletion, and FLP deletion has little effect on the plasmid stability. Raf1 binds to STB Using ChIP assay, we have demonstrated in vivo Raf1-STB interaction in the presence of the Rep proteins but not in their absence (Figure 2A-C) and that Raf1-STB interaction can be re-established in a [Cir 0 ] cell if either Rep1 or Rep2 is provided in trans ( Figure 2D). The interaction of Raf1 with STB, independent of any other plasmid-encoded protein, has previously been demonstrated in vitro by Surface Plasmon Resonance (SPR) study, but a similar result was not achieved with the gel retardation assay (29). More contradicting results were obtained such as Raf1 was not identified in a mono-hybrid library screening using STB as the bait but was identified through candidate approach where, Raf1-STB interaction was detected in a monohybrid strain harboring transcriptional AD-Raf1 fusion protein (22). It was also observed that at a low Raf1 level Raf1-STB interaction was lost (29). These observations suggest a transient interaction between Raf1 and STB. In this study, we could not detect Raf1-STB interaction independent of the Rep proteins (Figure 2), neither did our results produce higher enrichment per input at STB with Raf1 compared to the Rep proteins ( Figure 2E), which is contradictory to the earlier findings using SPR where Raf1-STB interaction was shown to be independent of the Rep proteins, and a stronger interaction was observed between Raf1 and STB as compared to both Rep1-STB and Rep2-STB interactions (29). These contradictions between earlier findings and our observation may arise due to the transient but rapid Raf1-STB interaction in the absence of the Rep proteins which could be detected by SPR but not by other assays (ChIP and gel retardation) where a more stable interaction is captured. We suspect, that under native conditions the Raf1-STB interaction is stabilized by the Rep proteins through a stable interaction of Raf1 with the Rep proteins ( Supplementary Figure S6). Rep1 and Rep2 but not Raf1 bind to the 2-m gene promoters The model for the transcriptional control circuit of the 2-m plasmid maintenance proposes the binding of the Rep1-Rep2 complex to the promoters of the plasmidencoded ORFs (summarized in (22)). However, it was only speculated that there are promoters immediately upstream of the four 2-m ORFs that can be occupied by the Rep1-Rep2 complex. In this study, we cloned a 200 bp upstream region of the four ORFs upstream of the LacZ reporter to demonstrate that these fragments can indeed possess promoter activity as they drive the expression of LacZ (Supplementary Figure S2). The quantitative estimation of galactosidase specific activity revealed the relative strengths of these promoters, with RAF1 promoter being the weakest. ChIP assay was performed confirming that both Rep1 and Rep2 bind at the REP1, FLP and RAF1 promoters as per the hypothesis ( Figure 2E, (22)). To be noted, a positive and consistent binding was also observed at the REP2 promoter. Since REP2 is believed to not undergo any transcriptional control by the Rep1-Rep2 complex (15) and therefore binding by either Rep1 or Rep2 at the REP2 promoter is unlikely, although a binding at the promoter may not necessarily warrants for a transcriptional repression. Moreover, STB and 5 regions of FLP and REP2 share homology with a consensus sequence (TC(T rich) 13, 15 ATCTTG) suggesting that REP2 promoter may contain a potential target sequence for the Rep1 and Rep2 interaction. It should be noted that the presence of this consensus sequence varies in number among the 5 REP2 (one repeat between −10 to +20 of REP2 ORF), the 5 FLP (three repeats within −90 to +1of FLP ORF) and the STB (three repeats within the PstI and SnaBI fragment) regions (39). We speculate that the binding of the Rep proteins at the |
REP2 promoter is possible but with lesser stringency due to the presence of only one repeat. On the other hand, ChIP with Raf1, as expected, did not show enrichment at any region of the 2-m plasmid chromatin except for the STB ( Figure 2E). This is in agreement with the earlier studies (29) and the observation that deletion of RAF1 alone can lead to plasmid instability ( Figure 1). However, it is paradoxical that while Raf1 binds to the STB through the Rep proteins, it does not do so at the promoters. While this may imply that there are two distinct kinds of the Rep1-Rep2 complexes (one at the STB and other at the promoters), no other observation from earlier studies or from this study obviates such an assumption. On the contrary it is well known that the chromatin architecture of STB is distinct from the rest of the plasmid (19,20,40,41). Therefore, it is possible that the in- teraction and dynamics of the Rep1-Rep2-STB and Rep1-Rep2-P FLP , may be affected not only by the sequence of the DNA target loci but also by the chromatin architecture of these loci. In that case presence of Rep proteins may not be the only determinant for Raf1 binding. Importantly, in a previous study a substantial changes in the chromatin structure was demonstrated at the 5 regions of FLP and REP1 ORFs and at STB in cells devoid of Rep1 or Rep2 or Raf1 (42). Therefore, we believe that our observation of Rep proteins binding to these regions and at REP2 and RAF1 promoters might bring about the biological functions through Nucleic Acids Research, 2017, Vol. 45, No. 12 7177 making crucial changes in the chromatin structure of these regions. Raf1 blocks the formation of Rep1-Rep2 repressor complex The central component of the network that ensures stable high copy maintenance of the 2-m plasmid is the Rep1-Rep2 repressor complex ( Figure 4A, (22)). It has been proposed that Raf1 controls Flp activity either by blocking the interaction of the Rep1-Rep2 repressor complex with the promoter of FLP (by occupying the promoter and hence blocking the accessibility of the Rep1-Rep2 complex to the promoter) or by blocking the physical interactions among the Rep proteins themselves to form the Rep1-Rep2 complex so that the complex cannot bind to the FLP promoter as a repressor (16); however, no direct evidence was available to support either of this hypothesis. Our data suggests that the former hypothesis might not be true since Raf1 does not occupy the promoter ( Figure 2E). We provide evidence in support of the latter hypothesis by showing that Raf1 directly hinders Rep1-Rep2 interactions ( Figure 3) perhaps by physically associating with these proteins (Supplementary Figure S6). However, we found that Raf1 cannot restrict association of Rep1 or Rep2 with STB ( Figure 2D and Supplementary Figure S3) when the Rep proteins were overexpressed along with Raf1. In an alternate experiment where the Rep proteins were driven by their native promoters, quantitative estimation demonstrated that the occupancy of the Rep proteins at STB was reduced (in case of Rep2) or completely abrogated (in case of Rep1) due to the overexpression of Raf1 (Supplementary Figure S7). It should be noted that in this experiment the Rep proteins were expressed from a single copy integrated form of the 2m plasmid and therefore resulted in very low dosage. The result should be interpreted cautiously, especially for Rep1 which is known to be insufficient for plasmid maintenance at low dosage (27). Both the Rep proteins have an N-terminal oligomerization domain and a C-terminal DNA binding domain. The oligomerization domain is required for association of Rep1 or Rep2 with themselves (Rep1-Rep1/Rep2-Rep2) and the formation of the Rep1-Rep2 complex (43). From our results we propose that Raf1 interacts with the N-terminal oligomerization domain of both Rep1 and Rep2 leading to a blockage of oligomerization site and the formation of the Rep1-Rep2 complex. Whereas, the Rep proteins can still bind to the STB through their C-terminal DNA binding domain. It has been demonstrated that the overexpression of Raf1 leads to an increased FLP expression when both Rep1 and Rep2 are present, but there is no effect on the expression of FLP when either of them is absent (16). Our data that Raf1 blocks the formation of the Rep1-Rep2 repressor complex also explain the dependence of Flp activation function of Raf1 on the presence of the Rep proteins. These results support the proposition that Raf1 blocks the formation of Rep1-Rep2 complex and hence the repression of FLP. An obvious prediction of this model is that the PCN should decrease in a raf1Δ strain as Flp is de-repressed. However, a higher PCN was observed which apparently contradicts this paradigm ( Figure 1D). This increase in PCN is unlikely due to plasmid missegregation as shown in raf1Δ strain (Figure 1). This is because all the copy number assay experiments were performed in non-selective (YPD) growth conditions and therefore, any increase in the copy number is not due to the selection of plasmid bearing cells in the population or due to the effect of plasmid missegregation. Removal of FLP caused highest drop in PCN ( Figure 1D) as expected; however, removal of RAF1 in flpΔ strain caused an increase in PCN albeit below the steady state level observed in the wild-type strain. Since Raf1 itself has no amplification function and in the absence of Flp the amplification system is completely defunct, it is difficult to explain this higher steady state copy number as any indirect activation of the amplification system. Based on the above observations we propose a working model (Figure 4) for the transcriptional regulation and the resulting observable effect on the plasmid stability and PCN. According to the model, the Rep1-Rep2 complex is central to the two pathways that control both the plasmid stability and the PCN. The observed effect of the removal of RAF1 on the stability and PCN might result from a decreased concentration of active Rep1-Rep2 complex. Interestingly, a similar phenotype is expected if Raf1 is overexpressed ( Figure 1C and D) indicating a balanced Raf1 dosage is essential for maintaining a certain level of Rep1-Rep2 concentration optimized to provide steady state PCN. We argue that the feedback loops operating among Rep1, Rep2 and Raf1 as discussed below are crucial to understand how plasmids are stably maintained at a steady state level of copy number. A Rep1-Rep2 repressor amplified negative-feedback motif may create a stable control system Oscillatory networks are common in the living systems. They lie at the core of the biological rhythms and are involved in synchronization of the myriad cellular processes. Sustained oscillations of the networks can be achieved through a variety of regulatory motifs ranging from simple two-component time-delayed negative feedback loops to complex multi-component hysteresis oscillators constructed by interlinked negative and positive feedback loops (30). We analyzed our results from this perspective to envisage a systems level explanation for rhythmic events such as the steady state PCN maintenance through the synchronization of Flp-mediated copy number amplification with the cyclic cell cycle dependent efficient plasmid segregation process (44). We propose that these cyclic and synchronous events are a result of a multi-component oscillatory network formed by two feedback loops. The regulatory network that maintains a stable steady state PCN results from the two feedback loops (1 and 2) shown in Figure 4A and C and is reminiscent to a repressor amplified negative-feedback motif discussed elsewhere (33). In this network, the two feedback loops are connected through the Rep1-Rep2 repressor complex. The first feed-back loop (numbered 1) can be viewed as a negative-feedback loop where the product (repressor complex Rep1-Rep2) represses its own expression. Since REP2 is shown to be unaffected by the repressor complex (27) and hence its dosage does not affect the control network, it can be reasonably assumed to be expressed constitutively and a non-limiting factor for the con-trol network. Under this assumption, the cellular concentration of the Rep1-Rep2 complex will be affected only by the fluctuations in the cellular concentration of Rep1. Consequently, the first negative-feedback loop can be perceived as a Rep1 negative-feedback where Rep1 represses its own expression. The first loop can also be seen as a negativefeedback loop with a time delay by a series of intermediate processes such as time delay due to the intermediate processes of transcription-translation-SUMOylation of Rep and Flp proteins-transport into the nucleus etc. (45-47. This time delay in feedback gives rise to an oscillatory response and causes the repressor (Rep1-Rep2) to oscillate around a steady state level (30,33). In the absence of the positive feedback loop when Raf1 is absent ( Figure 4B and D), the oscillations maintain a certain average steady state concentration of the repressor as operated by loop number 1 alone and the concentration is believed to be, as discussed below, less than what is observed in the wildtype. However, introduction of a positive feedback loop by Raf1 (number 2), as in the wild-type shifts the average steady state concentration of the repressor. The new oscillatory response formed due to the introduction of Raf1 is similar to a three component repressor amplified negativefeedback loop ((33) and Figure 4C) where the repressor (Rep1-Rep2 complex) represses the transcription of both Rep1 and Raf1. The Rep1-Rep2/Raf1 positive feedback loop pushes the Rep1-Rep2 complex away from the steady state, causing an upward shift in the average steady state concentration of the repressor due to a positive feedback from Raf1. Thus, this two-looped system approach explains the decreased stability and an increased PCN (and increased FLP mRNA) in the raf1Δ cells perhaps due to an overall decrease in the average steady state concentration of the Rep1-Rep2 repressor. It is important to investigate whether under native condition in the raf1Δ cells the steady state level of Rep1-Rp2 is indeed lower than the wild-type. Efforts to test this by expressing the VN173-Rep1 and VC155-Rep2 under their native promoters failed repeatedly due to several technical problems. VN173-Rep1 fusion in the native plasmid could not be achieved due to gross destabilization of the plasmid. Transformation of a wild-type [Cir + ] strain or a VC155-Rep2 [Cir + ] strain (where all the native plasmids harbored VC155-REP2) with VN173-Rep1 tagging cassette yielded sick colonies on the selective plates after transformation. Alternatively, Rep1 and Rep2 expressed from a chromosomally integrated version of the 2-m plasmid (pFV14) were fused with VN173 and VC155, respectively at their N-termini. However, in such cells BiFC between VN173-Rep1 and VC155-Rep2 driven by their native promoters was not observed. We suspect that this lack of BiFC was due to a very low concentration of the intracellular VN173-Rep1-VC155-Rep2 complex as they were expressed from single copy genes. Nevertheless, this approach provides a theoretical platform to explain the rhythmic changes in the components of the 2 m system synchronized with the cell cycle events. The oscillations in Rep1-Rep2 complex can be fine-tuned to match the cell cycle events by changing the kinetic parameters of the control system. Further investigation measuring the variations in the stoichiometry of Rep1-Rep2 complex in the presence and absence of RAF1 and coupling that information with a detailed modeling and analysis of the control network can help in better understanding the overall dynamics of the 2m plasmid maintenance mechanism. Oxidative Stress in Human Atherothrombosis: Sources, Markers and Therapeutic Targets Atherothrombosis remains one of the main causes of morbidity and mortality worldwide. The underlying pathology is a chronic pathological vascular remodeling of the arterial wall involving several pathways, including oxidative stress. Cellular and animal studies have provided compelling evidence of the direct role of oxidative stress in atherothrombosis, but such a relationship is not clearly established in humans and, to date, clinical trials on the possible beneficial effects of antioxidant therapy have provided equivocal results. Nicotinamide adenine dinucleotide phosphate (NADPH) oxidase is one of the main sources of reactive oxygen species (ROS) in human atherothrombosis. Moreover, leukocyte-derived myeloperoxidase (MPO) and red blood cell-derived iron could be involved in the oxidative modification of lipids/lipoproteins (LDL/HDL) in the arterial wall. Interestingly, oxidized lipoproteins, and antioxidants, have been analyzed as potential markers of oxidative stress in the plasma of patients with atherothrombosis. In this review, we will revise sources of ROS, focusing on NADPH oxidase, but also on MPO and iron. We will also discuss the impact of these |
oxidative systems on LDL and HDL, as well as the value of these modified lipoproteins as circulating markers of oxidative stress in atherothrombosis. We will finish by reviewing some antioxidant systems and compounds as therapeutic strategies to prevent pathological vascular remodeling. Introduction Atherothrombosis is the main cause of death in developed countries. The main feature underlying atherothrombosis is a chronic pathological remodeling of the vascular wall, characterized by lipid deposition, oxidative stress, immune-inflammatory and proliferative responses, along with proteolysis, neo-angiogenesis, apoptosis, calcification and fibrosis [1,2]. Reactive oxygen species (ROS) are considered crucial mediators of vascular homeostasis and pathogenesis in vascular diseases. Low levels of ROS are essential for the regulation of multiple cellular processes and signaling pathways, whereas uncontrolled ROS production, as occurs in several vascular diseases including atherosclerosis or abdominal aortic aneurysm (AAA), results in exacerbated oxidative stress that damages vascular cells through a myriad of processes [3][4][5][6][7][8]. Known risk factors for atherothrombosis include increased systemic low-density lipoprotein (LDL) and reduced high-density lipoprotein (HDL) cholesterol levels. This systemic alteration of lipoprotein particles is accompanied by the increased lipoprotein retention observed during the earlier stages of the development of the vessel wall remodeling. LDLs are highly susceptible to being modified by the oxidative milieu found inside the vascular wall. In fact, the "oxidative modification hypothesis of atherosclerosis" [9] was based on the evidence that modified oxidized LDLs are retained in atherosclerotic plaques and their uptake by scavenger receptors on phagocytes lead to foam cell formation. In addition, oxidative stress could also modify other lipoproteins (e.g., HDL) or other molecules involved in different initial processes associated with vessel wall remodeling (e.g., nitric oxide-related endothelial dysfunction). However, the precise sources of oxidative stress in these initial stages are not completely defined. In the more advanced stages, intraplaque hemorrhages in complicated atherothrombotic disease [10] and intraluminal thrombus (ILT) in AAA [11] both lead to clinical complications due to arterial wall rupture, involving intimal cap rupture in complicated atherothrombotic plaques and medial and adventitial rupture in AAA. No matter where there is intraplaque or intraluminal localization, hemorrhages and/or thrombi involve trapping of red blood cells (RBCs), leukocytes and activating platelets. In this context, RBC-derived, iron-rich heme group and leukocyte-derived oxidants (e.g., NADPH-dependent ROS and myeloperoxidase-MPO-), are the main sources of oxidative stress and are able to modify lipids, proteins and DNA, which leads to the progression of atherothrombotic pathology towards clinical events [12]. In the present review, we will summarize the molecules involved in redox imbalance in human atherothrombosis, highlighting the functional consequences of oxidative stress mainly in lipoproteins, due to their key role in vascular diseases. Moreover, we will describe studies by analyzing the potential use of some biomarkers of redox imbalance, as well as its potential therapeutic value, in these pathologies. Generation and Elimination of ROS ROS are reactive derivatives of oxygen metabolism. These include molecules with unpaired electrons, also termed free radicals such as superoxide anion (O 2 − ) and hydroxyl radical (OH), which are highly unstable and have short half-lives. Non-radical ROS include more stable molecules with longer half-lives such as hydrogen peroxide (H 2 O 2 ), peroxynitrite (ONOO − ) and hypochlorous acid (HOCl) [13]. The majority of O 2 − generated is rapidly converted to H 2 O 2 , which, in contrast to O 2 − , penetrates cell membranes easily, and functions as a second messenger that activates multiple signaling pathways. O 2 − is formed by the univalent reduction of molecular oxygen. This process is mediated by different enzymatic systems including NADPH oxidases (NOX), xanthine oxidase, lipoxygenase, cyclooxygenase, CYP450 isoforms, monoxygenases and uncoupled endothelial NO synthase (eNOS). O 2 − can also be generated non-enzymatically by the mitochondrial electron transport chain, the endoplasmic reticulum (ER), and peroxisomes ( Figure 1) [13][14][15][16]. O 2 − can be converted into H 2 O 2 spontaneously or by the superoxide dismutases (SOD) enzymes: cytosolic Cu/Zn-SOD (SOD1), mitochondrial Mn-SOD (SOD2) and extracellular EC-SOD (SOD3). Moreover, some types of NOX including NOX-4 and dual oxidases (DUOX)-1 and -2, can directly produce H 2 O 2 [16], which can also be synthesized as a by-product of different enzymes including some which are important in cardiovascular diseases (CVD) such as lysil oxidase [17,18] (Figure 1). Figure 1. Generation and elimination of ROS. Enzymatic systems (in red) including NADPH oxidases (NOXs), Xanthine Oxidase (XO), Lipoxygenase (LO), Cyclooxygenase (COX) and uncoupled eNOS produce O2 − that can also be generated non-enzymatically (in orange) by the mitochondrial electron transport chain (mETC), the endoplasmic reticulum (ER) and peroxisomes. O2 − is then transformed into H2O2 spontaneously or through superoxide dismutases (SODs) or can be synthesized directly by NOX-4 or as a by-product of lysyl oxidase (LOX). O2 − can rapidly react with NO leading to the formation of ONOO − . H2O2 can be then converted into more reactive molecules, including hydroxyl radical (OH − ) by Fenton reaction or into HOCl by myeloperoxidase (MPO). Furthermore, H2O2 can also be transformed into H2O by catalase (CAT) or by the glutathione peroxidase (GPx)/gluthathione reductase (GR) and the thioredoxin (Trx)/peroxiredoxin (PRx) systems. TrxR: thioredoxin reductase; TrxPrx: thioredoxin peroxidase. In the presence of reduced transition metals (e.g., ferrous or cuprous ions), H2O2 can be converted into the highly reactive OH that damages different macromolecules including lipids, proteins and DNA. Alternatively, H2O2 may be converted into water by the enzyme catalase or glutathione peroxidase-1 that catalyzes the reduction of H2O2 using reduced glutathione (GSH) as an electron donor. GSH is transformed into glutathione disulfide (GSSG) by glutathione peroxidase, which can then be converted back to GSH by glutathione reductase in an NADPH-consuming process ( Figure 1). The thioredoxin (TRX) system, in which thiol-dependent peroxidases (peroxiredoxins, PRDX) are provided with electrons to remove reactive oxygen and nitrogen species rapidly, is also an important H2O2 detoxifying system [19]. Myeloperoxidase (MPO) is a well-known enzyme, mainly released by activated neutrophils, characterized by powerful pro-oxidative and pro-inflammatory properties. MPO is a heme peroxidase that produces HOCl in the reaction between H2O2 and chloride ions. These mediators are not only important for the antimicrobial activities of the innate immune system but also contribute to immune inflammatory diseases, including atherosclerosis and AAA [6,20] (see Section 3.2). Other important ROS/reactive nitrogen species (RNS) is peroxynitrite (ONOO − ) which is formed by the reaction of NO with O2 − . RNS produce post-translational modifications of proteins, nitrative stress and different modifications such as tyrosine nitration. Moreover, not only ONOO − but also MPO pathways have been involved in protein nitration [20]. Increased levels of ROS activate nuclear factor (erythroid-derived 2)-like 2 (Nrf2), a master regulator of the antioxidant response, which is activated to counteract oxidative stress. Nrf2 controls the expression of about 250 genes including those encoding antioxidant enzymes such as those In the presence of reduced transition metals (e.g., ferrous or cuprous ions), H 2 O 2 can be converted into the highly reactive OH that damages different macromolecules including lipids, proteins and DNA. Alternatively, H 2 O 2 may be converted into water by the enzyme catalase or glutathione peroxidase-1 that catalyzes the reduction of H 2 O 2 using reduced glutathione (GSH) as an electron donor. GSH is transformed into glutathione disulfide (GSSG) by glutathione peroxidase, which can then be converted back to GSH by glutathione reductase in an NADPH-consuming process ( Figure 1). The thioredoxin (TRX) system, in which thiol-dependent peroxidases (peroxiredoxins, PRDX) are provided with electrons to remove reactive oxygen and nitrogen species rapidly, is also an important H 2 O 2 detoxifying system [19]. Myeloperoxidase (MPO) is a well-known enzyme, mainly released by activated neutrophils, characterized by powerful pro-oxidative and pro-inflammatory properties. MPO is a heme peroxidase that produces HOCl in the reaction between H 2 O 2 and chloride ions. These mediators are not only important for the antimicrobial activities of the innate immune system but also contribute to immune inflammatory diseases, including atherosclerosis and AAA [6,20] (see Section 3.2). Other important ROS/reactive nitrogen species (RNS) is peroxynitrite (ONOO − ) which is formed by the reaction of NO with O 2 − . RNS produce post-translational modifications of proteins, nitrative stress and different modifications such as tyrosine nitration. Moreover, not only ONOO − but also MPO pathways have been involved in protein nitration [20]. Increased levels of ROS activate nuclear factor (erythroid-derived 2)-like 2 (Nrf2), a master regulator of the antioxidant response, which is activated to counteract oxidative stress. Nrf2 controls the expression of about 250 genes including those encoding antioxidant enzymes such as those involved in glutathione and TRX systems, SOD, catalase, and hemoxygenase-1, among many others [21]. In addition to enzymatic degradation of ROS, various low-molecule compounds can directly react with ROS. These molecules can be endogenously synthesized or obtained from diet and include vitamins C and E, uric acid, glutathione, flavonoids and thiols, among others [3]. The misbalance between ROS generation and elimination determines oxidative stress. It is now accepted that oxidative stress responses are involved in many cellular and tissue processes in relation to CVD and its risk factors. In fact, all established cardiovascular risk factors such as hypercholesterolemia, hypertension, diabetes mellitus, and smoking enhance ROS generation. The ROS-modulated processes include proliferation and migration of vascular smooth muscle cells (VSMC), endothelial dysfunction with diminished NO availability and increased vasoconstriction, and increased production of isoprostanes. These isoprostanes are eicosanoids derived from nonenzymatic oxidation of arachidonic acid via the interaction with ROS and they cause artery vasoconstriction via TP receptors, VSMC proliferation and platelet aggregation. Moreover, endothelial activation with the expression of adhesion molecules, recruitment of inflammatory cells, lipid oxidation, platelet aggregation, activation of metalloproteinases and altered extracellular matrix deposition, are also activated by ROS [3][4][5][6][7]. All cells in the vascular wall including VSMC, endothelial cells and adventitial cells, together with circulating cells (such as platelets and RBC) are able to generate ROS. Moreover, inflammatory cell infiltration is now recognized as a potential source of ROS in different CVD including atherosclerosis and AAA. Most of the vast information available on the role of oxidative stress in CVD has been obtained from animal models and excellent reviews covering these issues are already available [7,8,[22][23][24][25][26][27]. For the sake of clarity, we now discuss in depth the possible role of oxidative stress in atherothrombosis in the human context. Sources of Oxidative Stress in Human Vascular Diseases In this section, we will focus on NADPH oxidase as a master of ROS production, but also on leukocyte-derived MPO and in RBC-derived iron that have been linked to lipid/lipoprotein oxidation in humans ( Figure 2). involved in glutathione and TRX systems, SOD, catalase, and hemoxygenase-1, among many others [21]. In addition to enzymatic degradation of ROS, various low-molecule compounds can directly react with ROS. These molecules can be endogenously synthesized or obtained from diet and include vitamins C and E, uric acid, glutathione, flavonoids and thiols, among others [3]. The misbalance between ROS generation and elimination determines oxidative stress. It is now accepted that oxidative stress responses are involved in many cellular and tissue processes in relation to CVD and its risk factors. In fact, all established cardiovascular risk factors such as hypercholesterolemia, hypertension, diabetes mellitus, and smoking enhance ROS generation. The ROS-modulated processes include proliferation and migration of vascular smooth muscle cells (VSMC), endothelial dysfunction with diminished NO availability and increased vasoconstriction, and increased production of isoprostanes. These isoprostanes are eicosanoids derived from nonenzymatic oxidation of arachidonic acid via the interaction with ROS and they cause artery vasoconstriction via TP receptors, VSMC proliferation and platelet aggregation. Moreover, endothelial activation with the expression of adhesion molecules, recruitment of inflammatory cells, lipid oxidation, platelet aggregation, activation of metalloproteinases and altered extracellular matrix deposition, are also activated by ROS [3][4][5][6][7]. All cells in the vascular wall including VSMC, endothelial cells and adventitial cells, together with circulating cells (such as platelets and RBC) are able to generate ROS. Moreover, inflammatory cell infiltration is now recognized as a potential source of ROS in different CVD including atherosclerosis and AAA. Most of the vast information available on the role of oxidative stress in CVD has been obtained from animal models and excellent reviews covering these issues are already available [7,8,[22][23][24][25][26][27]. For the sake of clarity, we now discuss in depth the possible role of oxidative stress in atherothrombosis in the human context. Sources of |
Oxidative Stress in Human Vascular Diseases In this section, we will focus on NADPH oxidase as a master of ROS production, but also on leukocyte-derived MPO and in RBC-derived iron that have been linked to lipid/lipoprotein oxidation in humans ( Figure 2). NADPH Oxidase The NADPH oxidase system is the main source of ROS in the vessel wall and is present in endothelial cells, VSMC, adventitial fibroblasts, and infiltrating monocytes/macrophages. The structure and function of NADPH oxidase in physiological and cardiovascular pathological conditions have been extensively reviewed [5,6,14,16,[28][29][30]. In contrast to the rest of the ROS-producing enzymes that produce ROS as a by-product of their activity, the main catalytic function of NADPH oxidases is the generation of ROS. NADPH oxidase isoforms in mammals have a catalytic subunit called NOX (NOX-1-5) or DUOX (DUOX-1-2) and up to seven regulatory subunits, leading to the formation of seven NADPH oxidase isoforms. NOX-1, NOX-2, NOX-4 and NOX-5 are expressed in the cardiovascular system. The classic NOX, NOX-2, was initially found and characterized in leukocytes. Cytosolic adaptor proteins called "NOX organizers" (p47phox or NOXO1 and p40phox) and "NOX activators" (p67phox or NOXA1) that bind GTP-Rac and affect the flow of electrons, regulate the activity of NOX-1, NOX-2 and NOX-3. When the p22phox component binds with NOX-1-4, a stable heterodimeric complex and then the active oxidase are formed [5,6,14,16,[28][29][30]. NOX isoforms are variably expressed in vascular cells with some of them coexisting in the same cell type, suggesting different cell functions for each NOX. Thus, endothelial cells express NOX-1, NOX-2, NOX-4 and NOX-5, the latter being expressed only in humans; VSMCs mainly express NOX-1, NOX-4 and NOX-5; and adventitial cells mainly express NOX-2 and NOX-4 [5,6,14,16,[28][29][30]. As mentioned, NOX-2 is mainly expressed in phagocytes (neutrophils and macrophages) but platelets also express NOX-2 where it has a central role in generating O 2 − [6,31]. Many studies in animal models have demonstrated a key role of vascular and phagocytic NADPH oxidase isoforms in the development of vascular diseases and to some degree these observations have also been extended to humans. Vascular production of O 2 − increases as a consequence of risk factors for atherosclerosis [32]. Atherosclerotic lesions contain abundant p22phox and NOX-2 (also termed gp91phox) that correlated with the severity of atherosclerosis [33,34] where p22phox is located in adventitial fibroblasts, VSMC, macrophages in the neointima and media, and in endothelial cells [35]. Importantly, clinical and experimental studies support the role of NOX-2 expressed in platelets in the atherothombotic process [7,8] by mechanisms that include the expression of the CD40 ligand, a protein with pro-inflammatory and prothrombotic properties on interaction with its receptor CD40 thereby modulating platelet function [6,31]. The role of NOX-1 in atherogenesis remains controversial since NOX-1 was undetected or had very low expression in human lesions [33,34]. Interestingly, NOX-1 upregulation was demonstrated in plaques from patients with cardiovascular events or established diabetes mellitus [36]. In contrast, NOX-4 was found exclusively in non-phagocytic cells with NOX-4 being highest in stage IV atherosclerosis and dramatically decreased in the most complicated plaques that are characterized by fibrosis and a reduction in intimal VSMC [34]. Another study also showed that NOX-4 mRNA levels were reduced in plaques from patients with cardiovascular events or established diabetes mellitus which was found that, together with experimental studies, pointed to a possible role of NOX-4 as a negative modulator of inflammation and remodeling to convey atheroprotection [36]. Finally, NOX-5 was found to be upregulated in atherosclerosis in the endothelium in the early lesions and in VSMC in the advanced coronary lesions [37] and more recently, NOX-5 was found in human monocytes and macrophages and in macrophage-rich areas within human carotid artery atherosclerotic plaques [38]. However, the fact that NOX-5 is only expressed in humans has slowed down the progress in the elucidation of the impact of NOX-5 in atherosclerosis. Analysis of tissue from patients undergoing bypass surgery revealed that, besides changing NOX isoforms, diabetes is characterized by increased expression of p22phox, p47phox, and p67phox compared with non-diabetics [39]. Moreover, NADPH oxidase activity in peripheral blood mononuclear cells positively correlated with carotid intima-media thickness, a surrogate marker of atherosclerosis, in asymptomatic subjects [40]. Increased expression and activity of NADPH oxidases are also important mechanisms underlying oxidative stress in human AAA [41]. Specifically, mRNA levels of p22phox, NOX-2 and NOX-5 were significantly increased in AAAs while NOX-4 mRNA expression was lower [41]. Notably, although human studies clearly suggest a role for increased oxidative stress in atherothrombosis, the specific cell responsible remains elusive. In humans, direct evidence of the relationship between ROS and atherothrombosis mainly originate from studies in patients with NOX-2 loss of function by genetic NOX-2 or p47phox deficiencies (chronic granulomatous disease). These patients show increased flow mediated dilation and therefore NO-induced vasodilation, and diminished carotid intima-media thickness, two surrogate markers of atherosclerosis [7,8,42,43]. Interestingly, reduced carotid but not coronary artery atherosclerosis was observed in patients with chronic granulomatous disease suggesting that NOX2-related mechanisms may play a lesser role in coronary atherosclerosis than in other arterial beds [44]. In the same line of evidence, patients with the C242T polymorphism of the p22phox subunit (associated with lower oxidative stress) [45], had less cardiovascular death, myocardial infarction and re-vascularization compared with those carrying the wild-type allele [46], demonstrating that the 242T allele was a predictor of lower risk of recurrence of cardiovascular events in high-risk patients. MPO MPO is a hemoprotein mainly released by activated leukocytes that catalyzes the reaction between H 2 O 2 and chloride ions to produce HOCl as the primary oxidant. MPO-derived oxidants generate a footprint of specific (e.g., 3-Chlorotyrosine, 3-Cl-Tyr) and nonspecific (e.g., protein carbonyls and 3-nitrotyrosine modifications) oxidation products. Moreover, MPO may serve as a source of free iron through a mechanism that involves heme depletion [47] and MPO has been also implicated in lipoprotein oxidation in vivo [48]. It was long established that MPO works as an NO-oxidase, consuming NO to lead to impaired endothelial relaxation [49]. Previous studies have shown that catalytically active MPO and its oxidative species are present in human atherothrombotic tissues [50][51][52][53]. Moreover, plasma MPO levels are increased in atherosclerotic and AAA patients [53,54]. It should be noted that MPO is a potent predictor of cardiovascular events in patients with chest pain [55] and MPO levels are a significantly better predictor of major adverse cardiovascular events than NT-proBNP levels in patients with ST-segment elevation in myocardial infarction who are treated with primary percutaneous coronary intervention [56]. More recently, increased MPO indexed to HDL particle concentration at baseline is associated with increased risk of incident cardiovascular events in a population initially free of CVD [57]. Iron Iron plays crucial roles in cell proliferation and metabolism by serving as a functional constituent of various enzymes, normally associated to the hemo group. The main iron pool in the body is found within the hemoglobin (Hb) of RBCs, which, after a mean half-life of 120 days, are taken up by resident macrophages by erythrophagocytosis. However, free iron is toxic through the generation of ROS via the Fenton reaction. Iron could be released due to microvessel rupture in atherosclerotic plaques [10] or after RBC lysis within the intraluminal thrombus of AAA [11]. Iron was observed in advanced carotid atherosclerotic plaques [58]. Moreover, it has been recently described that RBC efferocytosis by the arterial wall promotes oxidation in early-stage human atheroma [59]. In this respect, the role of iron in the pathogenesis of atherosclerosis was originally associated with its ability to catalyze the oxidation of lipoproteins [60], but potential novel mechanisms by which iron could modulate atherogenesis have been later described [61]. More recently, Sawada et al. described that iron was accumulating in human AAA walls compared with non-AAA walls and the extent of the iron-accumulated area positively correlated with that of the area of 8-hydroxy-2 -deoxyguanosine expression [62]. A long time ago, iron was proposed as a cardiovascular risk factor, suggesting that the lower incidence of CVD in premenopausal women could be explained by the lower body iron stores [63]. We also described local iron retention and altered iron recycling associated with high hepcidin and low transferrin systemic concentrations in AAA patients, potentially leading to reduced circulating Hb levels [64]. Moreover, low Hb levels were associated with AAA progression. In this respect, anemia has been associated with several chronic (inflammatory) diseases, including CVD, probably related to the diversion of iron recycling [65,66]. Markers of Oxidative Stress in Human Vascular Diseases A biomarker is a marker reflecting or integrating one or several biological activities. In the case of biomarkers of oxidative stress, we will refer to both pro-and antioxidant biomarkers. Such markers may be any detectable and quantifiable molecules including proteins, peptides, lipids, nucleic acids, etc. Many studies have reported oxidative stress markers in tissues and plasma of patients with atherothrombosis. However, although markers assessing the oxidation of phospholipid and protein components of LDL were among the first to be developed, clinical trials including cross-sectional and retrospective and prospective studies provided equivocal results [6]. Among the reasons explaining these conflicting results, methodological issues have been highlighted and it is beyond the scope of this article to review this aspect in detail and the reader is referred to excellent reviews on this aspect [6,20,67]. In any case, although promising advances in this field are being carried out in the last years, it is still premature to unequivocally affirm the clinical validity of a specific oxidative stress biomarker for the management of patients with higher risk of CVD [6]. [68]. LDL has been associated to atherosclerosis development as in the subendothelial space, LDL becomes modified by either aggregation, acetylation and/or oxidation. Modified LDL induces endothelial injury, increases the expression of adhesion molecules, favoring monocyte adhesion and its differentiation to macrophages [69]. Moreover, oxidized LDL stimulates platelet aggregation and inhibits endothelial NO synthase expression/activity, promoting atherogenesis. Therefore, modified LDL is a main mediator inducing vascular damage and atherosclerotic disease development. Low-density lipoprotein (LDL) is the main player in cholesterol transport to the cells and high concentration of LDL is a well-established cardiovascular risk factor Oxidized LDLs are mainly present in ceroids that can be formed within the cell and are similar to cholesterol crystals [70]. Ceroids are autofluorescent, insoluble and sudanophilic polymers composed of aggregated proteins entrapping lipids. Iron deposits, Hb and MPO colocalized with ceroids within cells and tissues such as atherosclerotic plaques and AAA [71][72][73]. After the initial finding of the presence of ceroid/lipofuscin and of peroxidized lipids in atherosclerotic lesions [74,75], it was discovered that oxLDLs are present in atherosclerotic lesions [76]. Oxidation of LDL can be carried out by, among others, transition metals, Hb, lipoxygenases, and ROS generated by vascular cells or phagocytes. Interestingly, oxidation of LDL by NOX-2 containing platelets may represent another mechanism through which NOX-2 activates platelets in a self-perpetuating mechanism [7,8]. HOCl can modify LDLs at the lipid and the protein moieties in vitro and/or in vivo [77]. Malondialdehyde (MDA), a lipid peroxide product released by oxidation from prostanoid metabolism, reacts with the positively charged epsilonamino group of apo B-100 protein lysyl residues, a constituent of the LDL molecular complex [78]. MDA-modified LDLs induce lipid accumulation in macrophages [79]. Circulating MDA-LDL levels has been proposed as a marker of oxidative stress in atherosclerotic CVD [80][81][82] and clinical studies have demonstrated that MDA-LDL levels are associated with the severity of coronary artery disease [83], coronary plaque vulnerability [84] and adverse clinical outcomes after percutaneous coronary intervention with drug eluting stents [85]. Oxidized LDL is immunogenic and the oxidative modifications of apolipoprotein B-100 resulted in the formation of neoepitopes. Oxidation-specific epitopes (OSE) may be indirectly reflected by the presence of circulating antibodies and immune complexes. These are often measured as IgG and IgM autoantibodies to MDA-LDL and apoB-immune complexes [86]. These biomarkers can predict CVD and associated events [87]. In atherosclerotic CVD, IgG and IgM titters to OSE such as MDA-LDL, were predictive of recurrent events in a prospective study with a 15-year follow-up [88]. In general, high levels of IgM OSE biomarkers predict lower risk, consistent with their potential protective function as natural antibodies, and higher levels of IgG biomarkers predict a higher risk, consistent with their general properties of being acquired. Oxidized HDL Although most of the work related to the oxidation hypothesis of atherosclerosis has been performed on LDL, there is also |
evidence that oxidation of other lipoproteins, such as HDL, could also take place during vascular remodeling. In this respect, it is important to note that the majority of cholesterylester hydroperoxides are associated with HDLs rather than LDLs [89]. HDL is the responsible for the reverse cholesterol transport, which is the transport of excess cholesterol from the peripheral tissues to the liver for its elimination in feces and bile [90]. HDL is atheroprotective by several ways; one of them is cholesterol efflux from macrophages, controlling the accumulation of foam cells and atherosclerosis development. Beyond the role of HDL in reverse transport of cholesterol, the particle is protective through other functions such as anti-inflammatory, antioxidant, antithrombotic, anti-fibrotic and vasoprotective properties, protection against lipopolysaccharide and promotion of NO production [91,92]. Epidemiologically, an inverse relationship between HDL cholesterol (HDLc) and cardiovascular risk has been clearly demonstrated [93]. Low HDLc levels are also associated with both AAA presence and progression [94][95][96]. However, HDLc-raising therapies do not result in cardiovascular risk reduction [97]. These negative results have led to a new HDL perception, where the "quality" or "functionality" is more relevant than just HDL plasma levels [98]. It is now fully accepted that in pathological states, such as the oxidative and pro-inflammatory environment present in atherothrombosis, HDL is remodeled, modifying the functionality of the particle. Among functional assays measuring HDL quality, it has been shown that decreased cholesterol efflux capacity is related to incident CVD and CV events [99,100]. In addition, HDL functionality is associated with its molecular (protein/lipid) composition [101]. Regarding AAA, it was previously demonstrated that HDL carries less alpha-1 antitrypsin and higher MPO levels, leading to dysfunctional HDL characterized by decreased antioxidant properties [73,102]. The functionality of the particles could also be derived by the presence of postranslationally-modified proteins. Among them, it has been previously demonstrated that ApoA1, the main constituent of HDLs, could be oxidatively modified, leading to dysfunctional HDLs [103][104][105][106][107]. Previous studies of ApoA1 from human aortic tissues revealed that ApoA1 in the human aorta was extensively oxidatively cross-linked and functionally impaired [108]. Moreover, elevated oxApoA1 levels in a large cohort of subjects presented to a cardiology clinic were associated with increased CVD risk [109]. Similarly, Yassine et al. demonstrated a significant increase in oxApoA1 in the HDL of participants with diabetes and CVD compared to participants without CVD [110]. Antioxidants Paraoxonase 1 (PON1) hydrolyzes lipoprotein-associated peroxides and lactones. PON1 is mainly synthesized in the liver and in circulation it is associated with HDL. However, PON1 is not a fixed component of HDL since the enzyme could also exert its protective functions outside the lipoprotein environment. It has been demonstrated that HDL transfers PON1 to cell membranes to improve cellular resistance to oxidative stress [111,112]. Previous studies supported a role of PON1 in atheroprotection, through its ability to prevent lipid oxidation and limit atherosclerotic lesion development; moreover, low PON1 activity has been associated with different cardiovascular pathologies, including atherosclerosis and AAA [113][114][115][116][117]. In addition, human population studies have suggested an association of PON-1 polymorphisms with CVD [118]. As mentioned, catalase is one of the most active catalysts that decompose H 2 O 2 at an extremely rapid rate and without consuming cellular reducing equivalents [119]. H 2 O 2 itself is not very reactive; however, the danger of H 2 O 2 comes from its ready conversion to hydroxyl radical by the interaction with a range of transition metal ions, of which the most important in vivo is probably iron. We have recently observed a decrease in catalase, along with SOD and thioredoxin (TRX) reductase, in polimorphonuclear cells (PMNs) of AAA patients compared to controls, which suggest a global decrease in antioxidant enzymes in PMNs under chronic pathological conditions [120]. In contrast, increased catalase immunostaining was shown in AAA tissue [120], which is similar to what is observed in atherosclerotic plaques [121]. Among protein thiol-disulfide oxidoreductases, TRX and PRDX have been widely associated with atherothrombosis. TRX is overexpressed in cells of the vascular wall, probably as a response to high oxidative stress [122]. In contrast, the truncated form, called TRX80, was associated with a pro-inflammatory status and increased atherosclerosis [123]. Moreover, increased expression of TRX has been observed in complicated human atherosclerotic plaques, associated with augmented ROS production and intraplaque hemorrhage [124,125]. Similarly, TRX reductase overexpression is observed in atherosclerotic plaques [126]. Different PRDX isoforms seem to modulate different cellular responses. PRDX1 diminishes leucocyte activation and adhesion to vascular endothelium. Moreover, PRDX-1 was observed in both VSMC and macrophages in human atherosclerotic plaques [127]. Similarly, PRDX-1 and -2 were detected in AAA tissue [128,129], probably as a response to increased oxidative stress [130]. TRX and PRX levels are elevated in plasma from atherothrombotic patients [131,132]. We reported an increase in serum TRX, but also PRX-1, from AAA patients compared with control subjects. Besides, TRX and PRX-1 correlates with AAA size and expansion rate, which suggests that TRX and PRX-1 could be good biomarkers of AAA evolution [128,133]. The increased levels of TRX-1/PRX-1 associated to disease have been suggested to represent a response to increased oxidative stress. In this regard, we recently observed that TRX-1/PRX-1 levels in plasma of asymptomatic subjects correlated with NADPH activity in peripheral blood mononuclear cells [127]. Antioxidants as a Potential Therapeutic Strategy to Prevent Pathological Vascular Remodeling Based on the above findings, antioxidant therapy seems to be a promising alternative for the treatment of atherothrombosis and its associated complications. However, disappointing results have been obtained when comparing results obtained in animal models and in patients. Thus, different antioxidants have in general, prevented, slowed or even reversed atherosclerosis or AAA in animal models. In addition, these findings have been greatly reinforced by the fact that knockout mice on different ROS producing enzymes including NADPH oxidases, or transgenic mice overexpressing detoxifying enzymes including catalase, are partially protected against different processes involved in the atherothrombotic process (see below in this section). To date, no antioxidant drugs have proven effective in the treatment of atherothombotic complications in patients. The majority of trials evaluated the effects of vitamins (mainly vitamin C and E) or folic acid and showed negative results. Among potential explanations, it has been suggested that there were probable differences in oxidative stress in the patients and that they were not assessed for their "oxidative stress status" [6,134]. Other factors such as the type of vitamin given alone or in combination, administration with or without meals, concomitant use with other potential antioxidant drugs, lack of site-specificity or dosage and duration of antioxidant used have also been questioned [8,134]. Other potential antioxidant therapeutic approaches in humans are N-acetylcysteine, an antioxidant precursor of the synthesis of GSH, or nutritional supplements mainly included in the Mediterranean diet such as polyphenols present in extra virgin olive oil, chocolate, red wine or black and green tea, which in general seems to be associated with lower CVD (reviewed in detail by Violi et al. [8]). In this line of evidence, novel antiplatelet and antithrombotic therapies using different antioxidant compounds including flavonoid conjugates, isoquercetine or N-acetylcysteine are being tested [135][136][137]; however, the influence of their antioxidant activity to the antithrombotic and antiplatelet activities remains to be established. Another interesting finding relies on the fact that many commonly used drugs for the treatment of atherosclerosis or cardiovascular risk factors, mainly statins but also angiotensin converting enzyme (ACE) inhibitors or angiotensin receptor AT1 antagonists [8,134,138] show pleiotropic antioxidant effects that might have contributed, at least in part, to the beneficial effects of these drugs on the treatment of atherothrombosis. Thus, atorvastatin acutely inhibits platelet Nox2, platelet isoprostanes and thromboxane A 2 production and this leads to decreased oxidative stress and platelet activation [138]. Moreover, we and others have previously observed antioxidant capacity and inhibition of NADPH oxidase activation by statins in preclinical models [139,140]. Regarding the renin angiotensin system, it is well accepted that Angiotensin II, mainly acting on AT1 receptors, is one of the most important stimulus for oxidative stress both locally and systemically, through the activation of the NADPH oxidase among other mechanisms [5,14,16,[28][29][30]141]. In fact, many studies have shown beneficial effects of ACE and renin inhibitors and AT1 antagonists on oxidative stress in different cardiovascular diseases (reviewed in detail [142,143]). It should be noted that recent evidence demonstrates that the selective AT2 receptor agonist Compound 21 decreases oxidative stress and atherosclerosis in an experimental model of diabetes-associated atherosclerosis [144] and might open new avenues for pharmacological treatment of atherothrombosis. Other drugs such as calcium channel blockers can also have potential antioxidant activities in the context of atherothrombosis [141,145]. However, only interventional well-controlled clinical trials with specific ROS inhibitors or supplement antioxidants will unequivocally confirm the role of ROS in atherothrombosis and the potential beneficial effects of these therapeutic approaches in patients at risk or having cardiovascular events. As mentioned above, animal models have provided important mechanistic information about the role of ROS and ROS-producing enzymes in atherothrombosis [24,25]. More importantly, preclinical animal models have set up the bases for a possible therapeutic effect of antioxidants in the clinic and therefore, although this review has focused on human studies, this aspect will be revised here in more detail. A general antioxidant melatonin ( Figure 3) has been recently proposed as a potential agent for prevention of AAA [146]. More specific inhibitors of ROS have also been tested. Thus, administration of apocynin (antioxidant with some abilities to inhibit NADPH oxidase) attenuates experimental AAA formation and atherosclerosis progression [147,148]. It has also been shown that MPO inhibitor 4-amino benzoic acid hydrazide (4-ABAH) decreased vascular oxidative stress, consecutively improved endothelial function and significantly reduced atherosclerotic plaque development [149]. Very recently, MPO gene deletion attenuates experimental AAA formation [150]. Moreover, oral administration of taurine, an amino acid known to react rapidly with MPO-generated oxidants such as HOCl, also prevented AAA formation, reducing aortic peroxidase activity and aortic protein-bound dityrosine (diTyr) levels [150]. Treatment with the iron chelator desferrioxamine decreases lesion iron concentrations and inhibits atherosclerotic lesion development in the cholesterol-fed rabbit [151]. Iron restriction reduced the incidence of AAA formation with attenuation of oxidative stress and inflammation [62]. one of the most important stimulus for oxidative stress both locally and systemically, through the activation of the NADPH oxidase among other mechanisms [5,14,16,[28][29][30]141]. In fact, many studies have shown beneficial effects of ACE and renin inhibitors and AT1 antagonists on oxidative stress in different cardiovascular diseases (reviewed in detail [142,143]). It should be noted that recent evidence demonstrates that the selective AT2 receptor agonist Compound 21 decreases oxidative stress and atherosclerosis in an experimental model of diabetes-associated atherosclerosis [144] and might open new avenues for pharmacological treatment of atherothrombosis. Other drugs such as calcium channel blockers can also have potential antioxidant activities in the context of atherothrombosis [141,145]. However, only interventional well-controlled clinical trials with specific ROS inhibitors or supplement antioxidants will unequivocally confirm the role of ROS in atherothrombosis and the potential beneficial effects of these therapeutic approaches in patients at risk or having cardiovascular events. As mentioned above, animal models have provided important mechanistic information about the role of ROS and ROS-producing enzymes in atherothrombosis [24,25]. More importantly, preclinical animal models have set up the bases for a possible therapeutic effect of antioxidants in the clinic and therefore, although this review has focused on human studies, this aspect will be revised here in more detail. A general antioxidant melatonin ( Figure 3) has been recently proposed as a potential agent for prevention of AAA [146]. More specific inhibitors of ROS have also been tested. Thus, administration of apocynin (antioxidant with some abilities to inhibit NADPH oxidase) attenuates experimental AAA formation and atherosclerosis progression [147,148]. It has also been shown that MPO inhibitor 4-amino benzoic acid hydrazide (4-ABAH) decreased vascular oxidative stress, consecutively improved endothelial function and significantly reduced atherosclerotic plaque development [149]. Very recently, MPO gene deletion attenuates experimental AAA formation [150]. Moreover, oral administration of taurine, an amino acid known to react rapidly with MPO-generated oxidants such as HOCl, also prevented AAA formation, reducing aortic peroxidase activity and aortic protein-bound dityrosine (diTyr) levels [150]. Treatment with the iron chelator desferrioxamine decreases lesion iron concentrations and inhibits atherosclerotic lesion development in the cholesterol-fed rabbit [151]. Iron restriction reduced the incidence of AAA formation with attenuation of oxidative stress and inflammation [62]. Raising HDLc using both genetic and direct |
infusion models similarly show global anti-atherosclerotic functions of HDL [152][153][154][155]. Similarly, treatment with HDL or fenofibrate inhibits experimental AAA formation and progression [156,157]. ApoA-I or ApoA-1 mimetics reduced or regressed atherosclerosis in animals, altering HDL function (e.g., inhibiting LDL oxidation) without changing HDLc mass [158,159]. Proposed mechanisms include accelerating HDL-mediated cholesterol efflux/reverse cholesterol transport and enhancing HDL's anti-oxidant/anti-inflammatory properties [160]. In advanced aortic root atherosclerotic plaques of apolipoprotein E-deficient mice, native ApoA-I injections led to a significant decreases in lipid content, macrophage number, and an increase in collagen content; in contrast, oxidized human ApoA-I failed to mediate these changes [161]. Interestingly, modulation of HDL functionality by either ApoA1 mimetic D4F or PON1 overexpression decreased AAA formation in mice [95,117,162]. To add to this, PON1 overexpression in ApoE-KO mice displayed smaller atherosclerotic lesions as compared with control mice [163]. In relation to antioxidants, overexpression of catalase suppresses oxLDL-induced aortic smooth muscle cell death [164]. Atherosclerotic mice overexpressing catalase had smaller and relatively early stages of vascular lesions [165]. More recently, mitochondrial oxidative stress was successfully suppressed by catalase overexpression in mitochondria of macrophages or lesional myeloid cells of ApoE −/− mice, and this led to a significant reduction in the aortic root lesional area [166,167]. In addition, catalase overexpression in aortic smooth muscle cells prevented pathological mechanical changes underlying AAA formation [168,169]. In addition, hemoxygenase-1 deficiency aggravates Angiotesin-II induced aortic aneurysms in the ApoE −/− model [170]. Conclusions Atherothrombosis is a very complex pathology that involves, among many other processes, lipid deposition, oxidative stress, inflammatory cell recruitment and platelet activation. Excessive ROS and oxidative stress (likely arising from both increased ROS generation from NADPH oxidase, MPO and iron, and decreased antioxidant systems from PON1 or catalase, among others), play an important role in the initial phases of the disease by inducing endothelial dysfunction (i.e., impaired NO-dependent vasodilation and increased endothelial activation) and by facilitating oxidation of LDL and HDL. In the more advanced stages, RBC-derived iron-rich heme group and leukocyteand platelet-derived oxidants perpetuate the inflammatory process and eventually participate in the rupture of the arterial wall with subsequent platelet aggregation and thrombus formation. Findings from animal models and clinical studies prompted researchers to find oxidative stress markers in tissues and plasma of patients with atherothrombosis. However, although some of these oxidative stress markers predict increased CV risk, none of them have yet been incorporated into clinical practice. Notably, the concept of specificity does not imply their potential use as diagnostic biomarkers in the clinical setting as pathological biomarkers are not specific to a disease, but rather reflect a biological activity associated with pathology. In this respect, oxidative stress is underlying several different diseases and therefore we could observe modified levels of oxidative stress markers in different pathologies, not only in CVD. We have focused this review on markers of oxidative stress associated with lipid/lipoproteins as they are the main drivers of vascular pathology, but it is important to note that oxidized lipoproteins could only be a consequence of increased oxidative stress and just reflect vascular disease without clearly proving their implication in promoting atherothrombosis. In fact, findings in patients treated with different antioxidant therapies are not conclusive despite the overwhelming information on the causative role of ROS in animal models of atherosclerosis and aneurysms. Similarly, we have reviewed the data on antioxidants as they have been studied more globally probably due to the higher stability and easier methodology used to address these questions. In this respect, we envision that novel methodological approaches (e.g., mass-spectrometry) will help to test more specifically the contribution of oxidative stress markers in the mechanisms of human atherothrombosis. In fact, these more specific markers of oxidative stress as surrogate prognostic/therapeutic markers could also potentially give interesting information at the clinical level. Finally, further studies, with more specific ROS inhibitors or antioxidants and carefully designed clinical trials, will probably shed light on the clinical benefits of targeting oxidative stress in CVD and its risk factors. The impact of platelet indices on clinical outcome in heart failure: results from the MyoVasc study Abstract Aims Platelet indices have been associated with traditional cardiovascular risk factors, cardiovascular diseases and all‐cause mortality. This study aimed to investigate the role of platelet count, mean platelet volume (MPV) and platelet‐to‐leukocyte ratio, including platelet‐to‐monocyte and platelet‐to‐lymphocyte ratio with cardiac function, heart failure (HF) phenotypes and clinical outcome, worsening of HF. Methods and results Univariate and multivariable linear and Cox regression analyses were used to investigate the associations between platelet indices, cardiac function and worsening of HF in 3250 subjects enrolled in the MyoVasc study. Higher MPV, lower platelet count, lower platelet‐to‐leukocyte and platelet‐to‐monocyte ratios have been associated with reduced left ventricular ejection fraction (beta estimate [β]MPV [fL] = −0.05 [−0.09; −0.02], βplatelet count (× 10 /L) 9 = 3.4 [1.2; 5.6], βplatelet‐to‐leukocyte ratio = 1.4 [1.1; 1.8], βplatelet‐to‐monocyte ratio = 28 [20; 36]) and increased E/E' ratio (β MPV [fL] = 0.04 [0.003; 0.07], βplatelet count (× 10 /L) 9 = −3.1 [−5.3; −0.92], βplatelet‐to‐leukocyte ratio = −0.83 [−1.2; −0.46], βplatelet‐to‐monocyte ratio = −20 [−28; −12]), independent of age and sex. Cox regression demonstrated an increased risk for worsening of HF in subjects with MPV > 75th percentile (hazard ratio [HR] = 1.47 [1.16; 1.87]), platelet count < 25th percentile (HR = 1.36 [1.07; 1.74]), platelet‐to‐leukocyte < 25th percentile (HR = 1.53 [1.20; 1.95]), platelet‐to‐monocyte < 25th percentile (HR = 1.38 [1.08; 1.77]) and platelet‐to‐lymphocyte > 75th percentile (HR = 1.50 [1.17; 1.93]) ratios, independent of potential confounders. MPV > 75th percentile and platelet count < 25th percentile were strongly related to outcome in HFpEF vs. HFrEF (P for difference = 0.040). Platelet‐to‐leukocyte ratios were associated with worse outcome in both HF phenotypes, without a significant difference between HFpEF and HFrEF. Conclusions Platelet indices are linked with worse cardiac function and adverse clinical outcome, independent of subjects' underlying cardiovascular profile. This study emphasizes their important value to provide additional information on pathophysiology and risk stratification in HF syndrome. Introduction Heart failure (HF) is a major public health problem affecting more than 23 million individuals worldwide. 1 A recent large epidemiological study, including 3 million individuals from Germany with at least two documented HF-related diagnoses, demonstrated a prevalence of 3.96% and an incidence of 655 new cases per 100,000 persons at risk for HF in Germany only. 2 HF is a complex clinical syndrome including unspecific symptoms like shortness of breath and peripheral oedema and thus requires further invasive and non-invasive diagnostic tools. 1 The current HF classification is based on left ventricular ejection fraction (LVEF) into (1) HF with preserved ejection fraction (HFpEF) with signs and symptoms of HF and diastolic abnormalities on echocardiography, (2) HFpEF borderline or HF with mid-range ejection fraction (HFmrEF) with EF of 41-49% and (3) HF with reduced ejection fraction (HFrEF) with EF ≤ 40%. 3,4 Particularly for HFpEF, considerable uncertainty remains regarding its pathogenesis, diagnosis and optimal therapeutic approach. 5 Endothelial dysfunction, inflammation, cardiomyocyte dysfunction and myocardial fibrosis have been implicated as key factors in the development of HF. 6,7 Platelet activation has been described in patients with congestive HF as increased whole blood aggregation, higher mean platelet volume (MPV) and higher expression of platelet bound and soluble P-selectin. 8 Platelet markers including MPV have been associated with traditional cardiovascular risk factors (CVRFs) such as arterial hypertension, diabetes mellitus, obesity, hypercholesterolaemia and smoking, often concomitantly present in HF subjects. [9][10][11] Platelets have an important role as mediators of inflammation, particularly via their interaction with leukocytes. 12 In addition, plateletto-leukocyte ratios, including platelet-to-monocyte and platelet-to-lymphocyte ratios, have been suggested as novel biomarkers to assess systemic inflammation in various conditions. 13,14 Platelet indices and their significance have not yet been comprehensively explored in individuals with HF. This study aimed to investigate the relation of MPV, platelet count and platelet-to-leukocyte ratios with parameters of cardiac function, HF phenotypes and clinical outcome in the MyoVasc study, a cohort of individuals with HF. Study sample MyoVasc is a large epidemiological, prospective, cohort study at the University Medical Center of the Johannes Gutenberg-University Mainz in Germany conceptualized to investigate pathophysiology, diagnostics, clinical course and treatment of HF. 15 Information about inclusion and exclusion criteria of the MyoVasc study were provided in the Supporting Information. Baseline examination of the n = 3289 MyoVasc study participants took place between January 2013 and April 2018. All participants, aged from 35 to 84 years, underwent a comprehensive, highly standardized clinical investigation at the MyoVasc study centre. Platelet indices, measured in fresh blood samples within the routine laboratory at baseline examination, were available in 3250 individuals; n = 294 were controls with normal echocardiographic function ( Figure S1). Written informed consent was obtained from all study participants prior to entering the study. The study complies with the principles outlined in the Declaration of Helsinki, Good Clinical Practice and Good Epidemiological Practice. An approval from the responsible ethics committee (reference number 837.319.12 (8420-F)) and data safety commissioner was obtained in 2012, before study initiation. The MyoVasc study is registered at http://clinicaltrials.gov (identifier: NCT04064450). Assessment of cardiac structure and function Resting two-dimensional transthoracic echocardiograms were performed according to recommendations of the American and European Society of Echocardiography using an iE33 echocardiography system (Royal Philips Electronics, Amsterdam, The Netherlands). 16 The mitral inflow velocity pattern was recorded from the apical four-chamber view with the pulsed-wave Doppler sample volume positioned at the tips of the mitral valve leaflets during diastole in expiration. Peak early (E-wave) and late (A-wave) diastolic filling velocities were measured, and their ratio (E/A) was calculated. The lateral mitral annular early diastolic velocity (E') was measured by spectral tissue Doppler imaging, and the E/E' ratio determined. LVEF was calculated by measurement according to Simpson from the apical four-chamber view. Laboratory assessment Venous blood sampling for the present analysis on platelet indices was performed by using tripotassium ethylenediaminetetraacetic acid (K3-EDTA) tubes. Platelet and leukocyte counts, including monocyte and lymphocyte counts, and MPV were automatically determined within 30-90 min after blood withdrawal on an ADVIA 120 Hematology System (Siemens, Erlangen, Germany) in the central laboratory of the Institute for Clinical Chemistry and Laboratory Medicine, University Medical Center Mainz, Germany. Heart Association (NYHA) functional class ≥ II; (bilateral ankle swelling OR rales OR nocturia) AND N-terminal pro-B-type natriuretic peptide (NT-proBNP) > 125 pg/mL; NYHA Class I AND NT-proBNP > 125 pg/mL AND HF medication. HFpEF was defined as symptomatic HF with PEF, and HFrEF was defined as symptomatic HF with REF. Data assessment and statistical analysis According to these criteria, the analysis sample comprised n = 2111 individuals with PEF, n = 637 with HFpEF. n = 844 individuals were subjects with REF; n = 341 were diagnosed as HFrEF and n = 397 as HFpEF borderline ( Figure S1). HFpEF borderline individuals and not classifiable individuals (n = 343), with HF symptoms and PEF but without diastolic dysfunction, were excluded for those analysis where HFpEF vs. HFrEF was compared. Study outcome was defined as worsening of HF, a composite of transition from asymptomatic to symptomatic HF and cardiac death in asymptomatic HF individuals as well as a composite of hospitalization due to worsening of HF and cardiac death in symptomatic HF individuals. 15 Statistical analysis was performed after data quality control including a review for completeness and plausibility performed by the data management unit. Clinical characteristics of the study sample were described according to quartiles of MPV and platelet count. Additionally, clinical characteristics were presented for the total analysis sample, HFpEF and HFrEF individuals. Normally distributed values were described by using mean ± standard deviation. Categorical variables were expressed as absolute and relative frequencies. MPV, platelet count and platelet-to-leukocytes ratios were assessed by univariate and multivariable linear regression models adjusted for age, sex, cardiovascular risk profile and cancer or age, sex, systolic and diastolic cardiac function (by LVEF and E/E' ratio, respectively) as well as plus antithrombotic medication (ATC B01). Beta estimates for LVEF (%) and E/E' ratio were presented per 1 standard deviation (SD) of the trait. In addition, the distribution of LVEF (%) and E/E' ratio per increasing MPV (fL) or per increasing platelet count (10 9 /L) were depicted as scatter plots. The cardiovascular risk profile comprises CVRFs and cardiovascular diseases (CVDs) as described in the Supporting Information. The distributions of CVD per increasing MPV (fL) or per |
increasing platelet count (10 9 /L) were depicted as boxplots. Outcome data on worsening of HF were depicted as cumulative incidence plots for quartiles of MPV, platelet count, plateletto-leukocyte ratio, platelet-to-monocyte ratio and plateletto-lymphocyte ratio with Grey's test for differences between curves, respectively. A forest plot depicted the relation between platelet indices and worsening of HF, calculated by Cox regression analyses with hazard ratio (HR) and 95% confidence interval (CI) and adjusted for age and sex and additionally for the cardiovascular risk profile and cancer. The difference in worsening of HF between HFrEF and HFpEF was depicted by a cumulative incidence plot. Cox regression analyses were calculated to determine the role of platelet indices in clinical outcome within the phenotypes independent of CVRFs and cancer as well as to determine differences for the roles of platelet indices in HFrEF vs. HFpEF. Furthermore, the roles of antithrombotic agents (ATC B01) and history of cancer on the clinical outcome, worsening of HF, were analysed. Because of the explorative character of the analysis, a significance threshold was not defined for P-values. The P-value should be interpreted as continuous measure of statistical evidence. All statistical analyses were performed using R Version 3.6.0 software (http://www.r-project.org). Clinical characteristics of study participants Clinical characteristics of the study sample at baseline are reported according to quartiles of MPV and platelet count in Table 1 and Table S1. Increasing MPV quartiles were going along with increasing frequencies of individuals with diabetes mellitus, obesity and atrial fibrillation (AF) and history of cancer. Proportions of subjects with REF and HFrEF increased along with increasing MPV quartiles, whereas proportions of subjects with PEF and HFpEF decreased with higher MPV quartiles with the highest prevalence in the lowest MPV quartile (MPV ≤ 7.7 fL). Myocardial infarction (MI), coronary artery disease (CAD) and peripheral artery disease (PAD) showed a U-shape-like distribution with the highest proportions in both lowest and highest MPV quartiles. Individuals in the lowest quartile of platelet count (≤ 186 × 10 9 /L) were older, more male with higher prevalence of diabetes mellitus and dyslipidaemia (Table S1). In addition, MI, CAD, AF, PAD and venous thromboembolism (VTE) were more prevalent in the lowest platelet quartile. The highest frequencies of individuals with history of cancer, chronic kidney disease (CKD) and chronic liver disease (CLD) were present in individuals from the lowest platelet count quartile. Additionally, the distribution of CVD and co-morbidities were depicted in Figure S2A for increasing MPV (fL) and in Figure S2B for increasing platelet count (10 9 /L). The intake of antithrombotic agents (B01) was highest in the lowest platelet count quartile (74.1%) compared with quartiles with higher platelet count with a frequency of antithrombotic intake of less than 60%. The number of individuals with PEF showed an increasing trend with higher platelet counts, whereas subjects with REF showed the opposite relation with a higher proportion of individuals with REF in the lowest quartile of platelet count. Platelet indices and clinical outcome A total of 298 events were registered for worsening of HF during the follow-up period with a median follow-up time of 2.24 years (interquartile range: 1.18-3.97 years). As shown in Figure 1A, the highest quartile (Q4) of MPV (MPV > 8.7 fL, shown in Table S3) was associated with the highest cumulative incidence for worsening of HF compared with Q1-Q3, P-value < 0.0001. Subjects within the lowest quartiles of platelet count (platelets < 186 × 10 9 /L, Figure 1B), platelet-to-leukocyte ratio (platelet-to-leukocyte ratio < 25.8, Figure 1C) and platelet-to-monocyte ratio (platelet-monocyte ratio < 410, Figure 1D) showed a higher cumulative incidence for worsening of HF compared with subjects with higher platelet counts or platelet ratios (P-value platelet count = 0.00012, P-values plateletto-leukocyte and platelet-to-monocyte ratios < 0.0001, respectively). Inversely, the highest quartile of platelet-to-lymphocyte ratio was associated with a higher cumulative incidence for worsening of HF with P-value = 0.0021 ( Figure 1E). Cox regression analysis confirmed the worse outcome in subjects within the highest quartile of MPV in a model ad- Relation of platelet indices and outcome in HF phenotypes Looking into HF phenotypes, a higher incidence for worsening of HF was found among HFrEF individuals compared with HFpEF (P < 0.0001, Figure S4) Table 3. Similarly, the effect of platelet count differed between HFrEF and HFpEF independent of age, sex, CVRFs and cancer (P for difference = 0.0022) with a higher risk for worse outcome in Platelet-to-leukocyte and platelet-to-monocyte ratios of the lowest quartile and platelet-to-lymphocyte ratio above the 75th percentile did not show relevant different effects in HFrEF compared with HFpEF phenotype independent of age, sex, CVRFs and cancer. Whereas the effects on worse outcome of platelet-to-leukocyte and platelet-to-monocyte ratios were higher among HFpEF phenotype, for plateletto-lymphocyte ratio, the association to outcome was stronger in HFrEF phenotype. Table S4 presented the model additionally adjusted for antithrombotic medication (ATC code: B01). Figure 1 Presented are cumulative incidence plots for worsening of HF in the study sample (n = 3220) with a median follow-up time of 2.24 years (interquartile range: 1.18-3.97 years) according to quartiles of MPV (A), platelet count (B), platelet-to-leukocyte ratio (C), platelet-to-monocyte ratio (D) and platelet-to-lymphocyte ratio (E). The intake of antithrombotic agents did not substantially change the associations between platelet indices and risk for worsening of HF. In addition, to investigate if cancer history modifies the association with worsening of HF, an analysis excluding subjects with cancer history was performed in comparison with the whole sample that included subjects with cancer history. The subgroup without cancer history with MPV > 75th percentile and platelet-to-lymphocyte ratio > 75th percentile showed a higher risk for worsening of HF compared with the complete sample but lower risk for worsening of HF with platelet count < 25th, platelet-to-leukocyte ratio < 25th percentile or platelet-to-monocyte ratio < 25th percentile independent of age, sex, antithrombotic agents, CVRFs and co-morbidities (Table S5). Discussion This study investigated several platelet indices like MPV, platelet count and platelet-to-leukocyte ratio, including platelet-to-monocyte and platelet-to-lymphocyte ratio, in relation to cardiac function and clinical outcome in HF individuals. Higher levels of MPV were associated with reduced LVEF, a measure of systolic dysfunction, and increased E/E', a measure of diastolic dysfunction, independent of age and sex. In the same line, with opposite direction only, were the findings for the relation between platelet count and cardiac function measurements. The highest MPV quartile and the lowest quartile of platelet count were characterized by worse cardiovascular risk profile with higher frequencies of diabetes mellitus, CAD, AF, CKD and CLD. Higher MPV, a potential Figure 2 Forest plot presenting the association of MPV > 75th percentile, platelet count < 25th percentile, platelet-to-leukocyte ratio < 25th percentile, platelet-to-monocyte ratio < 25th percentile and platelet-to-lymphocyte ratio > 75th percentile and worsening of HF with hazard ratios (HRs) with 95% confidence interval (CI), adjusted for age and sex and additionally adjusted for CVRFs and cancer in n = 3188 individuals (298 events); Cardiovascular risk factors (CVRFs) are arterial hypertension, diabetes mellitus, smoking, obesity, dyslipidaemia, family history of myocardial infarction/stroke, myocardial infarction, stroke, coronary artery disease, atrial fibrillation, peripheral artery disease and venous thromboembolism; MPV, mean platelet volume. Cox regression analysis in n = 951 individuals for the association between mean platelet volume (MPV) > 75th percentile, platelet count < 25th percentile, platelet-to-leukocyte ratio < 25th percentile, platelet-to-monocyte ratio < 25th percentile or platelet-to-lymphocyte ratio > 75th percentile and worsening of HF (n = 174 events). Results are presented as hazard ratios (HRs) with 95% confidence interval (CI). In addition, differences for the effects of platelet indices for HFrEF vs. HFpEF were calculated. Cardiovascular risk factors (CVRFs) are arterial hypertension, diabetes mellitus, smoking, obesity, dyslipidaemia and family history of myocardial infarction/stroke. marker of platelet activation, 18,19 has previously been associated with traditional CVRFs and CVDs, particularly with diabetes mellitus, obesity and AF. [9][10][11]20,21 High levels of MPV have also been described in the setting of HF. 22 In a large adult population-based cohort, the relation between higher MPV and increased all-cause mortality was independent of traditional CVRFs. However, this relation was lost after adjusting for CVDs including HF, suggesting for a possible role of HF in the association between MPV and total mortality. 9 MPV has been reported to be associated with higher thrombin generation potential assessed in presence of platelets, particularly among individuals at risk for CVDs. 23 In addition, higher MPV was correlated with a higher percentage of platelets expressing surface P-selectin, another recognized marker of platelet activation. 3,8,20,23 The present results support an important role of platelets in HF pathophysiology and HF-related outcome in both HF phenotypes. The overall incidence of worsening of HF was higher among HFrEF compared with HFpEF, but with respect to platelet indices, higher MPV and lower platelet count showed a stronger effect on worse outcome in HFpEF phenotype. CVRFs and cancer did not substantially change the association between platelet indices and clinical outcome, even though the cardiovascular risk profile and laboratory parameters differed between HF phenotypes and co-morbidities have been shown to modulate platelet activation. 9,11,24,25 The risk for worsening of HF remained higher independent of intake of antithrombotic agents. Individuals without cancer history with higher MPV and/or higher platelet-to-lymphocyte ratio had even higher risk for worsening of HF compared with the total analysis sample including subjects with cancer history. This finding could potentially speak for the benefits of regular, closer follow-up of cancer patients for developing cardiovascular complication with particular consideration for the cardiovascular toxicities from cancer treatment. 26 In addition to the underlying cardiovascular risk profile, HF specific features such as haemodynamic and vascular changes including cardiac remodelling could also have an impact on platelet characteristics. 22 Platelets are recognized mediators of inflammation, particularly through their interaction with leukocytes and endothelial cells. [27][28][29] Increased release of cytokines and catecholamines observed in severe HF has been associated with platelet activation and higher levels of MPV. 22 Platelet ratios to leukocytes, to monocytes and particularly to lymphocytes have been reported as novel markers of inflammation and were linked to total mortality. 13 This study demonstrated that both platelet-to leukocyte and platelet-to-monocyte ratios have important associations to cardiac function parameters such as LVEF and E/E' that remained independent of age, sex and antithrombotic agents. However, a role of age and/or sex was observed for the associations to cardiac function parameters. Lower ratios were associated with worse systolic and diastolic function. Differently, a positive trend between platelet-to-lymphocyte ratio and LVEF, but no relation to E/E' ratio, has been also observed in models adjusted for age and sex. Higher MPV has been associated with increased mortality after MI, a strong risk factor for HFrEF, 5 whereas lower platelet count has been associated with increased risk of total, cancer and non-cardiovascular/non-cancer mortality but was unrelated to cardiovascular mortality. 9,11 Interestingly, this study showed that higher MPV and lower platelet count were more related to clinical outcome in HFpEF compared with HFrEF independent of CVRFs and cancer. For platelet-to-leukocyte ratios, including platelet-to-monocyte and plateletto-lymphocyte ratios, no differences for the risk prediction of worsening of HF have been found between HFpEF and HFrEF. Lower platelet-to-leukocyte ratio and plateletto-monocyte ratio showed an important trend towards worse clinical outcome particularly for HFpEF phenotype, as observed for MPV and platelet count. Increased leukocyte count has been associated with adverse clinical outcome in HFpEF subjects. 30 In this study, fibrinogen levels and leukocyte count were observed higher in HFrEF individuals compared with HFpEF. Lower platelet-to-leukocyte ratios resulting from higher leukocyte counts contribute to a proinflammatory state in HF that may promote activation of platelets and coagulation system in both phenotypes. An activation of the unspecific immune response in individuals with worse cardiac function could be anticipated, as C-reactive protein (CRP), fibrinogen and leukocyte count were higher in both symptomatic HF phenotypes compared with the rest of the analysis sample. Furthermore, due to the release of a plethora of inflammatory mediators by activated platelets, the inflammatory state in HF individuals could be further potentiated. 31 Higher platelet-to-lymphocyte ratio showed a stronger trend for worsening of HF among HFrEF subjects compared with HFpEF, independent of the underlying cardiovascular risk profile. Recent studies in acute HF individuals reported different results for the association of platelet-to-lymphocyte ratio and long-term mortality as independent predictor of outcome in acute HF. 13,14,32 In this study, within the highest quartile of |
platelet-to-lymphocyte ratio, HFrEF individuals showed a 2.65-fold increased risk and HFpEF individuals 1.56-fold increased risk for worsening of HF, indicating an important role for high platelet-to-lymphocyte ratio as a biomarker of clinical outcome related to reasons other than worse systolic and diastolic function. Strengths and limitations The major strength of this study is the comprehensive, highly standardized clinical investigation and follow-up of a large sample of individuals with HF syndrome. However, there are some limitations that should be considered: Despite the observed important links between platelet indices and HF, this study was not design to investigate a causal relationship. The impact of platelet indices on clinical outcome in heart failure: results from the MyoVasc study Furthermore, the lack of detailed information on the type and stage of cancer prevented us to investigate more in details the role of cancer history on the association with platelet indices and HF outcome. Further mechanistic studies are warranted to clarify the role of platelets as cause or result of HF pathophysiology and their role in the HF-related pathological response. Nevertheless, platelet indices were associated with measures of systolic and diastolic function, as well as with clinical outcome in HF individuals. According to the guidelines, HF is divided into three phenotypes: HFpEF, HFpEF borderline and HFrEF. 4 This study analysed only HF phenotypes with preserved and REF but excluded individuals with EF of 41-49%. The role of platelets in HFpEF borderline individuals needs to be further investigated as this phenotype presented with partial characteristics of HFpEF and some HFrEF properties. 4 Conclusion In conclusion, this study supports a role for platelets in the pathogenesis of HF demonstrating an important link to the clinical outcome in HFpEF and HFrEF phenotypes. Better characterization of platelet function is warranted to increase the knowledge on platelet-related molecular mechanisms involved in HF-related inflammation, especially in HFpEF phenotype, as well as to understand further if these biomarkers help to identify HF patients at risk for worse clinical outcome. Table S1. Characteristics of the study sample according to quartiles of platelet count (N = 3,250). Table S2. Characteristics of study participants (N = 3,250) and according to HFpEF (N = 637) and HFrEF (N = 341). Table S3. Quartiles of platelet indices. Table S4. Relation between platelet indices and worsening of HF in HF phenotypes with additional adjustment for Antithrombotic medication (ATC code: B01). Table S5. Relation between platelet indices and worsening of HF in total analysis sample and after excluding individuals with history of cancer. Figure S1. Derivation of the analysis sample. Figure S2. Boxplots for the distribution of cardiovascular diseases and comorbidities. Figure S3. Scatter plots of LVEF and E/E' per increasing MPV (fL) or per increasing platelet count (10 9 /L). Figure S4. Cumulative incidence plot for worsening of HF in HFrEF and HFpEF. Characterization of Plasmodium falciparum Adenylyl Cyclase-β and Its Role in Erythrocytic Stage Parasites The most severe form of human malaria is caused by the parasite Plasmodium falciparum. The second messenger cAMP has been shown to be important for the parasite’s ability to infect the host’s liver, but its role during parasite growth inside erythrocytes, the stage responsible for symptomatic malaria, is less clear. The P. falciparum genome encodes two adenylyl cyclases, the enzymes that synthesize cAMP, PfACα and PfACβ. We now show that one of these, PfACβ, plays an important role during the erythrocytic stage of the P. falciparum life cycle. Biochemical characterization of PfACβ revealed a marked pH dependence, and sensitivity to a number of small molecule inhibitors. These inhibitors kill parasites growing inside red blood cells. One particular inhibitor is selective for PfACβ relative to its human ortholog, soluble adenylyl cyclase (sAC); thus, PfACβ represents a potential target for development of safe and effective antimalarial therapeutics. Introduction Malaria remains a major burden in the developing world, causing approximately 1 million deaths per year. It is a vectorborne disease caused by protozoan parasites of the genus Plasmodium, the most lethal of which is Plasmodium falciparum. A diverse array of protozoal, fungal, and bacterial pathogens, including Plasmodium spp., depend upon the ubiquitous second messenger cyclic adenosine monophosphate (cAMP) for survival and environmental sensing [1]. In fact, two stages of the Plasmodium life cycle appear to depend upon cAMP: Sporozoites require cAMP generation for host cell invasion [2], and previous reports suggest that cAMP effectors play an important role in the asexual red blood cell stage of the life cycle. Specifically, inhibition of cAMP-catabolizing phosphodiesterases (PDEs) or addition of membrane-permeable cAMP analogs increase the percentage of schizonts in asynchronous, erythrocytic cultures of P. falciparum [3], and treatment of erythrocytic stage cultures with either pharmacological or genetic inhibitors of the main effector of cAMP, Protein Kinase A (PKA), inhibit growth [4,5]. While these data reveal that the cAMP pathway is required for progression through the erythrocytic, asexual stage of the life cycle, the stage of the life cycle that causes symptomatic malaria, it remains unclear how cAMP levels are controlled during this period. cAMP is synthesized by adenylyl cyclases (AC), and the P. falciparum genome encodes two such enzymes, PfACa and PfACb. Both enzymes contain class IIIB catalytic domains similar to mammalian soluble adenylyl cyclase (sAC) [6]. Mammalian sAC is structurally, molecularly, and biochemically distinct from other mammalian adenylyl cyclases, which are transmembrane proteins regulated by heterotrimeric G proteins (tmACs). Unlike tmACs, mammalian sAC is directly regulated by bicarbonate. In physiological systems, bicarbonate is in nearly instantaneous equilibrium with CO 2 and intracellular pH (pHi) due to the action of carbonic anhydrases [7]; thus, mammalian sAC serves as a physiological CO 2 /HCO 3 2 /pHi sensor [8,9], with specific roles in sperm activation [10,11], ciliary beat frequency in bronchii [12], pH homeostasis in epididymis [13], kidney [14,15], and shark gill [16], metabolism [17], and aqueous humor formation in the eye [18]. PfACa and PfACb differ in their modular architecture. PfACa . contains six predicted transmembrane domains and a single carboxy-terminal catalytic domain homologous to sAC-like ACs. The motifs required for metal cofactor binding, substrate binding, and catalysis are contained within this single catalytic domain, suggesting that this enzyme functions as a homodimer [19]. In contrast, PfACb has no predicted transmembrane regions and possesses two sAC-like AC catalytic domains. PfACb and ACb orthologs from other Plasmodium spp. possess all the motifs required for catalytic activity, but they are spread across the two presumptive catalytic domains suggesting that catalysis requires intramolecular heterodimerization, similar to mammalian sAC [20]. In addition, these ACs possess a threonine residue which is thought to be predictive for bicarbonate regulation in sAC-like . Samples were collected in triplicate. Luciferase activity is elevated between 4-16 hours due to increased promoter activity during primary round of infection. The peak of luciferase activity seen at ,44 hr under normal culture conditions, but absent in the absence of CO 2 /HCO 3 2 or presence of KH7, reflects reinvasion into RBCs. The graph was prepared with Prism software; error bars represent s.e.m of triplicate wells in the representative experiment. (D) Microscopic evaluation of Giemsa-stained parasites at 44 hr reveals parasites (P) maintained in normal culture completed mitosis and newly released merozoites are poised to reinvade new RBCs. Parasites treated with KH7 (E) or grown in low CO 2 /HCO 3 2 conditions (F) never form schizonts. doi:10.1371/journal.pone.0039769.g001 ACs [21]. Unlike other adenylyl cyclases including ACb orthologs from other Plasmodium spp., each catalytic domain of PfACb is interrupted by blocks of highly charged stretches of amino acids, which are encoded by low complexity regions of unknown function prevalent throughout the P. falciparum genome. PfACa has been studied both in vivo and in vitro. PfACa is a predicted bifunctional protein comprising both a K + channel and an AC that is conserved in alveolata protozoans [22]. PfACa transcripts are abundant in sexual stage gametocytes [19], suggesting a possible role during sexual stages. Additionally, ACa proteins in Plasmodium spp. appear to play a role during the liver sporozoite stage. Specifically, P. berghei sporozoites deficient in ACa were shown to have reduced infectivity of cultured hepatocytes and reduced liver infectivity in a mouse model, but they were viable and exhibited normal growth during asexual, erythrocytic growth [2]. In contrast, PfACb has not yet been heterologously expressed or biochemically characterized, and attempts to generate PfACb-deficient parasites using protocols that demand growth of the haploid mutant parasite in erythrocyte cultures were repeatedly unsuccessful [2]. Interestingly, its mRNA is highly expressed during the erythrocyte stage; PfACb transcript levels begin to increase in the trophozoite stage and peak during schizogeny [23,24]. We took advantage of a number of small molecule inhibitors of sAC-like adenylyl cyclases to identify the essential source of cAMP during erythrocytic growth. Three distinct AC inhibitors blocked growth of P. falciparum inside red blood cells. We established conditions for in vitro characterization of PfACb . and we tested sensitivity of these three inhibitors against the in vitro AC activities of both PfACa and PfACb. Consistent with the differential expression patterns of the two cyclases, only PfACb proved to be sensitive to all three, providing strong evidence that it is the source of cAMP essential during erythrocytic growth. Interestingly, one of the three inhibitors was also selective for PfACb relative to mammalian sAC demonstrating that small molecules can distinguish between the parasite and host enzymes. These data define PfACb as a target for development of novel antimalarial therapeutics. Results and Discussion We have identified two, structurally distinct inhibitors of sAClike ACs; catechol derivatives of estrogen and KH7 ( Figure S1). Catechol estrogens (CEs), such as 2-hydroxyestradiol (2-CE), inhibit Class III ACs, including mammalian and bacterial sAC-like ACs, by chelating the catalytic magnesium ion in the active site [25]. The second structurally unrelated inhibitor, KH7, was identified as a potent, specific inhibitor of mammalian sAC [11,26,27] in a small molecule screen [11] and was subsequently found to inhibit a number of bicarbonate-sensitive ACs [16,28]. To determine the effect of these compounds on parasite growth and viability inside red blood cells, we measured the luminescence of the wild-type NF54 P. falciparum strain transfected with the pHLIDH plasmid, which constitutively expresses firefly luciferase [29]. The luminescence of this parasite strain directly corresponds to the measures of viability determined with the widely-used tritiated hypoxanthine-uptake assay [30] ( Figure S2). Both KH7 and 2-CE killed rapidly ( Figure 1A,B) [LD 50 = 8.5 mM (95% C.I. = 7.8-9.2 mM) for KH7 and 60 mM (95% C.I. = 43-90 mM) for 2CE] with death observed within a single replicative cycle (48 hours) of synchronized parasites ( Figure 1C). Giemsa-stained slides prepared from parasites treated with KH7 revealed condensed, pyknotic parasites ( Figure 1E), confirming that these compounds lead to rapid parasite death rather than simply inhibiting proliferation or reporter activity. As a reference, the terminal phenotype of KH7-killed parasites was indistinguishable from that of parasites maintained in the absence of CO 2 /HCO 3 2 . Synchronized cultures grown in CO 2 / HCO 3 2 in the presence of the inhibitor KH7 or grown in the absence of CO 2 /HCO 3 2 lacked the burst of luciferase due to the reinvasion observed in normal cultures ( Figure 1C,D). Microscopic evaluation confirmed that the drug-treated parasites ( Figure 1E) resembled dead CO 2 /HCO 3 depleted parasites ( Figure 1F); neither formed merozoites, indicating they had not completed schizogeny. In addition, we tested KH7 against a chloroquineresistant P. falciparum strain (Dd2), and it was lethal, as determined microscopically, with similar efficacy as observed against the chloroquine-sensitive NF54 strain (data not shown). In order to determine the temporal effect of KH7 on synchronized parasites, we added KH7 to synchronized cultures at different time points throughout the cell cycle ( Figure 2A). Addition of KH7 in the first 24 hours of the cell cycle led to complete cell cycle arrest. However, if KH7 was added to the culture at a point well into schizogeny (34 hours), parasites were able to complete the cell cycle and invade new erythrocytes. In a complementary experiment to determine a ''window of KH7sensitivity,'' synchronized cultures were incubated in the presence of KH7 for various times, at which point the drug was washed out and cultures were grown for the remainder of a 48-hour cell cycle. When KH7 was removed at 24 hours or before, cultures were able to progress through the cell cycle, reinvade |
erythrocytes, and enter G1 ( Figure 2B). If KH7 remained on cultures past 24 hours, parasites appeared unable to recover within the 48-hour culture period. These data demonstrate that parasites are most sensitive to KH7 at 24-31 hours post-invasion. This corresponds to the period in the cell cycle during which PfACb mRNA levels are beginning to rise dramatically ( Figure S3). We next sought to determine whether the in vitro activities of PfACa and/or PfACb were sensitive to 2-CE and KH7. PfACa has been heterologously expressed and characterized previously [22], but the in vitro activity of PfACb has not yet been demonstrated. We expressed a synthetic gene encoding the catalytic domains of PfACb . AA 1-785) with mammalian codon usage as a fusion protein with a carboxy-terminal glutathione-Stransferase (GST) using a baculovirus (BV) expression system. GST-PfACb 1-785 was soluble, and we were able to purify it only under high salt conditions ( Figure S4). This high salt requirement for GST-PfACb 1-785 solubility may be due to the blocks of charged amino acids inserted into its catalytic domains. Similar to other sAC-like ACs [21,31,32,33,34], including PfACa [22], which exhibit much greater activity using Mn 2+ -ATP as a substrate relative to Mg 2+ -ATP, purified GST-PfACb 1-785 was active in the presence of Mn 2+ -ATP ( Figure 3A). We were unable to detect measurable activity in the presence of Mg 2+ -ATP ( Figure 3B). A similar Mn 2+ -ATPdependency was observed in assays of AC activity in erythrocytic stage P. falciparum lysates [35]. GST-PfACb 1-785 displayed Michaelis-Menten kinetics with a lack of cooperative binding of substrate at the active site ( Figure 3A). The enzyme has an apparent Michaelis constant (Km) for substrate ATP of ,0.6 mM using Mn 2+ as a cofactor with a maximum reaction velocity of ,265 nmol cAMP/min/mg. This Km value is similar to that obtained for human sAC (0.9 mM) [34]. The optimal ratio of divalent cation (Mn 2+ ) to substrate (ATP) was 4:1 ( Figure 3B), similar to mammalian sAC [34], and GST-PfACb 1-785 displayed minimal ability to produce cGMP when supplied with GTP as substrate (data not shown). Mammalian sAC is directly regulated by bicarbonate [34,36] and calcium [34,37], and the threonine residue thought to be predictive of bicarbonate stimulation [21] is found in PfACb and ACb orthologs from other Plasmodium spp. However, because bicarbonate precipitates in the presence of Mn 2+ , and because we found bicarbonate and calcium activation to be unique to Mg 2+ -ATP-dependent activity in mammalian sAC, we were unable to explore bicarbonate-or calcium-responsiveness of BV-expressed GST-PfACb 1-785 . Instead, we explored the pH responsiveness of GST-PfACb 1-785 . In contrast to mammalian sAC, GST-PfACb 1-785 exhibited a strong pH dependence ( Figure 3B) [36]. Varying the reaction pH from 7 through 9 revealed a pH optimum of 7.5, and activity decreased sharply at both higher and lower pH values. Thus, PfACb activity will be sensitive to changes in pHi, which, in physiological systems, is dependent upon the carbonic anhydrase-mediated equilibrium between CO 2 , bicarbonate, and protons. It is important to note that the pH dependence observed for PfACbis strikingly similar to the pH dependence of P. falciparum in culture. When pH of growth media is maintained between 7.1 and 7.5, parasitemias increase 20-30 fold after three days, with sharp reductions in yield outside of this pH range [38]. During the trophozoite stage, when PfACb mRNA is first expressed [23,24] (Figure S4), the intracellular pH (pHi) of parasites is approximately 7.3 [39,40,41]. Therefore, we speculate that PfACb functions as the parasite's pH sensor during growth inside red blood cells. GST-PfACb 1-785 activity was inhibited by both KH7 and 2-CE with affinities that reflect their observed efficacies in culture. KH7 inhibited GST-PfACb 1-785 with an IC 50 of 5 mM, and 2-CE showed inhibition with an IC 50 of 8 mM ( Figure 4A,B). In contrast, although PfACa adenylyl cyclase activity was inhibited by 2-CE, it was largely insensitive to KH7 ( Figure 4C). Thus, among adenylyl cyclases in P. falciparum, only PfACb is inhibited by the two structurally unrelated inhibitors which kill parasites in erythrocytic cultures. While these data suggest PfACb may be a relevant target for killing malaria parasites inside red blood cells, both KH7 and 2-CE are also known to inhibit mammalian sAC, leaving open the possibility that host red blood cell sAC may be the relevant target of these compounds. To address this concern, we sought to identify a PfACb selective inhibitor. During our screen to identify KH7 as a mammalian sAC inhibitor, we tested numerous KH7-like compounds ( Figure S1). Most of the KH7-like compounds were ineffective against sAC-like cyclases, and these proved to have little effect on P. falciparum growth ( Figure S5). However, one KH7-like compound, KH7.15, which is inert against mammalian sAC [27], inhibited GST-PfACb with an IC 50 of 150 mM ( Figure 4D). KH7.15 killed parasites ( Figure 4E) with a similar efficacy [LD50 = 67 mM (95% C.I. 58-78 mM)] as it inhibited PfACbactivity in vitro. The fact that parasites were killed by two structurally unrelated inhibitors (2-CE and KH7) and by a third inhibitor (KH7.15) selective for PfACb relative to both PfACa and to the host sAC suggest that PfACb is the relevant target of these compounds and is essential for parasite growth inside red blood cells. Our data include the first characterization of PfACb and suggest that PfACb is essential for erythrocytic-stage parasite viability. We have demonstrated PfACb is biochemically distinct from other Class IIIb adenylyl cyclases and exhibits significant pH-sensitivity. Additionally, we have shown that small molecule inhibitors can distinguish PfACb from mammalian sAC. Although the profile of KH7.15 is not ideal for clinical use, the data presented here provide proof-of-principle that PfACb can be selectively targeted, thereby identifying it as a therapeutic target for a new class of antimalarial drugs. Although effective pharmacological therapies for malaria exist, the widespread and expanding resistance to these drugs demands new approaches to therapeutic intervention. The spread of multidrug resistant strains of P. falciparum threatens to increase the malaria burden, and novel therapeutics to combat malaria are desperately needed. This work is an initial step in attempts to address that need by defining PfACbas a novel, attractive therapeutic target. Parasite Culture and Microscopy Compounds The parasite strains NF54 and NF54 transfected with pHLIDH were grown in 5% hematocrit in RPMI 1640 (Invitrogen/Life Technologies) supplemented with 0.5% Albumax II (Invitrogen/ Life Technologies), 0.25% sodium bicarbonate (standard media), and 0.01 mg/ml gentamycin. Human red blood cells for culture were obtained from human volunteers, cleared of leukocytes by passage through a Sepacell R-500 column (Baxter Health Care), and washed three times in RPMI 1640. Parasites were grown in sealed culture flasks under an atmosphere of 90% nitrogen, 5% oxygen, and 5% carbon dioxide. Parasitemias were maintained between 1 and 10%. Fixed parasites were stained with Giemsa to allow microscopic analysis of cultures using an Olympus BX40 compound microscope. A synthetic gene with mammalian codon usage was used as the template. A 4-nucleotide addition was included in the FWD primer for directional topoisomerase-based cloning, and a stop codon was included in the REV primer. Following the PCR reaction, fragments were resolved on a 1% Agarose gel. Bands corresponding to the appropriate size were excised and fragments were gel-purified (Qiagen gel purification kit). After quantification by gel electrophoresis and comparison to a High Mass Ladder (Invitrogen), 10 ng of each fragment was used in a 2-hr topoisomerase-based cloning reaction with pENTR/TEV-D-TOPO (Invitrogen). Two microliters of the cloning reaction was transformed into TOP10 E. coli (Invitrogen). Colonies were screened by restriction digest, and positive clones were sequenced using M13 forward and M13 reverse primers and multiple genespecific primers. Clones found to be correct by sequencing were subsequently recombined into the ''destination'' vector pDEST20 (N-terminal GST tag) using a 1-hr LR Clonase II recombination reaction (Invitrogen). pDEST20-PfACb plasmid was transformed into DH10Bac E. coli (Invitrogen). Transformed bacteria were plated onto LB agar plates containing 50 mg/mL kanamycin (Sigma-Aldrich), 7 mg/ mL gentamicin (Sigma-Aldrich), 10 mg/mL tetracycline (Sigma-Aldrich), 100 mg/mL Bluo-gal (Invitrogen), and 40 mg/mL isopropyl-b-D-1-thiogalactopyranoside (Sigma-Aldrich). White colonies, indicative of successful bacmid recombination, were picked and streaked on fresh plates to confirm the phenotype. Blue colonies were streaked on a separate area of the same plate as a control. Confirmed white colonies were cultured in 500 mL of LB containing 50 mg/mL kanamycin, 7 mg/mL gentamicin, and 10 mg/mL tetracycline. Subsequently, bacmid DNA was isolated from the cell pellet using the NucleoBond Bac 100 DNA isolation kit (Macherey-Nagel). Isolated bacmid DNA was immediately transfected into Sf9 cells plated at ,80% confluency on a 6-well plate (Becton-Dickenson) using Cellfectin reagent (Invitrogen). After transfection, successful recombination of bacmid DNA was confirmed by PCR analysis using M13 forward (Invitrogen), M13 reverse (Invitrogen), and the PfACb FWD primer indicated above. Four days post-transfection, cells showed significant signs of baculovirus infection. Cell media containing recombinant baculovirus was harvested and clarified by centrifugation at ,1,0006g. This P1 baculovirus stock was amplified first in a volume of 20 mL (400 mL P1 baculovirus was added) and subsequently in a volume of 500 mL (10 mL P2 baculovirus was added). For expression studies, 25 mL P3 baculovirus was added per liter of insect cells (either Sf9 or Hi-Five). Heterologous Protein Expression Insect cells are a proven system for expression and characterization of adenylyl cyclases [42]. Hi-Five cells at a density of 1610 6 cells/mL were infected with GST-PfACb 1-785 baculovirus at a concentration of 25 mL P3 baculovirus/L of culture. Infected cells were cultured for 40 hrs and harvested by centrifugation at ,10006g. Cells were frozen in liquid nitrogen and stored. Frozen pellets were resuspended in lysis buffer containing 50 mM Tris (pH 7.5), 5 mM DTT, 2 M NaCl, 10 mg/mL aprotinin/leupeptin, 1 mM PMSF, 1 mM benzamidine, 10 mM b-mercaptoethanol at a ratio of ,10 mL lysis buffer/100 mL of pelleted culture. This lysate was sonicated 5 times at 10-second intervals at 12 watts with a Misonix Microson cell disruptor. The sonicated lysate was clarified by centrifugation at 100,0006g using a Ti-75 rotor (Beckman). The resulting supernatant was passed over a Superdex G-25 column with a 5-mL bed volume for further clarification. The clarified lysate was incubated on ice with minor agitation for 1 hr with 1 mL (packed volume) of glutathione sepharose 4B (Amersham) per 100 mL of lysate. The lysate was allowed to flow through, and the resin was washed with 3610 bed volumes of lysis buffer. Finally, bound protein was eluted with 15 mM reduced glutathione in lysis buffer in 1 bed volume fractions. PfACb protein was detected by activity and anti-GST Western blot (data not shown). PfACa pressed as previously described [22]. Radioactivity-based Two-Column Adenylyl Cyclase Assay Adenylyl cyclase assays with purified PfACb and PfACa were performed according to the method of Salomon [43]. Purified GST-PfACb . 50-500 ng) was incubated in 50 mM Tris, pH 7.5 (unless otherwise indicated), 1 mM DTT, 300 mM NaCl, 10 mM MnCl 2 and 2.5 mM ATP (unless otherwise indicated) with ,1,000,000 cpm [a-32 P]ATP (Perkin Elmer) and ,5,000 cpm [ 3 H]cAMP (Perkin Elmer). (Tris buffers were pH-adjusted at room temperature for use at 37uC.) Reactions were performed in 100 mL for 20 minutes at 37uC and stopped with 150 mL 1.5% SDS. Product [ 32 P]cAMP was separated from substrate [a-32 P]ATP by sequential column chromatography over dowex 50WX4-400 resin (Fluka) followed by aluminum oxide resin (Sigma). Product [ 32 P]cAMP was eluted from dowex, directly onto the alumina by water, and the cAMP was eluted from alumina by 0.1 M imidazole, pH = 7.3. Viability Assays The NF54 strain transfected with pHLIDH expresses the firefly luciferase gene under the control of the constitutively active Hrp3 promoter [44]. This strain of parasites was created by transfection and stable integration of the plasmid pHLIDH into the genome of the NF54 wildtype parasite line. pHLIDH is a derivative of the pHLH-1 plasmid [44], in which the drug selectable marker hdhfr was inserted under the control of the PcDT59 promoter [45]. Parasites were plated on day 0 at 1% parasitemia in 96-well plates in standard media in the presence of the indicated concentrations of DMSO (vehicle control), KH7, 2-CE, or KH7.15. Media plus compounds were replenished |
on day 1. On day 2, red blood cells were lysed with Bright-Glo Lysis Buffer (Promega), and luminescence was read using a luminometer (Molecular Devices) after injection with 10 ml Bright-Glo Luciferase Reagent (Promega) for a 2-sec integration time and a 15-sec read time. Data shown are normalized to the luminescence of vehicle-treated control parasites. Parasite Synchronization NF54 parasites were synchronized as described [46]. Briefly, parasites in cultured RBCs were centrifuged for 4 min at 4000 rpm. The pellet was layered atop a 40%/70% Percoll-Sorbitol gradient and centrifuged for 20 min at 10,000 rpm. The late-stage fraction at the interface of the gradient was collected, washed in media, and reconstituted with fresh RBCs and media. Following erythrocyte invasion, the synchronized culture was expanded into 6 20-mL cultures at 3% parasitemia. At each indicated time point, one 20-ml culture was centrifuged for 2 min at 4000 rpm. The pellet was resuspended in 500 ml phosphatebuffered saline (PBS), and RBCs were lysed with 10 ml 10% saponin and microcentrifuged for 2 min at 13,000 rpm. The supernatant was aspirated, and pellets were frozen at 280uC until all time points were collected. Protection of Human Subjects Blood was purchased from the New York City Blood Center or obtained from healthy human volunteers for use in parasite culture. A protocol for acquisition and use of human blood has been approved and is on file with the Internal Review board at Weill Medical College of Cornell University (Protocol #0010004662). For blood purchased from the New York City Blood Center (NYBC), contact of blood donors will not be attempted and is not necessary for the livelihood of the study. Informed consent is not required (other than NYBC in-house protocol). The blood will be used for research purposes onlysolely for in vitro culture of Plasmodium falciparumand not for transfusion into humans or animals. NYBC policy states that only surplus blood will be made available for research purposes, and thus this study will not compromise blood supplies. Blood will be used for research purposes only -solely for in vitro culture of Plasmodium falciparum -not for transfusion into humans or animals. The blood purchased from NYBC will only be used as a resource for propagation of malaria parasites and no data will be collected with regard to the blood itself. Therefore the inclusion of women, minorities or children is not applicable. Ethics Statement Blood used in parasite cultures was obtained under a protocol approved by and on file with the Internal Review board at Weill Medical College of Cornell University or at New York Blood Center. All donors gave prior written consent. Figure S1 Structures of compounds used in this study. Parasite viability with measured with the luciferase-based (yellow curves) or tritiated hypoxanthine-based viability assay (red curves) in the presence of increasing concentrations of chloroquine (A), quinine (B), mefloquine (C), and artemisinin (D). Best-fit curves are shown. Y-axis is percentage assay readout; X-axis is log 10 drug concentration. EC 50 s for each drug are shown below the figure. Supporting Information Best-fit curves are highly similar for each drug. (TIFF) Figure S3 Expression levels of PfACb in the red blood cell. RT-PCR using PfACb-specific primers confirms publicly available microarray data [23,24]. Both primer sets 1 (blue bars) and 2 (red bars) amplify high levels of PfACb mRNA in the late trophozoite and schizont stages of the parasite. Representative photos of Giemsa-stained parasites corresponding to the time of RNA extraction for the RT-PCR analysis are shown below the graph. (TIFF) Figure S4 The solubility of His-tagged PfACb 1-785 is increased by high salt conditions. (Similar results were obtained with GST-PfACb 1-785 ). Hi-5 insect cells were infected with His-tagged PfACb 1-785 baculovirus and harvested after 42 hrs (determined to be the optimal time for maximal activity and expression of intact protein). Cell pellets were resuspended in a lysis buffer containing 50 mM Tris (pH = 7.5), 10 mg/mL aprotinin/leupetin, 1 mM PMSF, 1 mM benzamidine, 200 mM NaCl, and 1 mM DTT at ,10 mL lysis buffer/100 mL of pelleted culture. This lysate was sonicated five times at 10-second intervals at 12 watts with a Misonix Microson cell disruptor. Sonicated lysate was clarified by centrifugation at 100,0006g using a Ti-75 rotor (Beckman). The pellet fraction was resuspended in lysis buffer and adenylyl cyclase activity corresponding to PfACb 1-785 activity remained in the insoluble pellet fraction. The various additives indicated above were added to the resuspended pellet fraction, and the solution was again clarified by centrifugation. Soluble fractions were assayed for adenylyl cyclase activity. This was used as a measure of PfACb 1-785 amount. Only 2 M NaCl significantly solublized PfACb 1-785 . (TIFF) Figure S5 Effect of KH7-like compounds on parasite viability. P. falciparum cultures were maintained in a 96-well plate in the presence of 40 mM of the indicated compound. Luminescence was measured after 48 hrs. Reactions were performed in duplicate. (TIF) Optimal Medical Therapy of Chronic Stable Angina: Current Guidelines and Future Perspectives The optimal medical therapy of patients with chronic stable angina is a controversial issue. While the management of stable angina has certainly improved during the last decade, a number of challenges remain. The European Survey showed that up to 30% of patients do not receive antiplatelet drugs, 50% do not receive a beta-blocker, and 25% do not use a lipid-lowering agent [1]. Furthermore, guidelines for the management of stable angina are based on therapies that, at variance with other diseases, have not been extensively evaluated in large randomized trials. Also, both American and European guidelines have been published years ago, and therefore do not take into consideration the role of newer treatment options in management algorithms. As a consequence, many patients at greatest risk receive less than optimal treatment nowadays. The optimal medical therapy of patients with chronic stable angina is a controversial issue. While the management of stable angina has certainly improved during the last decade, a number of challenges remain. The European Survey showed that up to 30% of patients do not receive antiplatelet drugs, 50% do not receive a beta-blocker, and 25% do not use a lipid-lowering agent [1]. Furthermore, guidelines for the management of stable angina are based on therapies that, at variance with other diseases, have not been extensively evaluated in large randomized trials. Also, both American and European guidelines have been published years ago, and therefore do not take into consideration the role of newer treatment options in management algorithms. As a consequence, many patients at greatest risk receive less than optimal treatment nowadays. Current Guidelines The trials of medical therapy vs. interventional approaches in stable angina were carried out almost three decades ago, when routine managements differed greatly from those currently adopted. Given that European [2] and American [3] guidelines were published in 2006 and 2007, respectively, review is expecting soon. In 2011, the National Institute for Health and Clinical Excellence (NICE) published the recommendations for the management of stable angina [4]. According to the UK document, optimal medical therapy should comprise one or two anti-anginal agents as needed, without distinguishing between betablockers or calcium channel blockers as first-line anti-anginal therapy. The guidelines recommend switching between these treatments if needed, or combining them if symptoms are not satisfactorily controlled. The major limitations of NICE guidelines are twofold: (i) they equally support the use of beta-blockers and calcium-antagonists, despite the latter have been studied in the INVEST trial only in 2003 [5]; (ii) they allow the newer antianginal drugs, i.e. ivabradine, ranolazine and nicorandil, to be used only when older drugs fail. The Newer Anti-Anginal Treatments There is now evidence to suggest the possibility that the newer antianginal treatments may act as 'disease modifiers. ' Ivabradine acts on the I fcurrent which regulates the intrinsic pacemaker activity of the sinoatrial node, resulting in reduction in heart rate and myocardial oxygen demand. Ivabradine has a proven antianginal efficacy and is effective beyond beta-blockers, as it increases coronary blood flow and preserves myocardial contractility and relaxation [6]. Ranolazine inhibits the late sodium influx across the sarcolemma, thereby attenuating abnormalities in ventricular repolarisation and contractility associated with myocardial ischemia, as well as preventing calcium overload that causes cardiac ischemia. Several studies have already demonstrated the antianginal and anti-ischemic activities of ranolazine [7]. At variance with beta-blockers and calcium-antagonists, ranolazine does not significantly alter heart rate or increase systolic blood pressure, thus explaining its relevant role in those with angina refractory to maximal tolerated doses of other anti-anginal medications. In addition, ranolazine has the potential of limiting myocardial ischemia, thus playing a role in the pathophysiology of peri-procedural myocardial infarction during percutaneous coronary intervention. A recent randomized, pilot study has demonstrated that pretreatment with ranolazine 1,000 mg twice daily for 7 days significantly reduced procedural myocardial injury in elective PCI [8]. Nicorandil is a potassium channel activator, as well as acting as a nitrate-like epicardial coronary vasodilator, lowering preload through venodilation. Treatment is associated with protection of the myocardium during ischemia, shortening action potential duration and preventing intracellular calcium toxicity [9]. Future Perspectives First-line treatment should not only provide symptomatic relief, but also address the underlying disease mechanism(s). Beta-blockers are here to stay as first-line therapy of stable angina, as they have diseasemodifying properties and exert favourable effects on arrhythmias, heart failure and prevention of sudden death. In contrast, there is a lack of evidence that calcium channel blockers are disease-modifying and their therapeutic role will be probably reduced in the near future. Long-acting nitrates are used on an empirical basis as randomised controlled trials are lacking, and are therefore expected to be replaced by the newer antianginal treatments. Ranolazine, ivabradine or nicorandil are already indicated as second line options in patients inadequately controlled by first-line treatments or in whom such treatments are contraindicated or not tolerated. Ranolazine and ivabradine, which are associated with very low rates of side effects, will have increased roles in the management of stable angina patients that are older and with co-morbidities. Conclusion Chronic stable angina is still a major therapeutic challenge. Despite the UK guidelines in 2011, a number of unresolved issues remain, specifically disparities in the management of the older patient with multiple co-morbidities. It is expected that the newer anti-anginal treatments ranolazine and ivabradine will play an increasing role in the future clinical pathways for the management of stable angina. Organic Mood Disorder Following Left Anterior Temporal Lobectomy with Amygdalohippocampectomy One third of patients with antiepileptic-resistant temporal lobe epilepsy (TLE) will have to undergo surgery for a better seizure control. Anterior temporal lobectomy (ATL) is done for mesial temporal sclerosis that is the most common histopathological lesion associated with TLE. Psychiatric manifestations following ATL are not uncommon with depressive symptoms more common with left ATL and manic symptoms following right ATL. Mr. A is a 42-year-old left cerebral dominant (Confirmed by WADA test) male with no past history of psychiatric illness who had undergone anterior temporal lobectomy with amygdalohippocampectomy. He started having manic episodes post operatively which subsided with antipsychotics. He had multiple such episodes over the next 13 years with minimal inter episodic symptoms. This is a rare instance of manic symptoms following left-sided ATL that emphasizes the need for better understanding of the cerebral laterality of affective symptoms. INTRODUCTION Around 30% of the patients with drug-resistant temporal lobe epilepsy (TLE) have the option of a surgical treatment [1] as the most common pathology on histopathology and radiological examination is mesial temporal sclerosis (MTS). [2] Anterior temporal lobectomy (ATL) offers the best chance of achieving seizure control, resulting in seizure freedom in 50-80% of patients undergoing this procedure. [3] Psychiatric symptoms after ATL is not an uncommon finding with depressive symptoms being predominant. [4] Manic and hypomanic symptoms after ATL is a rare phenomenon [5] mostly reported in case series. We report a case with manic syndrome following leftsided ATL. CASE REPORT Mr. A is a 42-year-old left cerebral dominant (Confirmed by WADA test) male who is the third born, out of a non consanguineous marriage with no past history or family history of mental illness. He had several stereotyped attacks of complex partial seizures with aura and automatisms from young age and a seizure frequency of approximately one episode per month. His seizures were identified to be of left temporal in origin with the help of video EEG recordings. MRI brain |
showed left-sided MTS. His seizures were refractory to antiepileptic drugs and hence surgical option was considered. Anterior temporal lobectomy (ATL) with amygdalohippocampectomy (AH) was done on the left side in December 2001. No psychiatric problems were noted during premorbid evaluation by a psychiatrist except reports of 'proneness to irritability' from relatives. One third of patients with antiepileptic-resistant temporal lobe epilepsy (TLE) will have to undergo surgery for a better seizure control. Anterior temporal lobectomy (ATL) is done for mesial temporal sclerosis that is the most common histopathological lesion associated with TLE. Psychiatric manifestations following ATL are not uncommon with depressive symptoms more common with left ATL and manic symptoms following right ATL. Mr. A is a 42-year-old left cerebral dominant (Confirmed by WADA test) male with no past history of psychiatric illness who had undergone anterior temporal lobectomy with amygdalohippocampectomy. He started having manic episodes post operatively which subsided with antipsychotics. He had multiple such episodes over the next 13 years with minimal inter episodic symptoms. This is a rare instance of manic symptoms following left-sided ATL that emphasizes the need for better understanding of the cerebral laterality of affective symptoms. In the immediate postoperative period itself, he developed behavioural disturbance in the form of irritability, excessive talk, over familiarity, and making tall claims like he has a lot of property in his name. He had disturbed sleep and poor appetite. These symptoms lasted for around 3 months and he was on put antipsychotics. Following this, his symptoms improved. He later stopped his medications abruptly and had another episode of similar illness around a year later. Subsequently, he developed around six episodes of similar symptoms in the next 13 years with each episode lasting around 2-3 months. There were only minimal interepisodic symptoms and he was functioning well with no prophylactic medications in between the episodes. Seizures were not reported after the surgery. The current episode was of 1-month duration. On current mental state examination he was not fully cooperative for the interview and his talk was increased in tempo and amount. He was making tall claims and had an irritable mood during the interview. He was oriented to time, place, and person. His personal and social judgment was impaired and had a poor insight to his symptoms. He was alleging that it was his wife and not he, who had psychiatric illness for which she needed medications. Physical examination yielded no significant findings. A diagnosis of organic mood disorder was made. He is put on olanzapine and is on regular follow up. DISCUSSION A major complication of temporal lobe surgery is the development of a long-term psychiatric illness, which occurs in clear consciousness and is not related to seizure activity. [5] Around half of the postoperative patients report some psychopathology with depressive symptoms (55.6%) being the most common presentation. [6] Anxiety symptoms (27%), emotional liability (19%), and psychotic symptoms (4.8%) are also seen after ATL. [6] Mr. A is a person with fresh onset of manic symptoms immediately following left ATL with AH in the absence of positive family history. Several models are proposed for the emergence of manic symptoms following ATL. Valence theory suggests that the right hemisphere is more concerned with the negative emotions while the left side with positive emotions. [7] Hence, removal of the right side of brain can lead to an excessive activity of the left side of the brain and associated increase in positive emotions. [7,8] But in this case, it was the left side that was operated upon. Another hypothesis is regarding white matter dysfunction. When AH is done along with ATL there is disruption of white matter which results in a compensatory increase in synapses and loss of cortical inhibition. [5] Manic symptoms could also be due to the residual epileptiform activity occurring even after surgery. [9] It has been reported that more cortical dysfunction is seen with postoperative mania and more thalamic involvement in case of depressive symptoms. [10] Literature says that mania is more common following the right-sided surgeries and depression more common following the left-side ones. [5] Therefore, this case is one of the rare occasions of left-sided ATL precipitating manic symptoms and it demonstrates that our understanding of the relationship of mood and cerebral lateralization is far from complete. Circumcision with “no-flip Shang Ring” and “Dorsal Slit” methods for adult males: a single-centered, prospective, clinical study This paper was aimed to compare the clinical effectiveness and safety of adult male circumcision using the Shang Ring™ (SR) with the no-flip technique compared with Dorsal Slit (DS) surgical method. A single-centered, prospective study was conducted at the West China Hospital, where patients were circumcised using the no-flip SR (n = 408) or the DS (n = 94) procedure. The adverse events (AEs) and satisfaction were recorded for both groups, and ring-removal time and percentage of delayed removals were recorded for the SR group. Finally, complete follow-up data were collected for 76.1% of patients (SR: n = 306; DS: n = 76). The average ring-removal time for the SR group was 17.62 ± 6.30 days. The operation time (P < 0.001), pain scores during the procedure (P < 0.001) and at 24 h postoperatively (P < 0.001), bleeding (P = 0.001), infection (P = 0.034), and satisfaction with penile appearance (P < 0.001) in the SR group were superior to those in the DS group. After two postoperative weeks, the percentage of patients with edema in the SR group (P = 0.029) was higher but no differences were found at 4 weeks (P = 0.185) between the two groups. In conclusions, the no-flip SR method was found to be superior to the DS method for its short operation time (<5 min), involving less pain, bleeding, infection, and resulting in a satisfactory appearance. However, the time for recovery from edema took longer, and patients may wear device for 2–3 weeks after the procedure. INTRODUCTION Circumcision is one of the most ancient and common urinary surgeries dating back more than 5000 years. 1 To date, there have been hundreds of derivatives of the circumcision surgery, and the most common forms are the World Health Organization (WHO)-recommended forceps-guided, Dorsal Slit (DS), and sleeve resection methods. 2 In these traditional surgeries, it is necessary to cut-off the excess foreskin and suture the incision; this procedure is cumbersome and time-consuming, creating a longer learning curve for the surgeon. Among the above-mentioned methods, the DS method is the most widely used worldwide; 2 however, it requires superior surgical skill to avoid asymmetric removal of the foreskin on both sides, which, if not achieved, may result in an imperfect postoperative appearance. Therefore, there is an urgent need to reform the circumcision surgery to simplify the surgical process, shorten the operative time, reduce the adverse events (AEs) rate, and achieve a more acceptable appearance. In 2008, Peng et al. 3 first reported clinical data after applying a type of disposable circumcision device, the Chinese Shang Ring™ (SR), which was invented by Mr. Shang JZ from China. The study showed that compared with traditional circumcision, the novel device had a the corona without turning over the foreskin. This single-centered, prospective clinical study was conducted to compare the clinical efficacy and AEs between the "no-flip SR" and "DS" procedures for adult males with redundant prepuce or phimosis. Clinical data Patients attended the Department of Urology, West China Hospital, Chengdu, China, between October 2012 and April 2014. In total, 502 adult males aged 18-76 years (mean 25.4 years), 56 with phimosis and 446 with redundant prepuces, underwent the no-flip SR (n = 408) or DS (n = 94) procedures, respectively. Patients were given freedom to choose their procedure. However, the patients were sufficiently informed of the merits and costs of the two methods before deciding on which one to choose and providing written informed consent. The average cost of the SR procedure was 2-3 times higher than that of the DS procedure. Males with severe active balanitis, hypospadias, or a concealed penis were surgery taboos and were excluded. YCZ performed all these procedures. The study was approved by the Ethics Committee of the West China Hospital of Sichuan University. Surgical instruments For contents of the circumcision special surgical package in detail, please refer to the paper of Cheng et al. 4 The common SR sizes in China are C (34 mm), D (33 mm), E (32 mm), F (31 mm), G (30 mm), H (29 mm), and I (28 mm) (Wuhu Santa Medical Equipment Technology Co., Ltd.; Wuhu; China). If required, a special opener and scissors were used during ring-removal. 4 Preoperative preparation A routine blood test, the coagulation function test, was conducted on each patient to rule out infection, circulatory disease, and dysfunction of blood coagulation. With the foreskin returned to a natural "resting" position at 20°C room temperature, the circumference of the penis was measured just below the corona, following which the corresponding size of SR was chosen. If the reading was just between two sizes, the larger one was selected. 4 When measuring the circumference of the penis, the special SR tape should not be too tight or too loose. The measurement process should be quick to decrease the chance of erection, which could result in a measurement error. Surgical method The patient lay supine and was disinfected from the level of navel to mid-thigh. Approximately, 1 ml of lidocaine (2%, 2 g ml −1 ) was subcutaneously injected at the 12 o' clock position near the base of the penis. Following this, without withdrawing, the needle was advanced into the layer of dorsal nerves. At this point, 1 ml of lidocaine was injected, followed by further 5 ml on both the sides to block the dorsal nerves. Lidocaine (1 ml) was also injected for the frenum position. The injection from both sides was planned to converge at the 6 o' clock position. Before injecting any anesthetic, the surgeon gently withdrew the plunger of the syringe to ensure that no blood was drawn. The foreskin was grasped at the 3, 6, and 9 o' clock positions with mosquito clamps to expose the opening of the foreskin. Any adhesions were separated, and the smegma was cleaned away. The glans and corona were disinfected again. An inner ring of the appropriate model size was placed between the corona and the inner plate of the foreskin, the length of the inner and outer plates were adjusted to the proper extent, and the outer ring was gently placed onto the inner ring. The outer ring was tightened to the first fixing buckle. Approximately, 0.5 cm of the inner plate of the prepuce was conserved with the ring positioned higher at the front and lower at the back, forming a 30°-45° angle with the axis of the penis. The positions of the inner and outer rings were adjusted; the outer ring was tightened to the second fixing buckle. Dissection scissors were used to cut the redundant foreskin. The blade was used to perform an open relaxation incision at the 2, 5, 8, and 11 o' clock positions of the incisal edge. 4 The incisal edge was disinfected and the wound covered with gauze (Figures 1-4). Patients were requested to take more rest and to avoid sexual stimulation and strenuous exercise during the first postoperative week for preventing penile erection, SR slip, or hemorrhage. If the foreskin could not be fully retracted because of phimosis, a dorsal incision was performed. The incision position was clamped with a medium artery forceps for 1 min, and the dissection scissors were then used to cut the foreskin for reducing bleeding. Following this, the proximal incision was clamped to prevent tearing the inner plate when turning over the foreskin. For details of the DS procedure, please refer to the guidelines of the DS method of circumcision under local anesthesia. 2 The local block procedure was just the same between the two groups. Time-related indexes The operation time was defined as the duration from disinfecting the operative area to covering the surgical incision with gauze. The ring-removal time was defined as the number of days between SR placement and spontaneous ring-removal. Delayed ring-removal was defined as the lack of spontaneous ring-removal during 3 weeks after placement. Evaluation of AEs Bleeding, edema, pain, infection, wound dehiscence, and delayed |
ring-removal were observed. The official criteria 8 were applied to evaluate if those AEs occurred. The severity of each AE was graded to three levels: mild (if no intervention was needed), moderate (if nonsurgical intervention was indicated), and severity (if surgical intervention or hospitalization was required). 9 A visual analog scale (VAS) was used to evaluate the pain 10 on a scale of 0 to 10, with 0 representing no pain and 10 representing sharp pain. We recorded the pain scores during the procedure and at 24 h postoperatively. Follow-up We phoned the patients 24 h after surgery for postoperative pain scores. We recommended a hospital check-up 2 and 4 weeks after surgery; at this second-time point, data regarding AEs were collected. Telephone follow-ups were used for the patients who could not return. We instructed each patient to return to the hospital for surgical ring-removal if spontaneous ring-removal had not occurred in 3 weeks. The satisfaction with penile appearance was followed-up at 4 weeks after surgery. If purulence, cracking, or bleeding of the wound occurred, patients were strongly recommended to return to the hospital. All the follow-up data were collected by WBX. SPSS 13.0 (International Business Machines Co., Ltd.; Armonk; USA) was used for statistical analyses. Discrete variables and continuous variables were compared with the 2 and t-test, respectively. P < 0.05 was considered to reflect a significant difference. Two authors (LRL and LY) conducted the statistical analysis independently, and any disagreements were solved by discussion within the study group. Comparison of operation results and AEs The baseline data were comparable between two groups in age and number of the phimosis/redundant prepuce (P > 0.05 Specific adverse events and their treatment There were three cases in which part of the outer ring slipped in the first 7 days after surgery because of frequent penile erection. For patients with over 1 cm of wound dehiscence (2 of the 3), we sutured the inner and outer plates where the dehiscence occurred and asked them to avoid sexual stimulation and to cover the penis with a cold towel if erection occurred. Meanwhile, they were asked to disinfect the wound and the slipping position with 5% povidone iodine 3 times a day. SR fell off by itself in 3 weeks, and the wound healed well after stitch removal in the hospital for all three cases. Hemorrhage occurred in 13 patients within the first 3 days; however, they recovered after disinfection, binding and suturing when necessary. A further six cases in the SR group had serious edema beyond 4 weeks. We asked these patients to steep the penis in strong brine for 5 min (3 times a day), following which the edema faded within an additional 2 weeks. There were 16 cases of infection after surgery. They all resolved after the application of antibiotics for 1 week and a routine of washing the penis with 5% povidone iodine every day. For the SR arm, spontaneous ring-removal occurred in 323 cases and delayed removal of the ring occurred in the residual 85 cases. Using special equipment in our department, the rings were successfully removed. DISCUSSION Previous studies have shown that male circumcision can reduce the human immunodeficiency virus (HIV) infection rate in men by 60% and reduce urinary tract infections, penile cancer, and the female partners' incidence of cervical cancer. [11][12][13][14] In 2008, the WHO published three methods of adult male circumcision using local anesthesia: forceps-guided, DS, and sleeve resection methods; all three require cutting and suturing, and the operation time is rather long. 2 Circumcision using SR is an innovative circumcision surgery because it abandons the complex surgical procedures during traditional circumcision surgery. Because of the pressure between the inner and outer ring blocks, blood supply to the distal foreskin is stopped, resulting in necrosis and eventual shedding of the dead tissue. SR simplifies the surgical procedure, shortens the learning curve, reduces the operation time and difficulty, and substantially reduces the rate of AEs such as pain and infection. However, if choosing the method of waiting for spontaneous ring-removal, patients must wear the device for 2-3 weeks after the procedure, which is a major drawback. To our knowledge, this was the first study to compare the clinical efficacy and AEs of the "no-flip SR" and "DS" procedures for males with a redundant prepuce or phimosis. Most published clinical data on circumcision using SR are based on the so-called "standard program" (flip SR method). 4 The study by Xie et al. 6 shows that the pain score of the no-flip SR method is lower than that of the flip method at day 3 after surgery (2.23 ± 0.24 vs 4.92 ± 1.21, P < 0.05). This study showed that most patients can tolerate surgery after taking nonsteroidal analgesics orally 1 h before surgery. The pain scores during the procedure and at 24 h for the no-flip SR method were 1.78 ± 1.27 and 4.02 ± 1.16, respectively, which were lower than those for the DS method (P < 0.001 for both time points). We noted that the no-flip method left the foreskin in a more natural state, reducing the stretching on the foreskin by the ring, leading to less pain and discomfort. In contrast, the drawback of the long duration and complex and invasive procedure of the DS method may cause more pain to patients. However, because the inner ring of the no-flip method was located in the inner foreskin, forming a narrow space between the corona and the inner ring, depositions of smegma and residual urine were more common using the no-flip method than using the flip method; the latter were subsequently more likely to result in infection. This study found that the infection rate was 2.94%, which was higher than that using the flip method as reported by Cheng et al. and Xie et al. (0.6% and 1.5%, respectively) but still lower than that using the DS method (P = 0.034). 4,6 In addition, Yang et al. 15 reported that the incidence of infection for no-flip adult SR was 0.56% (3/528), which is lower than that detected in our study (2.94%, 29/306). They did not report the complete follow-up data of how many cases were available among the 528 cases; they may assume that no AEs occurred for cases lost to follow-up. In contrast, because the patients subjected to the DS method underwent a complex, invasive, and open procedure, they suffered a higher chance of being infected. Therefore, we suggested that each patient should wash his penis with 5% povidone iodine and clean the narrow space and the incision with povidone iodine swabs. Meanwhile, an obvious flaw of the SR procedure is the high incidence of edema. The incidence of edema at 2 weeks after surgery was higher than that in the DS group (P = 0.029) and was nearly 10% even at 4 weeks after surgery. This may be a result of some of the following factors: (i) lymphatic vessels could not be rebuilt in time, unlike after traditional surgery; (ii) a relatively narrow space between the corona and the inner ring increased the risk of infection, and infection could aggravate the edema; (iii) the inflammation reaction of the foreskin; or (iv) repeated erection of the penis after surgery. Therefore, it was advised to avoid sexual stimulation. Washing the penis with 5% povidone iodine mixed with 10% hypertonic saline during the early postoperative stage not only removed the secretion of the inner plate of the foreskin and the incision, thereby reducing the incidence of infection, but also relieved the foreskin edema. The study by Cheng et al. 4 shows that the pain score of manual ring-removal is the highest among the pain scores collected at different time points and that returning to the hospital for ring-removal increases the financial burden on patients. The randomized comparison in the study by Barone et al. 16 confirmed that within the 3 weeks after surgery, the occurrence of spontaneous ring-removal was more likely when the , and severity (if surgical intervention or hospitalization was required); *The P value was applied to the comparison between the total events of SR arm and the total events of DS arm. NA: not available; AEs: adverse events; DS: dorsal slit; SR: shang ring patients waited for longer. Therefore, we recommended the patients to come back to the hospital only in cases of delayed removal. We showed that the proportion of spontaneous ring-removal was 79.17 (323/408); this reduced the cost and avoided the pain of surgical ring-removal in most patients. For all patients in all cases of delayed removal, 2% lidocaine was applied to the surface to soften the scab, thereby reducing the pain and the occurrence of incision cracking and bleeding. For patients who are not willing to wait for SR to fall off on its own, the ring-removal procedure should be conducted at least 7 to 10 days postoperatively. 4 Importantly, a randomized controlled trial (RCT) 17 with circumcision intervention, which was performed at Uganda, demonstrated that among the couples with HIV+ females and HIV− males at baseline who resumed intercourse more than 5 days before the male partner's wound was certified as completely healed and within the 5 days before or any time after certified wound healing, 27.8% (5/18) and 9.5% (6/63) of males were confirmed with the seroconversion of HIV (P = 0.06), respectively. Therefore, it is necessary and important to remind the patients to delay resuming intercourse or to use condoms. The most exciting result for the SR group was a near-perfect level of satisfaction with a penile appearance in comparison with the DS group (96.41%, P < 0.001). The inner ring of SR determined the edge of the residual prepuce; therefore, it could objectively produce an appropriate length of prepuce and frenula, bilateral symmetry appearance, and smooth edge. For the DS procedure, it may be difficult for surgeons to determine the prepuce edge and the ultimate appearance, largely depending on their surgical skill. Finally, this study still had several limitations. First, although the prospective study provided evidence to support the use of the "Chinese SR, " the rate of follow-up was only 76.1%, thereby leading to potential bias, particularly for AEs. Further, RCTs with large-scale, multi-center, and rigorous design would provide more robust conclusions. Second, the self-reported data (24.6%) may result in reporting bias; therefore, the actual prevalence of AEs was higher than the current outcome. In addition, we failed to investigate their sex life quality after circumcision because most patients refused to answer related questions. A precise cost effectiveness of the two methods also was not assessed, which should have been considered, particularly in developing countries such as China. CONCLUSION In general, our study shows that circumcision using the no-flip SR method is superior to that using the DS method. The operation time was only approximately 5 min, with less pain and AEs such as bleeding and infection, producing a neat healing edge and a higher satisfaction with penile appearance. However, the edema recovery period was longer, and patients may wear the device for 2-3 weeks after the procedure. AUTHOR CONTRIBUTIONS JHL and LRL wrote the first edition of the paper. WBX and TRS collected the data. SBY and LY analyzed the data. QW, YCZ and PH commented in detail on the drafts. All authors read and approved the final manuscript. Complementing Cancer Metastasis Complement is an effector of innate immunity and a bridge connecting innate immunity and subsequent adaptive immune responses. It is essential for protection against infections and for orchestrating inflammatory responses. Recent studies have also demonstrated contribution of the complement system to several homeostatic processes that are traditionally not considered to be involved in immunity. Thus, complement regulates homeostasis and immunity. However, dysregulation of this system contributes to several pathologies including inflammatory and autoimmune diseases. Unexpectedly, studies of the last decade have also revealed that complement promotes cancer progression. Since the initial discovery of tumor promoting role of complement, numerous preclinical and clinical studies demonstrated contribution of several complement components to regulation of tumor growth through their direct interactions with the corresponding receptors on tumor cells or through suppression of antitumor immunity. Most of this work, however, focused on a role of complement in regulating growth of primary tumors. Only recently, a few studies showed that complement promotes cancer |
metastasis through its contribution to epithelial-to-mesenchymal transition and the premetastatic niche. This latter work has shown that complement activation and generation of complement effectors including C5a occur in organs that are target for metastasis prior to arrival of the very first tumor cells. C5a through its interactions with C5a receptor 1 inhibits antitumor immunity by activating and recruiting immunosuppressive cells from the bone marrow to the premetastatic niche and by regulating function and self-renewal of pulmonary tissue-resident alveolar macrophages. These new advancements provide additional evidence for multifaceted functions of complement in cancer. iNTRODUCTiON In both mouse models of cancer and patients, the expression of several complement genes is increased, resulting in higher than normal concentrations of complement proteins in plasma or other body fluids (1,2,3). In addition, complement activation is thought to occur in cancers because activated complement fragments are deposited within tumors (4,5). This deposition of complement cleavage products and complement protein complexes including the C5b-9 terminal complement complex was observed in breast cancer (6) and in papillary thyroid carcinoma (7,8). Complement activation through the lectin pathway was shown in colorectal carcinoma (9,10). Complement fragments were detected in ascites from ovarian carcinoma patients (11). Complement activation in cancer patients is also supported by detection of C5a circulating in plasma of non-small cell lung carcinoma patients (2). The early studies reporting upregulation and activation of the complement pathway led to a notion that complement, similar to lysing bacteria, may contribute to lysis of tumor cells and, consequently, participates in tumor immune surveillance. However, this is disputable because of the resistance of cancer cells to complement-mediated lysis, which, however, become obvious mainly in the context of use of monoclonal antibodies for cancer immunotherapy (12,13). This resistance results from high expression of membrane complement regulatory proteins (CRPs) on tumor cells (14) and secretion of soluble complement regulators from these cells (15), especially in solid tumors (13,16). In contrast, in hematologic malignancies, complement mediated killing can be relevant, at least in the therapeutic context. For example, rituximab, a chimeric CD20 monoclonal antibody used to treat B cell lymphomas utilizes complement-mediated cytotoxicity (CDC) to kill tumor cells. There is growing interest in targeting complement regulators to improve efficacy of monoclonal antibody therapy in cancer (13,17). Another approach to improve complement-mediated killing of tumor cells is the use of the "hexabody" platform. This technology stems from a seminal discovery that IgGs form hexamers after binding to antigen on the activating surface. This process is mediated by noncovalent interactions between Fc fragments of IgGs (18). Engineering Fc segments can be utilized to enhance formation of hexamers and, consequently, improvement of CDC toward tumor cells (19). Additional example of antitumor complement functions is participation of the complement anaphylatoxins C3a and C5a in enhancing antitumor immunity after radiotherapy. Interestingly, dexamethasone, a drug often administrated during radiotherapy limited complement activation and, consequently, inhibited antitumor immunity (20). In contrast to these beneficial outcomes of complement activity, it is conceivable that without the discussed here therapeutic interventions, complement enhances tumor growth through its proinflammatory properties (4,5). This possibility is consistent with a well-established tumor promoting role of chronic inflammation (21). Indeed, the first work to demonstrate tumor promoting properties of complement showed that several complement deficiencies were associated with reduced tumor growth through mechanisms linked to improvement of antitumor immunity (22). Several follow-up studies demonstrated immunoregulatory properties of various complement proteins (23). In addition, complement enhances tumor growth through direct regulation of tumor cell proliferation and invasiveness through C3a and C5a receptors expressed on carcinoma cells (24). Interestingly, the receptors for anaphylatoxins are also expressed in several leukemia and lymphoma cell lines and the blasts from chronic myeloid leukemia and acute myeloid leukemia patients. These cells responded robustly to C3a and C5a stimulation in vitro through chemotaxis and this process is negatively regulated by heme oxygenase 1 (HO-1) (25). These findings indicate that trafficking and spread of tumor cells in hematologic malignancies is perhaps, at least partially, controlled by complement system, therefore, inhibiting complement or upregulating HO-1 offer a new therapeutic opportunity for hematologic malignancies. Together, studies of the last decade provide compelling evidence for a pivotal role of the complement system in tumor growth and targeting complement for anticancer therapy. Interestingly, recent developments point to regulation of cancer metastasis by complement, which appears, in some studies, to be independent from complement functions in primary tumors. This work links complement to a phase of metastatic process that only recently has been proved experimentally and is termed the premetastatic niche (26). We focus our discussion here on these new advancements on complement in metastasis. We also discuss contributions of complement to epithelial-to-mesenchymal transition (EMT), which initiates metastasis in primary tumors. COMPLeX COMPLeMeNT The complement system is an assembly of more than 50 proteins that work together to provide immunity from infections, regulate several homeostatic processes, and trigger responses to tissue damage or injury (23). Although the textbook definition places complement in the center of innate immunity, recent developments demonstrated that this versatile system functions beyond limits of the immune system, regulating, for example, synaptic pruning (27), tissue regeneration/repair (28,29), and bone homeostasis (30). In addition to its key function in innate immunity, complement regulates adaptive immunity. The receptors for the complement activation fragments are expressed in B and T cells and their signaling is pivotal for maintaining efficient protection against infection (31,32). The stimulation of the complement receptor 2 (CR2) through antigen coated with C3d reduces the threshold for B cell activation rendering costimulation for best antibody production (33,34). The studies on a role of complement in regulating T cell responses has led to surprising discovery that complement proteins in the cytoplasm regulate several intracellular process, mainly of a metabolic nature, essential for T cell homeostasis. The intracellular complement, termed "complosome, " interacts with other intracellular innate sensor systems to control processes that are fundamental for adaptive immune responses such as metabolic reprograming necessary for generation of effector T cells (35). The complement system also includes soluble fluid-phase or membrane-bound proteins, cofactors, regulators, and receptors (36). Upon stimulation by either pathogen or danger-associated molecular patterns, or antibodies, a cascade of events occurs that leads to activation of complement through different complement pathways. The alternative pathway is initiated by bacterial surfaces or unconstrained fluid phase hydrolysis of the complement C3 thioester (37). The lectin pathway is triggered through binding of mannose binding lectin or the ficolins (termed ficolin-1, ficolin-2, ficolin-3) to particular carbohydrates or N-acteyl residues (38,39). The classical pathway starts when C1q binds to at least two IgG molecules (or one IgM) in a complex with antigen (40). In addition, complement fragments can be cleaved and thus activated through proteolytic enzymes that are not traditionally linked to the complement system. We grouped these additional ways of complement activation under the "umbrella" of the "fourth extrinsic pathway" (41). All three traditional complement activation pathways lead to cleavage of a complement fragment C3, which results in generation of C3a anaphylatoxin (10 kDa) and a large component-C3b. The C3b is deposited on the bacterial or other activating surfaces (42,43). Following cleavage of C3 by an enzymatic complex-C3 convertase, C5 is cleaved by C5 convertase and similar to C3 cleavage, small C5a and large C5b fragments are generated. C3b and C4b opsonize pathogens, e.g., flag them for phagocytosis by myeloid professional phagocytes that express receptors for C3 cleavage fragments. The large C5 cleavage product, C5b, binds to an activating surface and supports subsequent binding of C6, C7, C8, and finally C9 [membrane attack complex (MAC)]. The multiple C9 fragments polymerize and form a pore in the cell membrane resulting in cell lysis or cell activation in certain circumstances (44). The complement anaphylatoxins C3a, C4a, and C5a are potent mediators that orchestrate events of inflammation (41,45). Excessive complement activation can be deleterious; therefore, this process is tightly controlled by CRPs (46). There are both soluble and membrane bound CRPs that can be grouped into several functional categories: (i) CRPs with decay-acceleration activity that increases the rate of C3 convertase breakdown and (ii) with cofactor activity resulting in the cleavage of C3b and C4b, thus, stopping C3 convertase formation (47). Three additional important CRPs are factor H, C1 inhibitor (C1NH), and CD59. Factor H acts in the alternative pathway as a C3 convertase decay accelerator and as a cofactor for factor I-mediated cleavage of C3b. CD59 is the only CRP, which acts to prevent assembly of the MAC. C1NH acts in both the classical and lectin pathways by inactivating C1r, C1s, and mannose-binding lectin serine proteases (47,48,49). COMPLeMeNT AND CANCeR In 2008, complement C3, C4, and C5a receptor 1-deficient mice were shown to have slower tumor growth in a model of human papilloma virus-induced cancer (22). This paper was the first study to contradict a well-accepted, at that time, notion of complement participation in immune surveillance. Tumor promoting functions of complement, at least in this model, were linked to C5a/C5a receptor 1 (C5aR1)-mediated activation and recruitment of myeloid-derived suppressor cells (MDSC) to tumors and inhibition of antitumor immunity. At the time of this publication, concerns were raised that the observed phenotypes may be restricted to a single tumor model (4,5,50). However, multiple preclinical and clinical studies in the last decade supported tumor-promoting properties of different complement components (16,49). For example, studies by Corrales and colleagues demonstrated that C5a regulates MDSC in a lung cancer model (2). The blockade of C5aR1 led to a reduction in expression of genes that suppress antitumor immunity including Arg1, Il-6, Il-10, Ctla4, Lag3, and Cd234 (PDL1) (2). Recently, it has been shown that C3aR and C5aR1 signaling have an important impact on the IL-10-mediated cytotoxic properties of CD8 + T cells infiltrating tumors in models of melanoma and breast cancer (E0771) (51). In this manuscript, tumor infiltrating CD8 + T cells were shown to produce C3, which in autocrine manner inhibited the expression of IL-10. This cytokine appears to be essential for the cytotoxic properties of these cells. Mechanistically, IL-10 was associated with C3aR and C5aR1 signaling in CD8 + T cells. Complement's role in recruiting tumor-associated macrophages (TAMs) and controlling their proangiogenic characteristics was proposed in a work, exploring antitumor functions of pentraxin 3 (52). In addition to research demonstrating contributions of complement to inhibition of antitumor immunity, several studies showed other mechanisms behind tumor promoting functions of complement. In ovarian carcinoma models, tumor cells were demonstrated to produce complement components. C3a and C5a generated through activation of complement fragments produced in tumor cells regulated proliferation and invasiveness of tumor cells in autocrine fashion (24). C1q deposited in several human malignancies and mouse tumors seems to accelerate tumor growth through its proangiogenic properties and direct regulation of tumor cell motility and proliferation (53). Of value, Ajona et. al recently demonstrated improved efficacy of programmed celldeath 1 (PD-1) blockade in the presence of complement inhibition in reducing progression of tumors in a model of lung cancer (54). These new findings divulge a feasible path for targeting the complement system with the use of immunotherapeutic agents along with T cell check inhibitors. The detailed and comprehensive descriptions of a role in regulating tumor growth can be found in recent reviews (16,23,49). Here, we focus the discussion on the role of complement in regulating metastasis, a role that seems to involve different mechanisms. MeTASTASiS A HALLMARK OF MALiGNANCY Cancer metastasis is a process of relocation of tumor cells from a primary to a distant (disconnected from primary tumor) site, through lymph or blood. In fact, a metastatic potential determines the malignant character of primary growth (55). Cancer metastasis are responsible for approximately 90 percent of cancer-associated deaths, however, paradoxically, mechanisms regulating metastasis remain the most obscure aspect of cancer biology (56). The metastatic spread of cancer is a multistep and complex chain of alterations in tumor and host cells, and tumor stroma, known as the invasion-metastasis cascade (56,57). This cascade involves processes in primary tumor sites, circulation, and metastasis-targeted organs. Some of the first steps in the metastatic cascade involve acquisition of the ability to migrate and invade and degrade the tumor stroma by tumor cells. This goal is achieved through triggering in tumor cells several cellular programs that are collectively termed EMT, which is also an essential process during embryogenesis and wound healing (58). The EMT occurs perhaps in several malignancies; however, |
the current understanding of these cellular adaptations stems from studies in the models of epithelial-origin neoplasms carcinomas (59). Complement has been linked to EMT in two recent studies (Figure 1) (60,61). In the first study, increased expression of C5aR1 was found in hepatocellular carcinoma and hepatocellular carcinoma-derived cell lines and positively correlated with stage and invasion of liver capsule by tumor cells. The stimulation of C5aR1 via C5a induced EMT, as demonstrated by downregulation of E-cadherin and Claudin-1 expression, and upregulation FiGURe 1 | Overview of a role of complement in cancer metastasis. The C5a/C5aR1 axis contributes to the initial step of the invasion-metastasis cascade epithelialto-mesenchymal transition, which is essential for tumor cell motility, invasion of extracellular matrix blood vessels. C5aR1 signaling contributes to the formation of the premetastatic niche by recruiting immunosuppressive myeloid-derived suppressor cell from bone marrow to the lungs and by regulating self-renewal of alveolar macrophages that impair antitumor immunity via reducing antigen-presenting capacity of dendritic cells (APC) and polarizing T cell response toward Th2 phenotype. of Snail. Mechanistically, C5aR1-mediated EMT was linked to ERK1/2 signaling (61). In another study, C3 expressed in ovarian carcinoma-derived cells reduced expression of E-cadherin through C3a and Krüppel-like factor 5. Interestingly, C3 expression in tumor cells is transcriptionally regulated by twist basic helix-loop-helix transcription factor 1 (TWIST1), which binds to the C3 promoter and enhances its expression. TWIST1 and C3 colocalized at the invasive tumor edges, and in the neural crest and limb buds of mouse embryos. Therefore, this work identified TWIST1 as a transcription factor that regulates C3 expression during pathologic and physiologic EMT (60). The phenotypes associated with EMT program resemble phenotypes of cancer stem cells (CSCs) that are essential for metastatic spread. The recent work showed that CD10 + cancer-associated fibroblasts that express a second C5a receptor (C5L2) provide a survival niche for CSCs through C5L2-mediated NF-kβ activation (62). Through EMT, tumor cells reduce their attachment to neighboring tumor cells and surrounding stromal elements, increase motility, and acquire the ability to invade stroma, blood, or lymphatic vessels, thereby gaining access to the vasculature. The invasion of blood or lymphatic vessels enables tumor cells to intravasate and enter the circulation. The histopathological identification of vasculature invasion is itself a poor prognostic factor and often correlates with advanced metastatic disease (63). The lymph node metastases are a critical factor in cancer staging and are independent prognostic factors in several malignancies (64). However, mortality in cancer patients results from hematogenous spread to the vital organs including lungs, liver, and ultimately brain. Although initially lymph node metastases were thought to precede the subsequent hematogenous spread of cancer, evidence that draining lymph nodes are just temporary "parking" sites for cancer cells, before their departure to blood, is rather limited. It seems that lymph nodes represent a final destination for some cancer cells while other tumor cells, for unclear reasons, spread through the blood vessels (59). Upon successful intravasation, tumor cells move with the bloodstream to distant sites. However, only a small fraction of tumor cells that enter circulation safely reach their destination in the capillary beds of lungs and liver or cross the blood-brain barrier. This low efficacy of metastatic spread in blood results from hemodynamic stress and elimination of circulating tumor cells by the innate immunity, mainly natural killer (NK) cells (65). In contrast to NK cells, interactions with platelets (66) and neutrophils (67,68) appear to facilitate metastasis. After reaching their final destination, tumor cells are trapped in the capillary beds of the vital organs because their size is usually larger than the diameter of a single capillary. The halting of tumor cells in narrow capillaries facilitates their interaction with endothelium that is required for adhesion to endothelial cells and subsequent crossing of this endothelial barrier by extravasating tumor cells (transendothelial migration). Several substances secreted by tumor and host cells in the capillary beds enhance adhesiveness of tumor and endothelial cells and increase vascular permeability (69,70), thereby, facilitating tumor cell extravasation. In the liver and kidneys, the fenestrated endothelium seems to facilitate seeding of these organs by metastasizing tumor cells. Perhaps the mechanisms contributing to extravasation of tumor cells in different organs vary, depending on the location and intrinsic properties of metastasizing cells. After successful seeding of distant sites, tumor cells usually persist in an indolent state as single disseminated tumor cells or subclinical microscopic metastases, sometimes for years. The reasons for tumor cells to remain in a dormant state are unclear; however, poor adaptation of tumor cells to new microenvironment of metastasis-targeted organs seems to play a significant role (59). In addition, transition to rapidly growing and clinically overt metastasis, known as metastatic colonization, requires robust angiogenesis and immune evasion that may not be evident during a dormant phase of metastatic progression (71). For breast, prostate, and kidney cancers, a dormant phase may last even for decades after initial therapy and eradication of a primary tumor (59). Therefore, dormant tumor cells need to find a microenvironment-niche that allows them to slowly self-renew, provides needed nutrients, and protects from anticancer drugs and elimination by the immune system (72). For example, prostate carcinoma cells often metastasize to bones where they compete for residence in the endosteal niche with hematopoietic stem cells (73). In multiple organs including lungs, bones, and brain, tumor cells reside in close proximity to blood vessels in a region known as the perivascular niche (72,74). Interestingly, as much as EMT is necessary to trigger the invasion-metastasis cascade in primary sites, the reversal of this process, called mesenchymal-to-epithelial transition (MET), contributes to metastatic colonization, which is a final stage of metastatic disease. In metastatic tumors, MET appears to be critical in restoring a complex and heterogonous structure resembling primary tumors (75). Metastatic colonization leads to development of clinically overt and rapidly growing metastatic lesions, which are the ultimate reason for cancer-associated mortality. The transition from dormant to rapidly growing metastases requires acquisition of specific cellular programs by tumor cells, such as, discussed already MET, but also complex and well-orchestrated changes in the microenvironment of metastasis-targeted organs that include angiogenesis (72,76), inflammation (77), remodeling of extracellular matrix (78,79,80), and evasion of antitumor immunity (81). THe PReMeTASTATiC NiCHe Surprisingly, in several mouse models of cancer, changes that appear to be essential for metastatic colonization, e.g., a final stage of the invasion-metastatic cascade, including vascular alterations, remodeling of extracellular matrix, inflammation, and immunosuppression are observed in certain organs that seem to be marked for metastasis even before the arrival of the tumor cells. These alterations, collectively known as the premetastatic niche, are thought to facilitate seeding of these organs by disseminated tumor cells and their survival after they arrive to distant sites. The establishment of the premetastatic niche is triggered by the primary tumors (82) because efficiency of seeding of metastasistargeted organs by intravenously (i.v.) injected tumor cells is greatly enhanced by the presence of these tumors (3). Tumor-free mice i.v. injected with murine cancer cells developed significantly less lung metastases-derived from these i.v. injected cells than breast tumor-bearing mice i.v. injected with the same amounts of cells, indicating that the presence of primary breast malignancy facilitated seeding of the lungs by circulating (i.v. injected) tumor cells (3). It also appears that different types of cancer selectively prepare the premetastatic niche in different organs. This reflects the tendency of some malignancies to metastasize preferentially to specific locations. This specificity, known also as organotropism, was initially noted by Stephan Paget in 1889 (82), however, mechanisms regulating organotropism remain unclear until now. These mechanisms perhaps involve complex interactions between tumor cells and metastasis targeted organs that were proposed by Paget in his "seeds (tumor cells) and soil (microenvironment of premetastatic sites) theory. " However, until seminal studies of the last decade (83,84), which indeed established the field of premetastatic niche, the experimental proof for Paget's theory was missing. It is increasingly accepted that the premetastatic niche is created by tumor-secreted factors and tumor-shed extracellular vesicles, mainly exosomes. These factors seem to collectively control the stepwise development of premetastatic niche that begins with vascular alterations and progresses through activation of resident cells, extracellular matrix remodeling, and recruitment of bone marrow-derived cells (85). Secreted Factors The evidence that tumor-secreted factors contribute to the premetastatic niche and organotropism, was, perhaps, first provided by experiments showing that melanoma-conditioned medium injected into mice, directed the metastasis of Lewis lung carcinoma cells (which normally metastasize only to the lungs) to sites typical for experimental melanoma metastasis (84). Among several identified factors secreted by tumors, vascular endothelial growth factor A (VEGFA), placental growth factor (84), transforming growth factor β (TGF-β), and tumor necrosis factor (TNF) were first demonstrated to prepare "soil" for tumor cells (26,83). exosomes Exosomes, small extracellular vesicles formed on the cell surface through a budding mechanism, contain diverse cargo that facilitates cell-to-cell communication and homeostatic cell regulation (86). However, in patients and mouse models, formation of exosomes by tumor cells is increased compared to normal cells (87). Tumor-derived exosomes were isolated from plasma of cancer patients and mice with experimental tumors and found to carry tumor-derived cargo that promotes disease progression (87). This exosomal cargo, which includes tumor-derived miRNA and proteins, reprograms the target cells toward a prometastatic and pro-inflammatory phenotype, resulting in their contribution to the formation of the premetastatic niche. For example, melanoma B16-derived exosomes increase the expression of the receptor tyrosine kinase and MET in bone marrow progenitors, causing their exit from bone marrow and migration to the lungs, where they contributed to the premetastatic niche. Importantly, MET expression is also elevated in circulating CD45 − C-KIT low/+ TIE2 + bone marrow progenitors from patients with metastatic melanoma (87). B16-derived exosomes increase vascular permeability and enhance expression of TNF, S100A8, and S100A9, contributing to recruitment of bone marrow cells to the lung premetastatic niche. Of note, the source of S100 proteins was not identified in this study (87). Abundant Complement In contrast to tumor-derived secreted factors and exosomes, complement proteins are present in abundance in plasma and body fluids (41,88) and, therefore, are readily available to participate in the premetastatic niche in patients or mice even with very small tumors. Increased concentration of complement components in plasma and other bodily fluids has been observed in both cancer patients and mouse models of cancer (1, 2, 3) suggesting upregulation of the complement pathway. These higher amounts of complement proteins may be linked to enhanced expression and production of complement by the liver, however, local increases in expression of complement genes in tumors and organs targeted by metastasis contribute to augmented levels of complement fragments because endothelial and immune cells synthesize complement fragments (89,90) and these cells are an integral component of the tumor microenvironment (81,91) and the premetastatic niche (26). Thus, they are possible sites of origin for several complement proteins in tumors and metastatic sites. For example, in a mouse model of breast cancer with spontaneous metastatic spread mimicking human malignancy, increased concentrations of C3 were found in plasma and bronchoalveolar lavage indicating increased production of complement proteins (3). These higher levels of complement fragments correlated with increases in expression of C3 and C5 genes in the lungs (3,92). Cytotoxic CD8 + T cells were found to synthesize C3 in mouse models of melanoma and breast cancer (51). Importantly, tumor cells also produce complement proteins. Mouse ovarian carcinoma tumor cells and human ovarian carcinoma cell lines were demonstrated to produce C3 (24). In a squamous cell carcinoma model, expression of C3, factor B, and factor I were also demonstrated (93,94). Boire and colleagues recently demonstrated that C3 produced and secreted from tumor cells has prosurvival functions and facilitates leptomeningeal metastasis (95). Complement proteins are secreted from cells in their inactive forms, as zymogens. To exert their functions, these fragments are activated through a series of proteolytic cleavages that form a complement cascade, which ends with generation of complement effectors (41). Therefore, if complement plays a role in regulating metastatic progression, complement activation in metastasistargeted organs or tumors is anticipated. This activation can be revealed by detecting deposited complement fragments in tissue or secreted effectors such as complement anaphylatoxins, C3a, C4a, and C5a. The cleavage fragments of C3 were found to be deposited in the lungs prior to metastasis, indicating complement activation and participation of the complement system in the lung premetastatic (3). These data were |
obtained through a use of a syngeneic mouse model of metastatic breast cancer (4T1), in which tumor cells are injected into the mammary fat pad, and breast tumors formed there subsequently metastasize to distant sites, similar to human malignancy (96). The deposition of C3 cleavage fragments in the lungs correlated with increasing levels of C5a in plasma over time (Figure 1) (3). vasculature Increased vascular permeability is one of the earliest changes observed in the premetastatic niche and is associated with increased metastatic burden (97,98). The factors secreted from primary tumors, including epithelial growth factor receptor ligand epiregulin, metalloproteinases MMP1, and MMP2, are known to impact vascular permeability in primary tumors and distant sites, helping tumor cells to intravasate in a primary and then extravasate in a distant site, respectively (99). In a melanoma model, tumor cells secrete factors upregulating angiopoietin 2, MMP3, and MMP10 that synergistically destabilize vasculature in the premetastatic organs (97). Factors affecting vascular permeability can also be secreted from different cell populations recruited to the premetastatic niche. For example, myeloid cells were shown to produce MMP9 (100) and VEGFA (101). Endothelial cells, which are often targets for vasoactive substances, can themselves participate in vascular alterations in the premetastatic niche. VEGFA-dependent upregulation of E-selectin on the luminal surface of endothelium facilities adhesion of tumor cells to endothelium and subsequent extravasation (102). The complement effectors, especially C5a, are powerful inflammatory mediators that are actively engaged in bringing leukocytes to sites of inflammation. C5a can achieve this goal, acting as a potent chemoattractant that causes cytoskeleton changes in leukocytes that are responsible for cell movement (103). However, it also enhances (directly and indirectly) vascular permeability, further adding to accumulation of leukocytes in inflammatory foci (103). The complement C3 cleavage fragments were found to be deposited in the premetastatic lungs as early as 4 days after injecting tumor cells into the mammary fat pad in a model of breast cancer (before any tumor cells are present in the lungs). Since C3 is a central component of complement cascade, on which all complement activation pathways converge, deposition of C3 cleavage fragments indicates complement activation and subsequent generation of C5a (88), which indeed was present in sera of these mice (3). It is, therefore, conceivable that complement contributes directly to increased vascular permeability in the premetastatic niche similar to its participation in inflammatory vascular alterations; however, a direct experimental evidence for these C5a functions in the premetastatic niche has yet to be provided. The indirect impact of complement on vascular changes can be attributed to recruitment of MDSC (3) because these cells can produce and release several vasoactive factors including MMP9, which is intimately involved in regulating vascular integrity in the premetastatic niche (83,84). Genetic ablation of Mmp9 was shown to normalize the aberrant vasculature in the premetastatic lungs and reduce metastatic burden (100). The seminal recent work has also demonstrated that cancer-cellderived C3/C3a through C3aR in the choroid plexus disrupts the blood-cerebrospinal fluid barrier. The increased permeability of this barrier facilities the entry of plasma proteins that are essential for tumor growth into the cerebrospinal fluid, thereby, facilitating leptomeningeal metastasis (95). Resident Cells Resident cells in metastasis-targeted organs are naturally suited to participate in the premetastatic niche because they are present before the arrival of tumor cells. Tumors can reach and potentially hijack these cells through several mechanisms including secreted tumor-derived factors, exosomes, and recruitment of bone marrow-derived cells that subsequently interact with resident components of the premetastatic niche. Recent work also demonstrated that complement activation regulates resident cells in the lungs (Figure 1) (92). As discussed already, endothelial cells are targets for several vasoactive substances whether derived from tumors, recruited cells, or generated locally. Fibroblasts contribute to remodeling of extracellular matrix through deposition of new extracellular matrix components or by secreting enzymes that affect preexisting components of the matrix (104). S100A4 expressing pulmonary fibroblasts incorporate exosomal cargo derived from breast cancer cells and through this mechanism upregulates S100 proteins (82). Exosomal cargo from pancreatic carcinoma induces similar changes in the liver-resident macrophages, Kupffer cells (105). S100 proteins were linked to recruitment of myeloid-origin cells to the premetastatic niche (106). Similar to Kupffer cells in the liver, another population of tissue-resident macrophages, pulmonary alveolar macrophages, were recently demonstrated to contribute to the premetastatic niche (92). These recent developments on participation of tissue-resident macrophages to metastasis are of particular interest because roles of these cells in cancer remain unclear, in contrast to very well-studied TAMs or inflammatory monocytes/macrophages recruited to the lungs with metastases by CCL2 (101). Several early studies yielded conflicting results on how liver Kupffer cells and lung alveolar macrophages contribute to cancer progression (107). A role of these cells in cancer requires revision because recently published linagetracing data demonstrated that these cells have a different origin and biology than inflammatory macrophages (108). Unlike inflammatory macrophages and TAMs that are recruited from bone marrow to sites of inflammation or tumors, respectively, tissue-resident macrophages migrate to different organs during embryogenesis prior to hematopoiesis, and self-renew thereafter (109). Pulmonary alveolar macrophages are the resident macrophages of the lungs, and while they have well-known immunoregulatory and homeostatic roles in healthy lungs (110), they appear to be well-suited to partake in preconditioning the lungs for metastasis through their immunoregulatory properties. In support of this notion, alveolar macrophages were found to accumulate in premetastatic lungs and this accumulation was the result of cell proliferation rather than recruitment from bone marrow (92). The mechanisms controlling proliferation of these cells were linked to C5aR1 signaling because tumor-bearing C5aR1-deficient mice presented with a lower total number of these cells in the lungs compared to tumor-bearing wild-type controls and this reduced cell number associated with reduced Ki-67 expression. Immunoregulatory functions of these cells appeared to be related to skewing effector CD4 + T cells responses toward Th2 phenotype, which plays limited role in antitumor immunity in contrast to Th1 responses (92). In addition, alveolar macrophages in tumor-bearing hosts reduced number and antigen-presenting capacity of lung dendritic cells through regulation of TGF-β1 in lung infiltrating leukocytes (Figure 1). The depletion of alveolar macrophages reversed immunosuppression and reduced lung metastatic burden (92). extracellular Matrix Remodeling The continuous remodeling of extracellular matrix by tumorderived secreted factors, resident fibroblasts, and recruited bone marrow-derived cells is an integral part of the premetastatic niche (111). This remodeling is achieved through deposition of new extracellular matrix components or modification of existing components. For example, deposition of fibronectin produced by activated fibroblast provides a docking site for bone marrow-derived cells that express fibronectin receptor VLA-4 (84). These stromal fibroblasts, stimulated with tumorderived TGF-β, produce periostin in a mouse model of breast cancer (78). Periostin directly interacts with type I collagen, fibronectin, and Notch1 through its amino-terminal EMI domain and interacts with tenascin-C and BMP-1 through its fas I domains. These periostin interactions with mainly extracellular matrix molecules occur at first intracellularly. In addition, periostin serves as a ligand for integrins such as αvβ3 and αvβ5 and promotes cell motility by acting outside the cell (112). Periostin was demonstrated to facilitate melanoma metastasis to wounds (113) and to regulate immunosuppressive functions of MDSC during early stages of breast cancer metastasis (114). MDSC (mainly monocytic-MDSC), which accumulated in the premetastatic lung of MMTV-PyMT spontaneous breast tumor-bearing mice, secrete versican, an extracellular matrix proteoglycan. Versican contributed to MET and the formation of macrometastasis in the lungs (115). Enzymatic modulation of extracellular matrix proteins also occurs in the premetastatic niche and is mediated mainly by metalloproteinases produced by cells that are recruited to the premetastatic niche. In addition, the members of the LOX family crosslink collagen type I and IV and this crosslinking facilitates adhesion of bone marrow-derived cells to the extracellular matrix of the premetastatic niche. These cells produce more metalloproteinases contributing to further remodeling of extracellular matrix. Although complement was not directly linked to extracellularmatrix remodeling in the premetastatic niche, studies in different model systems demonstrated that fibronectin can interact with several complement components including C1q (116) and C3 cleavage fragments (117). C3 cleavage fragments can also bind to different components of extracellular matrix including collagen (118). Interestingly, binding of C1q to fibronectin was not associated with complement activation but was connected to enhancement of phagocytosis of C1q coated particles through fibronectin. Therefore, functional significance of complement interactions with extracellular matrix proteins in the premetastatic niche remain to be elucidated. However, it is reasonable to theorize that complement C3 cleavage fragments, bound to extracellular matrix proteins, interact with its receptors broadly expressed on myeloid-origin cells that are recruited to the premetastatic niche. The receptors for C3 degradation products include CR1, which binds C3b and iC3b, CR2 (CD21), which binds the degradation products of C3b (iC3b, C3dg, C3d), CR3 (CD11b/CD18 or Mac-1), which binds iC3b, and CR4 (CD11c/ CD18), which binds iC3b, however, through a different domain than CR3 (119). C5a leads to upregulation of CR3 on MDSC, which may facilitate adhesion of these cells to endothelium and recruitment to tumors (22); however, it may also contribute to adhesion of MDSC to extracellular matrix in the premetastatic niches. Recruited Cells The identification of bone-marrow derived cells in the premetastatic niche and discovery of their roles in facilitating seeding of these niches by tumor cells, provided perhaps the first experimental evidence confirming the "seed and soil" theory (83,84). A seminal work by Hiratsuka and colleagues defined Mac1 (CR3) positive macrophages as a source of MMP9 in the lungs in addition to endothelial cells. These macrophages were recruited to the lungs because resident alveolar macrophages do not express CD11b (https://www.immgen.org/). The study by Kaplan and colleagues demonstrated recruitment of VEGFR1 and VLA-4 expressing hematopoietic progenitors to the premetastatic lungs and their participation in the premetastatic niche (84). These studies opened an avenue for further investigations into discovery of other recruited components of the premetastatic niche. Less than a decade later, MDSC, which were long recognized as modulators of the primary tumor microenvironment (120), were identified as contributors to the premetastatic niche (100,115). However, these studies reported on metastasis promoting properties of MDSC linked to increased vascular permeability (100) and remodeling of extracellular matrix (115) rather than to their well-established immunoregulatory roles. The C5aR1/C5a signaling axis recruits MDSC to primary tumors (2,22), therefore, it was explored whether similar mechanisms operate in the premetastatic niche. Utilizing a syngeneic mouse model of metastatic breast cancer, it has been demonstrated that C5aR1 knockout or wild-type mice administrated with a specific C5aR1 inhibitor (PMX-53) had decreased lung and liver metastatic burden compared to control mice. Interestingly, C5aR1 appear to regulate only metastasis in this model because lack of C5aR1in mice did not affect primary breast tumors. The differences in lung metastasis were associated with differences in a degree of infiltration of the lungs and livers by MDSC. The lung infiltrating MDSC were found mainly in interavleolar septa and due to intensity of this infiltration, the morphological picture resembled interstitial pneumonia. Therefore, the term the premetastatic pneumonia has been proposed to emphasize intensity of MDSC infiltration and specific localization of these cells in the lungs (3). These MDSC were recruited to the premetastatic sites through C5a since C5aR1 was expressed in blood MDSC and complement activation, leading to C5a generation, was observed in the premetastatic niche (Figure 1). To further investigate the role of C3 cleavage fragments and MDSC, tumor-draining lymph nodes from breast cancer patients were examined; they observed that C3 fragments' deposition and local C3 production were both intensified in lymph nodes with metastases (3). The decreased metastasis in mice lacking C5aR1 resulted from improved antitumor immunity due to escalated infiltration of the lungs by CD8 + and CD4 + T cells. In addition, the increases in these T cell subsets were found in peripheral blood and C5aR1-deficiency favored Th1 response. Also observed was a decrease in Tregs in both the blood and the lungs in C5aR1-knockout mice. The T cell subsets including CD4 + and CD8 + T cells isolated from the lungs of C5aR1 knockout mice produced increased amounts of IFN-γ. The elimination of cytotoxic CD8 + T cells by neutralizing antibody erased the inhibitory effect of C5aR1-deficiency on metastasis, supporting notion that this effect was caused by stimulating antitumor immunity. Importantly, these data also indicate immunoregulatory functions of MDSC in the premetastatic niche (3). CONCLUDiNG ReMARKS The evidence supporting contributions of |
complement to cancer metastasis is scarce and limited to a few recent papers. However, it appears that complement affects key steps in the invasionmetastasis cascade including EMT and the premetastatic niche (Figure 1). Given ubiquitous presence of complement in body fluids and tissues, the potential contributions of complement to regulating metastasis are significant. Our recent work demonstrated that C5aR1 regulates resident (alveolar macrophages) (92) and recruited (MDSC) (3) cells in the premetastatic niche. The significance of this regulation was underscored by complete protection from lung metastasis in mice depleted of alveolar macrophages and treated with C5aR1 inhibitor (92). Thus, despite an early phase of studies on complement participation in metastasis, several complement fragments appear to be promising targets for therapies seeking to stop cancer metastasis. Mitochondrial Dysfunction Secondary to Endoplasmic Reticulum Stress in Acute Myocardial Ischemic Injury in Rats Background The relationship between endoplasmic reticulum and mitochondria during acute myocardial ischemic injury is still unclear. Our study aimed to define the dynamics of endoplasmic reticulum stress and mitochondrial dysfunction during acute ischemic injury. Material/Methods A rat model of acute myocardial infarction and hypoxic cardiomyocytes were used in this study. Groups were set at 0 hours, 1 hour, 2 hours, 4 hours, and 6 hours after ischemic injury for both in vivo and in vitro studies. ATF6 and GRP-78 were examined to indicate endoplasmic reticulum stress. Cellular ATP and cytosolic levels of mitochondrial DNA and cytochrome c were detected to evaluate mitochondrial dysfunction. Caspase-3 was used for apoptosis analysis. Result Our results showed that both mRNA and protein levels of ATF6 and GRP-78 were elevated from 1 hour after ischemic injury in vivo and in vitro (P<0.05). However, ATP levels were increased at 2 hours after ischemic injury and significantly decreased from 4 hours after ischemic injury in vivo, while ATP level of cultured cardiomyocytes decreased remarkably from 2 hours after ischemic injury (P<0.05). Cytosolic mitochondrial DNA levels began to increase from 2 hours after ischemic injury (P<0.05). Cytosolic levels of cytochrome c increased from 4 hours after ischemic injury. Additionally, both mRNA and protein expressions of caspase-3 started to significantly elevate at 6 hours after ischemic injury (P<0.05). Conclusions The present study suggested that mitochondrial dysfunction was secondary to endoplasmic reticulum stress, which provides a novel experimental foundation for further exploration of the detailed mechanism after ischemic injury, especially the interaction between endoplasmic reticulum and mitochondria. Background Coronary artery disease is the leading cause of death and will remains so for the next few decades [1]. In addition to its high mortality rate, it is also a leading cause of morbidity and loss of quality of life [2]. Acute myocardial infarction usually occurs due to primary coronary artery disease and there are many well-known risk factors [3]. Although some causes and risk factors have been gradually elucidated, the molecular mechanisms of cardiomyocytes during acute myocardial infarction are still unclear and efforts are still needed to salvage para-infarction heart muscle and prevent post-infarction complications. Recently, an increasing number of studies have focused on the role of endoplasmic reticulum (ER) stress and mitochondrial dysfunction in acute myocardial infarction. It has been demonstrated that inhibition of endoplasmic reticulum stress can exert heart-protective effects after acute myocardial infarction [4]. Meanwhile, production of massive reactive oxygen species (ROS) is a well-known harmful factor to heart tissue, and inhibition of its production can limit mitochondrial damage and save heart tissue from damage [5]. In addition, mitochondria recently have been regarded as a promising target for cardio-protection in acute myocardial infarction [6]. However, the relationship between endoplasmic reticulum stress and mitochondrial dysfunction during acute myocardial infarction is still unclear. Both ATF6 and GRP-78 are 2 key proteins during the development of the endoplasmic reticulum stress. Dissociation of GRP78 and ATF6 could activated unfolded protein response [7,8]. It has been reported that the endoplasmic reticulum stress may result in mitochondrial damage and cell death, presenting as the production of ROS, releasing of mitochondrial DNA, and cleavage of caspases [9,10]. Many studies have revealed the possible relationships between the endoplasmic reticulum stress and mitochondrial damage in other pathophysiological scenarios [11,12]. However, in acute myocardial ischemic injury, these relationships have not been well-established, and insightful studies were imperatively needed. Therefore, in order to understand the pathophysiological processes during acute myocardial infarction and further explore the molecular mechanisms between endoplasmic reticulum and mitochondria, the present study aims to first define the time course of endoplasmic reticulum stress and mitochondrial dysfunction during acute ischemic injury. Material and Methods Experimental protocols in the acute myocardial infarction rat model We used 10-week-old Sprague Dawley (SD) male rats, weighting 220 to 250 g, purchased from Da-Shuo Biotech Company (Chengdu, China). All procedures followed the Guide for the Care and Use of Laboratory Animals and the protocols of the study were approved by the Ethic Committee of West China Hospital, Sichuan University. All rats were incubated and ventilated by an animal ventilation with room air. The rats had a left thoracotomy through a left parasternal incision performed to expose the heart. After 20 minutes stabilization, except for the control group, rats were induced to acute myocardial infarction by ligating the left anterior descending coronary artery. After 1 hour, 2 hours, 4 hours, and 6 hours, rats were sacrificed by cervical dislocation and their hearts were harvested for further analysis. Each group included 5 rats. Primary neonatal rat cardiomyocyte culture Neonatal rat hearts were collected within 24 hours and the heart tissue was cut into small pieces, followed by digestion with 22.5 µg/L Liberase blendzyme 4 (Roche, Germany) at 37°C for 40 minutes. The isolated cells were pre-plated for 90 minutes to remove non-cardiomyocytes, after which cardiomyocytes were cultured in Dulbecco's Modified Eagle Medium (DMEM) containing 10% fetal bovine serum (FBS) in 24-well culture plates pre-coated with 1% gelatin (Sigma). After 48 hours of cell culture, cardiomyocytes were subjected to hypoxia condition with 5% O 2 , 5% CO 2 and 90% N 2 for different time sets (0 hours, 1 hour, 2 hours, 4 hours, and 6 hours). Cytosol extraction Cytosol extraction was obtained by using the Mitochondria Isolation Kit for Tissue and Cultured Cells (KC010100, BioChain, USA). All procedures were strictly performed following the manufacturer's instruction. Briefly, myocardium or cultured cardiomyocytes were collected and washed twice with 10 mL icecold phosphate-buffered saline (PBS). Then the myocardium or cultured cardiomyocytes were homogenized and transferred into an Eppendorf tube, which was centrifuged at 600 g for 10 minutes at 4°C. The supernatant was carefully transferred into a new tube. The supernatant was centrifuged at 12 000 g for 15 minutes at 4°C to obtain the cytosol extraction, which is stored at -80°C for further analysis. Real-time polymerase chain reaction (RT-PCR) analysis Total RNA was extracted from myocardium or cultured cardiomyocytes by using TRIzol reagent (Sigma, USA). And cDNA was obtained by using the M-MLV reverse transcriptase kit (Invitrogen, USA). SYBR green PCR master mix was used in the PCR system following the manufactory's instruction (Bio-Rad Laboratories, USA). b-actin was used as a loading control. The primer sequences were as follows: 1) Atf6: forward: 5'-GATTTGATGCCTTGGGAGTC-3', e923124-2 ATP levels detection ATP levels from myocardium and cultured cardiomyocytes were detected using a specific kit (Jiancheng Bioengineering Institute, China). All procedures were performed following the manufacturer's instruction. For the last step, the ATP level was detected by spectrophotometry at 636 nm. The detections were equilibrated by the total protein concentration. Western blots Protein expressions of cleaved-ATF-6, GRP-78, cleaved-caspase-3, and cytochrome c were assessed by western blots. The myocardium was homogenized, and the cultured cardiomyocytes were lysed at 4°C for protein extraction, which was subjected to bicinchoninic acid (BCA) assay for measuring protein contraction. 50 µg protein extraction was added into 10% sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and then transferred to a nitrocellulose membrane. The membrane was incubated with specific primary antibodies against rat cleaved-ATF6, GRP-78, cleaved-caspase-3, cytochrome c and actin at 1: 1000 dilution overnight at 4°C, followed by incubation with horseradish peroxidase (HRP) conjugated secondary antibodies (Cell Signaling, USA) at room temperature for 1 hour. Protein bands were developed on an x-ray film. Densitometric ratios of cleaved-ATF6, GRP-78, cleaved-caspase-3, and cytochrome c to actin were obtained. Cytosolic mitochondrial DNA levels detection The whole cytosolic DNA was obtained by using the DNeasy Blood & Tissue Kit (No. 69504, Qiagen, China) as previously described [13]. The mitochondrial DNA level was detected using the SYBR green PCR master mix and the PCR system the CFX96™ Real-Time PCR Detection System (Bio-Rad Laboratories, USA). The primer sequences used to detect mitochondrial DNA (rat NADH dehydrogenase 1 gene) were as follows: CGCCTGACCAATAGCCATAA (forward); ATTCGACGTTAAAGCCTGAGA (reverse). All samples were analyzed 3 times for quality control. Statistical analysis All descriptive data was expressed as mean±standard error of the mean (SEM). Statistical studies were performed by GraphPad Prim 6 (GraphPad Software). One-way ANOVA followed by Bonferroni's test were used between multiple groups. P<0.05 was considered as statistical significance. Results As shown in Figures 1 and 2, both mRNA levels and protein levels of ATF6 and GRP-78 were significantly increased at the first hour and continued to increase in vivo and in vitro (P<0.05). The in vivo study revealed that the ATP levels were increased at the first 2 hours after ischemic injury and significantly decreased from 4 hours after ischemic injury (P<0.05). Intriguingly, ATP level of cultured cardiomyocytes decreased remarkably from 2 hours after ischemic injury (P<0.05). Both in vivo and in vitro studies showed cytosolic levels of mitochondrial DNA increased at 2 hours (P<0.05). While cytochrome c increased at 4 hours in vitro and 6 hours in vivo after ischemic injury (P<0.05). In addition, caspase-3, as the central biomarker of cell apoptosis, was used to examine the apoptotic status of the cardiomyocytes. Significant increasing of mRNA expression of caspase-3 and protein expression of cleaved caspase-3 were found at 6 hours after ischemic injury in vivo (P<0.05). In addition, in vitro study also revealed that increased mRNA level of caspase-3 at 6 hours and protein level of cleaved caspase-3 at 6 hours after ischemic injury (P<0.05). Discussion The present study first showed that after acute ischemic injury to cardiomyocytes, indicators for endoplasmic reticulum stress significantly increased at 1 hour, which demonstrated that endoplasmic reticulum stress occurred in the early stage of disease. In addition, ATP levels, which indicate energy metabolism were significant inhibited at 4 hours. In addition, mitochondrial DNA, which is a marker of mitochondrial damage, was elevated significantly at 2 hours, which demonstrated that the mitochondrial damage presented later than the endoplasmic reticulum stress. Our study implied that the mitochondrial dysfunction was secondary to the endoplasmic reticulum stress under ischemic insults. GRP-78-initiated unfolded protein response plays a critical and corrective role in endoplasmic reticulum stress, in which ATF6 is a key factor in the unfolded protein response [14,15]. It is reported that acute ischemic injury could cause significant endoplasmic reticulum stress, which induced cardiomyocytes e923124-3 apoptosis through endoplasmic reticulum stress-related apoptotic pathways [16]. Recently, increasing studies have demonstrated that inhibition of endoplasmic reticulum stress had great benefits on the ischemic injury cardiomyocytes [17,18]. Therefore, targeting endoplasmic reticulum stress might be a compromising strategy for treating acute myocardial infarction. Additionally, many studies have focused on acute myocardial infarction-induced mitochondrial dysfunction. As a major energy-producing organism in the cardiomyocyte, the mitochondria are essential for contraction through constant oxidative phosphorylation [19]. Under ischemic insults, multiple signaling pathways are activated to cause the uncoupling of the electron transport chain, opening of the mitochondrial permeability transition pore, and releasing of cytochrome c, which induces mitochondrial dysfunction-mediated cell death [5,20]. It is well-known that mitochondrial dysfunction-induced massive ROS production and cytochrome c release are responsible for apoptosis and necrosis, which have negative effects on adjacent cardiomyocytes and expands the infarct area [21]. Growing evidence has identified the inflammatory properties of free mitochondrial DNA, and that mitochondrial dysfunction or cell injury could result in releasing mitochondrial DNA from the mitochondria [22,23]. In addition, our previous study showed elevated circulatory mitochondrial DNA levels in patients with acute myocardial infarction [24]. In the present study, we found that mitochondrial dysfunction occurred later than endoplasmic reticulum stress under acute ischemic injury, which implied that mitochondrial dysfunction was secondary to endoplasmic reticulum stress. A study revealed a relationship between endoplasmic reticulum and mitochondria-regulated synthesis of proteins in the endoplasmic reticulum, and production of ATP in |
the mitochondria and apoptosis [25]. Another study demonstrated that endoplasmic reticulum stress could affect not only mitochondrial biosynthesis, but also mitochondrial function, in which mitochondrial-associated membrane (MAM), acting as the physical connecting site between the endoplasmic reticulum and the mitochondria, plays a critical role [26]. It has been suggested that endoplasmic reticulum stress could induce mitochondrial dysfunction and cell death through the disruption of Ca 2+ homeostasis [27]. Deniaud Conclusions From our preliminary data, we need to further explore the interaction between endoplasmic reticulum stress and mitochondrial dysfunction and identify the critical components, like some calcium channel proteins and small genetic regulators, in the ischemic injury. In summary, the present study first revealed the time-dependent variations of endoplasmic reticulum stress and mitochondrial dysfunction after the insult of ischemic injury, which suggested that mitochondrial dysfunction was secondary to endoplasmic reticulum stress. Our study provided a novel experimental foundation for further exploring the detail mechanism after ischemic injury, especially the interaction between endoplasmic reticulum and mitochondria, which could help identify more pharmacological targets for rescuing the myocardium after ischemic injury. Growth Hormone and the Human Hair Follicle Ever since the discoveries that human hair follicles (HFs) display the functional peripheral equivalent of the hypothalamic-pituitary-adrenal axis, exhibit elements of the hypothalamic-pituitary-thyroid axis, and even generate melatonin and prolactin, human hair research has proven to be a treasure chest for the exploration of neurohormone functions. However, growth hormone (GH), one of the dominant neurohormones of human neuroendocrine physiology, remains to be fully explored in this context. This is interesting since it has long been appreciated clinically that excessive GH serum levels induce distinct human skin pathology. Acromegaly, or GH excess, is associated with hypertrichosis, excessive androgen-independent growth of body hair, and hirsutism in females, while dysfunctional GH receptor-mediated signaling (Laron syndrome) is associated with alopecia and prominent HF defects. The outer root sheath keratinocytes have recently been shown to express functional GH receptors. Furthermore, and contrary to its name, recombinant human GH is known to inhibit female human scalp HFs’ growth ex vivo, likely via stimulating the expression of the catagen-inducing growth factor, TGF-β2. These limited available data encourage one to systematically explore the largely uncharted role of GH in human HF biology to uncover nonclassical functions of this core neurohormone in human skin physiology. Introduction The human hair follicle (HF) behaves as a neuroendocrine organ, even when isolated from systemic (blood flow, peripheral nervous system) stimuli, and shows hormone and receptor expression analogous to several central pituitary neuroendocrine axes [1,2]. Namely, the synthesis, secretion, and regulation of hormones of the hypothalamus-pituitary-adrenal (HPA) axis have been documented in human scalp HFs ex vivo. In the absence of systemic connections, cultured human scalp HFs express and respond to corticotropin-releasing hormone (CRH) and adrenocorticotropic hormone (ACTH), resulting in the HF synthesis of cortisol and activation of classical neuroendocrine feedback loops. Just as in the central HPA axis, the expression of pro-opiomelanocortin (POMC), the precursor for ACTH, α-MSH, and β-endorphin, is upregulated by CRH, while cortisol downregulates intrafollicular CRH protein synthesis [3]. All three POMC-derived peptides listed above regulate HF melanogenesis [4], while insufficient HF synthesis of melanotropic HPA axis hormones may contribute to HF greying [5]. HFs are also extra-pituitary sources of prolactin [6] and thyrotropin-releasing hormone (TRH) [7]. TRH and estradiol both regulate prolactin and prolactin receptor expression in human HFs in a similar manner as they do in the pituitary gland [8]. However, the HF expression of prolactin also underlies distinct controls, namely, it is not regulated by dopamine (as in the pituitary gland) but by substance P and the proinflammatory cytokine interferon-gamma [9]. Human HFs also express functional thyrotropin receptors [10], whose stimulation promotes intrafollicular mitochondrial activity and biogenesis [11]. Yet, the role of another key neurohormone, growth hormone (GH, somatotropin), in human HF biology remains insufficiently explored. After providing basic background on general GH biology, we delineate in the current review clinical and experimental evidence in support of GH as a potentially important regulator of human HF physiology. We argue that the limited available data encourages one to systematically dissect the role of GH in human HF biology in order to uncover nonclassical functions of this core neurohormone in human skin physiology and to develop novel GH or GH receptor-targeting neuroendocrine strategies for the therapeutic manipulation of hair loss (effluvium, alopecia) and unwanted hair growth (hirsutism, hypertrichosis). The Hypothalamus-Pituitary-Somatotropic Axis The hypothalamus-pituitary-somatotropic (HPS) axis refers to the neuroendocrine control of GH secretion and its downstream signaling. Growth hormone-releasing hormone (GHRH) produced in the hypothalamus upregulates GH gene expression and stimulates the release of GH from pituitary somatotrophs. Somatostatin (SST), also produced in the hypothalamus, inhibits GH release (but not GH synthesis) in pituitary somatotrophs. Both hormones act on the pituitary via the adenohypophyseal portal venous system. The orexigenic gastric peptide ghrelin stimulates hypothalamic GHRH secretion and pituitary GH release [12]. GH acts on peripheral cells in virtually every human tissue directly through the growth hormone receptor (GHR) ( Figure S1) and indirectly through the insulinlike growth factor 1 (IGF-1). The downstream signaling from activating the GHR varies by cell population but commonly involves the JAK2-STAT1/3/5 and/or MAPK pathways. Downstream, suppressors of cytokine signaling (SOCS) are known to inhibit GHR signaling effects. Interestingly, SOCS are upregulated by estrogen, which may cause sex-dependent differences when studying the HPS [13]. IGF-1, usually upregulated by peripheral GH signaling, inhibits GH secretion via negative feedback at the pituitary and hypothalamic levels [12] (Figure 1). The somatotrophic axis is closely linked to sleep and the circadian rhythm. GHRH has sleep-promoting effects, and GH secretion occurs in a pulsatile fashion, with maximal levels occurring after the onset of slow-wave sleep. Abnormal circadian rhythm disorders like narcolepsy are associated with abnormalities in the HPS axis [14]. Interestingly, the HF demonstrates circadian-dependent clock gene activity in the absence of central clock influences ex vivo [15]. PER-1 and BMAL-1 were both shown to regulate human HF cycling, as well as melanogenesis [16]. The peripheral clock activity has also been shown to be modulated by neurohormones like thyroxine [17], suggesting that other neurohormones like GH may also modulate HF biology directly or indirectly via modulation of clock genes. GH serum levels correlate with serum estradiol levels, with higher concentrations found in young people when compared with older people, as well as in females when compared to males. The bulk of GH secretion in males occurs during the night, whereas in females, nighttime secretion of GH corresponds to a smaller fraction of total daily GH secretion [18], which confirms that GH and clock genes are in direct connection. Growth Hormone-Releasing Hormone GHRH belongs to the secretin family of peptide hormones, which includes glucagon, secretin, vasoactive intestinal polypeptide, and others [19]. GHRH undergoes rapid enzymatic degradation in the blood via dipeptidyl peptidase IV [20] and therefore has negligent serum levels. While the primary function of GHRH is considered to be regulating pituitary GH synthesis and release, GHRH has been observed to be produced and have autocrine/paracrine effects in extra-pituitary tissues ( Table 1) that stimulate cell proliferation and inhibit apoptosis [19]. GHRH has been shown to promote wound healing by stimulating proliferation and survival of human dermal fibroblasts via signaling of the GHRH receptor [21]. stimulating proliferation and survival of human dermal fibroblasts via signaling of the GHRH receptor [21]. 1 GHR mRNA and protein are found in almost every human tissue, so only relevant hair follicle and skin cell populations are listed. The GHRH receptor (GHRHR) is a class II B GPCR found on the cell membrane of the pituitary somatotroph. Activating this receptor stimulates the exocytosis of GH and transcription of the GHRHR gene [20]. GHRHR activation is also vital to somatotroph cell proliferation via βγ subunit-mediated activation of Ras-MAP kinase and ERK phosphorylation. Extrapituitary GHRH activity is mediated by GHRHR and its splice variant type 1 (SV1), found in several cancerous and noncancerous human tissues, including apocrine glands and dermal fibroblasts (Table 1) [21,25,26]. GHRH signaling, via the GHRHR and the SV1, has been implicated in the growth of human apocrine tumors and metastatic melanoma [21,27]. GHRHR antagonists have been shown to inhibit cancer growth in vitro and in vivo and have anti-inflammatory and antioxidative effects [28,29]. Insulin-like Growth Factor-1 IGF-1, also called somatomedin-c, is a 70 amino acid protein with structural homology to pro-insulin. Gene expression of IGF-1 has traditionally been thought to be regulated by GH stimulation primarily in the liver. However, IGF-1 is expressed in most, if not all, tissues. GH stimulation is known to regulate both IGF-1 and many IGF-binding proteins (IGFBPs) [30]. IGF-1 receptor (IGF1R) is a transmembrane tyrosine kinase receptor consisting of two α subunits and two β subunits synthesized from a single mRNA precursor. Activating IGF1R leads to autophosphorylation of tyrosine kinases, leading to activation of several downstream signaling pathways, all of which stimulate the growth and proliferation of different cell populations [30]. For example, IGFs stimulate fibroblast proliferation, survival, migration, and production of growth factors like platelet-derived growth factors A and B [31]. IGF-1 is also known to be the most potent anagen prolonging growth factor in HFs [31,32]. In addition, IGF1Rs were found to be expressed in the hair matrix and outer root sheath keratinocytes of human scalp HF, where their signaling promotes proliferation and maintains the anagen phase [23,24,32,33]. Human fibroblast culture in vitro and human skin ex vivo has been shown to increase the expression of IGF-1/-2 and their receptors in response to GH and other factors [23,24]. IGF-1 plays a critical role in both skin and hair physiology, so it is not surprising to observe GH influencing hair growth. Ex Vivo, rGH Induces Premature Catagen Entry in Female Hair Follicles The known clinical hair phenotype associated with Laron syndrome or reduced GH serum levels (Table 2), which is also associated with decreased expression of IGF-1, would have led one to expect that the growth of organ-cultured human scalp HFs would be promoted by GH treatment. Unexpectedly, GH-treated microdissected human female scalp HFs showed premature catagen induction, most probably mediated via the upregulation of the potent catagen-inducting growth factor, TGF-β2 [23]. Even though IGF-1 expression in the outer root sheath keratinocyte was also upregulated, as expected, the overall increase of TGF-β2 expression in response to GH treatment may have been dominant over IGF-1, resulting in the observed growth inhibition in female HFs. However, these phenomena may not necessarily reflect only the direct effects of GHR stimulation within the HF itself. They might represent the overall HF response to complex neuroendocrine changes associated with excessive or insufficient GH/GHRmediated signaling. For example, chronic excessive GH signaling interferes with insulin and creates GH-induced insulin resistance [34]. Accordingly, many cutaneous findings (listed in Table 2) are found both in settings of insulin resistance and GH excess. Cutaneous Effects of Excessive or Reduced GH Receptor-Mediated Signaling Levels Numerous extrapituitary tissues and cells express mRNA and protein for GH, GHRH, along with their receptors, including the skin [22,43] and HF [23], but it remains unknown whether human skin and its appendages transcribe and translate the GH and GHRH genes in vivo ( Table 1). Most of what we currently know about the effects of GHR-mediated signaling arises from clinical observations in patients with excessive or insufficient serum levels of GH or defective GHR-mediated signaling. Pathologies leading to GH deficiency, like Noonan Syndrome, Turner Syndrome, and Prader-Willi syndrome, are associated with alopecia, telogen effluvium, and frontal hairline recession [35]. These syndromes are also associated with hypogonadism, which is also known to be associated with alopecia [44]. However, the interplay of androgens and GH on hair pathology is unknown and needs to be kept in mind. Male patients with GH deficiency of any cause were found to have reduced sweating [35]. These clinical findings clearly suggest that the HF and pilosebaceous unit [45,46] is a target for GH and can serve as a model system for studying how GH impacts a human mini-organ model as previously shown for other hormones (i.e., TRH, TSH, prolactin, CRH, ACTH). In this context, Laron syndrome, characterized by a loss of function mutation in the growth hormone receptor gene, leading to high levels of GH combined with low levels of IGF-1 [47], is particularly instructive. Laron syndrome is associated with sparse hair growth, various degrees of alopecia, and frontal hairline recession (Figure |
2A). Structural defects are found under microscopy such as grooving, tapered hair, pili torti and canaliculi, and trichorrhexis nodosa [48] as well as hypotrichosis [49] (Table 2). Recently, Laron syndrome has been mimicked in porcine models with GHR knockout mutations [50]. Both humans and the porcine model develop juvenile hypoglycemia with preservation of glucose tolerance and the development of normoglycemia with the onset of puberty [51]. Simulating GHR deficiency in organ-cultured human HFs by knocking down GHR using our established gene silencing methodology by transient siRNA transfection ex vivo [32,52,53] should instructively complement the use of this porcine model. deficiency in organ-cultured human HFs by knocking down GHR using our established gene silencing methodology by transient siRNA transfection ex vivo [32,52,53] should instructively complement the use of this porcine model. On the contrary, conditions of GH excess or deficiency have well-documented cutaneous manifestations ( Table 2). Increased plasma GH level in burn patients leads to improved re-epithelialization, increased granulation tissue, and reduced healing time [54]. Conditions with excess GH result in acromegaly in adulthood, and gigantism in childhood (before the epiphyseal growth plates fuse). Clinically, the leading cause of GH excess is a GH-producing pituitary adenoma (incidence/prevalence: 0.4-1.1 cases per 100,000/4-13 cases per 100,000 [55]), which occurs much more rarely than prolactin-secreting pituitary adenomas. McCune Albright syndrome, neurorofibromatosis-1, multiple endocrine neoplasia type 1, and Carney complex, which can also be associated with excessive GH serum levels, are even more rarely encountered orphan diseases [35]. Despite their rarity, the skin abnormalities seen in these diseases ( Table 2) provide important clinical pointers to the overall net impact of excessive GH serum levels on human skin and skin appendages in vivo. Excess GH is associated with hypertrichosis and hirsutism ( Figure 2B), as well as hyperhidrosis and increased sebum production [35]. Supplementing recombinant human growth hormone (rGH) may be key to therapies for encouraging wound healing and preventing or reversing aging-related damage. Treating elderly men with rGH has led to an increase in skin thickness [56]. Increased plasma GH in burn patients leads to improved epithelialization, increased granulation tissue, and reduced healing time [54]. Recombinant GH in human skin mice models has been shown to accelerate healing in pressure ulcer wounds [57]. Moreover, a large meta-analysis study suggested that rGH treatment may be used in the treatment of diabetic foot ulcers in humans [58]. Indeed, GH is known to promote cell proliferation, stimulate immune cells, and promote angiogenesis [57,59], all of which are known to be deficient in DFU patients. On the contrary, conditions of GH excess or deficiency have well-documented cutaneous manifestations ( Table 2). Increased plasma GH level in burn patients leads to improved re-epithelialization, increased granulation tissue, and reduced healing time [54]. Conditions with excess GH result in acromegaly in adulthood, and gigantism in childhood (before the epiphyseal growth plates fuse). Clinically, the leading cause of GH excess is a GH-producing pituitary adenoma (incidence/prevalence: 0.4-1.1 cases per 100,000/4-13 cases per 100,000 [55]), which occurs much more rarely than prolactin-secreting pituitary adenomas. McCune Albright syndrome, neurorofibromatosis-1, multiple endocrine neoplasia type 1, and Carney complex, which can also be associated with excessive GH serum levels, are even more rarely encountered orphan diseases [35]. Despite their rarity, the skin abnormalities seen in these diseases (Table 2) provide important clinical pointers to the overall net impact of excessive GH serum levels on human skin and skin appendages in vivo. Excess GH is associated with hypertrichosis and hirsutism ( Figure 2B), as well as hyperhidrosis and increased sebum production [35]. Supplementing recombinant human growth hormone (rGH) may be key to therapies for encouraging wound healing and preventing or reversing aging-related damage. Treating elderly men with rGH has led to an increase in skin thickness [56]. Increased plasma GH in burn patients leads to improved epithelialization, increased granulation tissue, and reduced healing time [54]. Recombinant GH in human skin mice models has been shown to accelerate healing in pressure ulcer wounds [57]. Moreover, a large meta-analysis study suggested that rGH treatment may be used in the treatment of diabetic foot ulcers in humans [58]. Indeed, GH is known to promote cell proliferation, stimulate immune cells, and promote angiogenesis [57,59], all of which are known to be deficient in DFU patients. Other Skin Phenotype Changes Associated with Signaling Abnormalities in the Hypothalamic-Pituitary Somatotropic (HPS) Axis SST analogues are used in therapies for several pathologies, including Merkel cell carcinoma, pancreatic endocrine neoplasms, and pituitary adenomas [60,61]. The use of SST analogues in therapies has been associated with scalp hair loss that resolves with discontinuation of treatment [37][38][39][40]. Classically, SST downregulates GHRH and GH signaling, which decreases IGF-1 signaling downstream. Decreased IGF-1 levels in dermal papillary fibroblasts of the hair follicle are found in balding scalp follicles when compared to nonbalding scalp HFs [62]. Furthermore, low circulating IGF-1 levels were associated with hair loss in middle-aged women [41]. In one study observing patients post transsphenoidal adenomectomy, 54% of patients who had acromegaly experienced hair loss 3 to 6 months postoperatively, compared to 6% of patients who had nonfunctional adenomas [36]. This study also showed hair loss was more common in patients cured by surgery than in non-cured patients, i.e., hair loss was more common in patients that experienced acute decreases in GH and/or IGF-1, behaving as a relative insufficiency [36]. More interestingly, topical liposomal IGF-1 was associated with more rapid hair growth and thicker hair in a hamster model [63]. In mice, GHRH treatment was found to reverse age-related changes, increasing the thickness of the epidermis and dermis, increasing moisture content, and improving the morphology of the skin tissue and collagen fibers [64]. GHRH deficiency in a Brazilian cohort showed delayed pigmentation, and reported to have youthful hair and no alopecia, even with profoundly decreased serum GH and IGF-1 levels [42]. This cohort aligns with GH acting as an inhibitor of hair growth previously seen ex vivo in female donors [23]. This population did show wrinkly skin, which implicates the HPS axis' role in age-related damaging of human skin even more. Major Open Questions The surprising hair growth-inhibitory results reported above ex vivo, including the upregulation of TGF-β2, question conventional concepts that the direct stimulation of GHR in human peripheral tissues always has a growth-stimulatory effect. It also raises the question to which extent GH/GHR-mediated signaling has tissue-and context (sex)dependent outcomes in human skin and its appendages. We know that a strong positive correlation exists between excess GH levels and insulin resistance and that both higher GH serum levels and insulin resistance are found in females. This may, at least partially, explain why the upregulation of IGF-1 after GH treatment in female scalp HFs ex vivo did not prolong hair growth. One possibility to answer this question might be to inves-tigate if the same GH concentration tested in female HFs prolongs the anagen phase in male scalp HFs via prominently increasing IGF-1 over any potential increase of TGF-β2 expression. Clinically, primary decreases in GH and IGF-1 have been associated with hair loss and alopecia [36], while decreases in GH and IGF-1 due to GHRH deficiency have not [42]. Furthermore, hair growth stimulation has been seen in acromegaly patients [35]. Understanding how hair and skin respond to GH and GHRH stimulation separately may be key in understanding these discrepancies. Dissecting the effects of GH stimulation in human HF and skin organ culture [65][66][67][68] in the presence/absence of GHR siRNA, followed by laser capture microdissection-based RNAseq analysis of defined HF and skin compartments or single-cell RNAseq, should help to clarify which cell population(s) in human skin are most receptive to GHR stimulation. These studies can also shed light on how they differ in their target genes and signaling pathways. Furthermore, GH signaling in the skin and HF still needs to be further characterized in the context of GHRH, SST, and IGF-1 expression, with any of their potential negative feedback mechanisms. Clinically, GH excess and deficiencies are often accompanied by other hormonal abnormalities, such as hypothyroidism and hyperprolactinemia, both of which are also known to affect hair growth [6][7][8]10,[69][70][71]. It will then be important to understand how these GH effects are amplified or hampered by other hormone profiles (e.g., prolactin, TRH, TSH). Understanding the growth hormone's effect on the HF and skin and its related signaling pathways is key to understanding future clinical therapies for dermatopathology. It might be interesting to investigate the expression levels of the different members of the HPS axis, keeping in mind potential sex differences in healthy skin and HFs as well as in some hair growth disorders (telogen effluvium, female/male pattern hair loss, alopecia areata). Moreover, since GH influences the expression level of IGF-1 and TGF-β2, it would be interesting to evaluate the impact of GH/GHR signaling on the hair follicle immune privilege (and consequently in alopecia areata) as both growth factors are well-known immune privilege guardians. This raises the question regarding the potential insensibility of the hair matrix and outer root sheath keratinocytes to some immune privilege guardians under excessive GH stimulation, and if that contributes to a patient's susceptibility to hair disorders, such as alopecia areata, with excessive GH stimulation. The effect of GHRH and GH on hair physiology and wound healing should be further explored as well. GH and GHRH have both shown wound healing properties ex vivo and shown to have a strong effect on human dermal fibroblast proliferation and differentiation [21]. Robust clinical trials with rGH or rGHRH have yet to be done regarding wound healing in humans. This physiological loss of the HPS axis hormones with age may be a key player in the aging process. Restoring GHRH and GH signaling could very well be a key player in anti-aging therapies to inhibit or reverse age-related damage of the skin. The hormones of the HPS axis are known to decrease with age, and mouse models have found antiaging properties with treatment of GHRH via reduction of malondialdehyde and matrix metalloproteinases [64]. Recombinant growth hormone is very well-established and is safely used therapeutically [72]. GH and GHRH have already been shown to have physiological effects on catagen promotion, carcinogenesis, and wound healing. Understanding the potential effects of the somatotrophic hormones in the HF can further help regulate hair cycling and hair pathologies and guide novel therapies in wound healing and cancer. Indeed, as suggested by our study [23], a slight change in GH levels may have a dramatic effect on hair growth, suggesting GH levels and GHR stimulation needs to be fine-tuned and tightly regulated. It might then be essential to measure with precision GH levels (not only serum levels) to avoid unwanted effects, suggesting that GH disorders might require personalized treatment. Conclusions The human HF has been shown to express a wide array of neurohormones and even display negative feedback mechanisms that mirror central neurohormone axes. Both GH release and the HF exhibit circadian-dependent regulation that may be interdependent. Pathological GH serum levels produce profound clinical effects on hair. Excess GH levels, and therefore excess GHR stimulation and excess IGF-1 levels are associated with hypertrichosis and hirsutism. Absent GHR stimulation, and thus severely decreased IGF-1 levels, is associated with alopecia, telogen effluvium, frontal hairline recession, as well as severe HF structural changes like pili torti et canaliculi and trichorrhexis nodosa. GHRs are found in virtually every human tissue and prominently in the HF. Stimulation of the GHR may have a profound effect on hair growth. Ex vivo, female human scalp HFs were inhibited by GH stimulation, suggesting a complex sex-dependent interaction between hair growth and GH stimulation. Further investigation of how GH and GHR stimulation affect hair follicle biology can guide feasible treatment options for different hair disorders. A qualitative analysis of Medicaid beneficiaries perceptions of prenatal and immediate postpartum contraception counseling Objectives: In the United States, about four out of every ten births are financed by Medicaid, making it a program that is key in addressing racial disparities in maternal health. Many women covered by Medicaid have access to prenatal and immediate postpartum contraception counseling that can aid them in their postpartum contraception decision-making. However, existing inequities within Medicaid and a history of reproductive harms targeting Black women and women with low incomes may contribute to women with Medicaid having different experiences of contraception counseling. This qualitative study explores how Black women and White women insured by |
Medicaid perceive prenatal and immediate postpartum contraception counseling and identifies additional factors that shape their contraception decision-making. Methods: We conducted semi-structured interviews with 15 Medicaid beneficiaries who delivered at a public teaching hospital in North Carolina. Interviews focused on women’s beliefs about planning for pregnancy, experiences with prenatal and immediate postpartum contraception counseling, and perceived need for postpartum contraception. We used a priori and emergent codes to analyze interviews. Results: Seven Black women and eight White women completed interviews 14–60 days postpartum. All women reported receiving prenatal and immediate postpartum counseling. Several women described receiving prenatal counseling, reflective of patient-centered contraception counseling, that helped in their postpartum contraception decision-making; one woman described receiving immediate postpartum counseling that helped in her decision-making. Some Black women reported receiving unsupportive/coercive contraception counseling. In addition to contraception counseling, past reproductive health experiences and future pregnancy intentions were salient to women’s contraception decision-making. Conclusions: Prenatal and immediate postpartum contraception counseling can help some Medicaid beneficiaries with their postpartum contraception decision, but past reproductive health experiences and future pregnancy intentions are also relevant. Counseling that does not consider these experiences may be harmful, particularly to Black women, further contributing to racial disparities in maternal postpartum health outcomes. Introduction Approximately 42% of all births in the United States are financed by Medicaid. 1 Given the size and scope of the program, Medicaid plays an important role in addressing racial disparities in maternal health outcomes, including ensuring access to postpartum contraception. [2][3][4] Providers are encouraged to discuss options for postpartum contraception with all women during pregnancy, and again after delivery to help them reach their reproductive goals. 5 The American Academy of Pediatrics and the American College of Obstetricians and Gynecologists recommend providing contraception counseling to all women during their prenatal visits and immediately postpartum. 5 One multi-state sample found that postpartum contraception use was highest among women who received both prenatal and immediate postpartum contraception counseling. 6 For women with low incomes, contraception counseling during the prenatal and immediate postpartum periods is key. Women who qualify for Medicaid because of pregnancy may lose access to helpful contraception services within weeks after delivery. 7 Although there are increased state efforts to extend postpartum Medicaid coverage, 8 inequities within the Medicaid program persist 9 and may negatively impact some women's postpartum contraception decision-making. 10,11 Racial inequities within Medicaid are the result of structural racism which is the "normalization and legitimization of an array of dynamics-historical, cultural, institutional and interpersonal-that routinely advantage whites while producing cumulative and chronic adverse outcomes for people of color." 12 These inequities are evident in the racial disparities among Medicaid beneficiaries. For example, Black women are twice as likely to have their births covered by Medicaid, compared to White women. 13 Black women's overrepresentation in the Medicaid population is the result of structural inequities that contribute to Black women having lower incomes. 14 Inequities in Medicaid are further exacerbated by obstetric racism. Obstetric racism, systematic racism effecting the ways in which women are treated during conception, pregnancy, labor and delivery, and postpartum, 15,16 harms contraception counseling discussions that take place between providers and women. For example, prior research has shown that Black and Latina women with low incomes are more likely to report being told to limit their childbearing compared to middle-class White women. 17 Other studies have found that providers are more likely to recommend permanent contraception or long-acting reversible contraception (LARC) to Black women and women with low incomes. 18,19 In addition, there is a history of reproductive coercion (including the forced sterilization of poor, Black, Latino, and developmentally disabled individuals) and the targeted distribution of some contraception methods to poor Black communities. 20 For some women, this history deepens their mistrust for contraception and contraceptive services. 21 Prenatal and immediate postpartum contraception counseling can offer important benefits and may be particularly useful to women with Medicaid, but because of inequities within the Medicaid program and a history of reproductive harms targeting Black women and women with low incomes, it is important to assess the contraception counseling experiences of women insured by Medicaid. This article examines Black women's and White women's, insured by Medicaid, decisions about postpartum contraception, including their perceptions of contraception counseling. We also describe the beliefs, knowledge, and reproductive health needs that influence postpartum contraception decision-making. Procedures We conducted semi-structured interviews with women who had live, singleton, full-term births at a public teaching hospital in North Carolina between December 2019 and October 2020. Participants were Medicaid beneficiaries, aged 18 years or older who identified as non-Hispanic Black or non-Hispanic White and spoke English. Women were excluded if they experienced a complicated birth, with a serious adverse event affecting the mother or infant. Women were recruited through (1) inpatient recruitment and (2) a website that provided information about research opportunities available to the public. From December 2019 to February 2020 the first author (L.Y.) reviewed the daily inpatient census and approached all eligible women during their delivery hospitalization about the study. From May 2020 to December 2020 the website listed relevant study information. Women who contacted study staff through the study website were screened to determine their eligibility. The first author (L.Y.) contacted all women who expressed interest in the study by phone, email, or text message approximately 14 days after delivery to schedule telephone interviews. The first author (L.Y.) conducted all interviews with participants. All eligible and interested participants received a written copy of the consent. All participants completed the verbal informed consent process before the recorded interview as part of the interview procedures. All participants were sent a US$25 gift card after the interview. All study procedures were approved and exempted by the University of North Carolina Institutional Review Board (study no.: 19-0798). Measures Interview questions were derived from the Behavioral Model of Health Services Use (BMHS) (Figure 1). 22 This model posits that various factors influence an individuals' decision to seek or receive care for a health condition. We used this model to examine postpartum contraception decision-making. Factors highlighted by the model include predisposing factors such as beliefs about planning for pregnancy, enabling factors such as contraception counseling received during the prenatal and immediate postpartum periods, and need factors such as perceived need for postpartum contraception. 23 The interview questions explored several topics including beliefs about contraception, use of contraception, choice of contraception after delivery, sources and resources used to learn about contraception, experiences with prenatal and immediate postpartum care and contraception counseling, and perceptions about short-interval pregnancies. Interview questions were piloted before participants were enrolled. Information collected from the pilot interviews were used to improve the interview format and not included in the final analysis. Data analysis The interviews were recorded and transcribed verbatim and analyzed by two study team members, one Black woman (L.Y.) and one White woman (B.W.), using template analysis. The coding template was based on the BMHS. Both team members used MAXQDA, version 12 (VERBI Software GmbH, Berlin), to code each transcript independently. The team members reviewed each transcript together to reach consensus on themes and applied the appropriate codes for each interview. As new themes emerged, team members re-reviewed interviews using the updated coding template. We reached data saturation when we could not identify any additional codes. A copy of the Consolidated Criteria for Reporting Qualitative Research (COREQ) checklist and the interview guide in supplementary materials. Results We completed interviews with seven Black and eight White women 14 to 60 days postpartum. The participants ranged in age (20-38 years) and had various pregnancy histories (1-10 pregnancies; Table 1). Women reported receiving prenatal care from teaching hospitals, local health departments, or other medical practices. Four women reported transferring care from one practice to another practice for their prenatal care. Thirteen women reported that they attended all or most of their prenatal visits. Twelve of the 15 women reported selecting a contraceptive method by the time of their interview. Three Black and two White women selected bilateral tubal ligation. Three white women selected LARC that they planned to receive at their postpartum visit. Two women, one Black and one white, were considering LARC, but had not made a final decision. None of the five women who selected or considered LARC reported receiving prenatal counseling about the option to receive a LARC method after delivery, but before discharge. A summary of findings is available in Table 2. We describe five themes among the predisposing, enabling, and need factors of the BMHS. When appropriate, we note differences within themes among the Black respondents and White respondents. We present quotes with the pseudonyms (selected by the first author) and age of the participants. Theme 1. Beliefs, knowledge, and other sources of information shape perceptions of contraception Beliefs about planning for pregnancy. There were no notable racial differences among participants about their beliefs. All women described planned pregnancies as more desirable than unplanned pregnancies. However, they acknowledged that planning for pregnancy has challenges, especially if they experience difficulties getting pregnant. They also reported stable income, family leave, and emotional support from partners, friends, and family as resources that women should assess when planning for pregnancy. Knowledge and experience with short-interval pregnancy. All participants reported that either they or someone they knew became pregnant shortly after giving birth. Several women specifically stated that pregnancy within "six weeks" or before the "six-week postpartum visit" was possible. Most women viewed a short-interval pregnancy as undesirable and reported planning to use contraception after delivery to prevent this. A few women with recent short-interval pregnancies noted that their current pregnancy after the short-interval pregnancy was a reason for their desire to use a long-acting method. Information from other sources. All but one participant reported receiving information or feedback about specific types of contraception from friends and family, online searches, and/or social media. A few women reported trusting the information they received from other sources more than information they received from their provider. Multiple women reported that discussions with family were encouraging; one Black woman and one White woman found advice from family members was less helpful because it did not consider their specific needs. For example, Ashely, a 28-year-old White woman described feeling frustrated by her mother's suggestions to get bilateral tubal ligation: All women received one of three types of prenatal counseling: (1) active, ongoing counseling, (2) limited but satisfactory counseling, and (3) unsupportive, coercive counseling. Four White women and two Black women described their prenatal contraception counseling as helpful in their decision-making. Three Black women described receiving unsupportive, coercive counseling. Theme 3. Immediate postpartum counseling has limited effects on contraception decision-making All women received postpartum contraception counseling during their delivery hospitalization. Generally, these discussions focused on confirming the contraception decision made during prenatal care. One White woman found these discussions helpful in her decision-making. The three Black women who reported unsupportive, coercive prenatal counseling reported continuing to feel pressured about their contraception choice after delivery. Theme 4. Medicaid coverage influences access to contraception and prenatal care The low/no-cost of contraception under Medicaid was relevant to some women's decision-making and many women were able to easily access Medicaid. For a few the 60-day limit on postpartum coverage influenced their contraception decision, and for one woman Medicaid approval came later in her pregnancy impacting her access to prenatal care. Need Factors Theme 5. Physiological effects of contraception and pregnancy impact decisions about contraception All women reported prior contraception use. Negative health effects resulting from the use prescription contraception (e.g. weight gain, pain or discomfort, mood swings, irregular bleeding), influenced their selection of postpartum contraception. Their plans and desire for more children were also salient to their decision-making. It aggravated me because . . . she was like, "You don't need no more kids, you don't need any more kids, you don't need any more kids" and when I had my third baby . These women wanted conversations that were more considerate of their choices and discussed more comprehensive information about their contraception options. Theme 3. Immediate postpartum counseling has less effect on contraception decisionmaking Generally, participants who did not receive permanent contraception reported that the postpartum contraception counseling they received before discharge focused on confirming the method they selected during prenatal care. A small number of women reported slightly different experiences with postpartum contraception counseling during the immediate postpartum period. In |
one case, a White woman reported her postpartum provider suggested a different method than her prenatal provider because of potential contraindications with the previously recommended method. Two Black women who desired a bilateral tubal ligation immediately following delivery had their procedures rescheduled due to complications they experienced. A White woman who was undecided but considering an implant was offered immediate postpartum implant placement but declined. The Black women who reported unsupportive/coercive prenatal counseling reported continuing to feel pressured about their contraception choice after delivery. Jackie (a 20-year-old Black woman) and Angel (a 21-year-old Black woman) ultimately selected contraception inconsistent with their preference to avoid further discussion: . . . They were on me in the hospital like, "Birth control! Birth control!" I was just like, Oh my god, I just had a baby. I don't know if I want to be on birth control because I'm trying to heal from a scar . . . Yeah, that's when I said it again. I said, "I think I am just going to get the IUD," because they asked me when I got there, when I was in labor, and they asked me after labor. (Jackie, 20 years) They asked me again and I told the lady again, and then she tried to talk to me out of it also, so then that's when I just decided on the pills. Because I had already forgot about all the other choices. I was just, "I'll just do the pills." Because you know, I just had a baby I was still drugged up and everything so I was like I'll just do the pills. (Angel, 21 years) Two women described challenges related to contraception uptake and postpartum care after discharge. Jackie, a 20-year-old Black woman, contacted her postpartum provider after discharge and felt dismissed when she told the provider she was not getting contraception: I was supposed to go to the doctor, you know, right after you give birth you go to the doctor because I had a C-section-but I feel like the fact that I wasn't getting birth control is the reason they pushed my appointment so far back. (Jackie, 20 years) Amber, a 27-year-old White woman, said she was not given adequate information about when or how to take the pills. When she called her postpartum provider for help, she was told she needed to attend her postpartum appointment to receive more information. Theme 4. Medicaid coverage influences access to contraception and prenatal Having Medicaid also influenced women's choice of contraception. Multiple women, specifically noted that Medicaid covered the cost of their chosen contraceptive method, including a few White women who described choosing bilateral tubal ligation because Medicaid covered the procedure costs. However, a few women experienced coverage challenges with Medicaid. For example, two women knew that Medicaid for Pregnant Women ends 60-days postpartum, which impacted their contraception decision-making. Crystal, a 35-year-old Black woman, who was deciding between bilateral tubal ligation and an IUD said, Emily, a 24-year-old White woman, was concerned that if she had issues with a selected contraception method, she would be unable to change it after her Medicaid coverage ended: I wasn't really too big on taking that risk of having the same problems over again, and then me having to go through another surgery in the future. Because that's something that I was eventually going to have to get done if the birth control didn't work . . . Which is harder to get our insurance to pay for, too . . . And I don't have Medicaid. I did have pregnancy Medicaid, so the thing with that is once you've surpassed your six to eight-week postpartum period, what do you do after that? So even though you can still get birth control options for free, if I decided to have a [tubal ligation] after that point, it would be like, okay, well, how am I going to pay for this? That's an invasive surgery, and at least a one-night hospital stay, so ten grand at the cheapest. That's a lot of money for the average person to come up with . . . (Emily, 24 years) Several women noted that applying for Medicaid was simple, and their local health departments were instrumental in their application and approval process. One Black woman and one White woman reported that their Medicaid approval came late in their pregnancies. For the Black woman that delay contributed to her beginning prenatal care later in pregnancy. Theme 5. Physiological effects of contraception and pregnancy impact decisions about contraception All participants reported prior contraception use. Their postpartum contraception preference was primarily shaped by previous experiences including side-effects (e.g. weight gain, pain or discomfort, mood swings, irregular bleeding). Women who selected the Mirena IUD postpartum had very positive prior experiences, including not having a period. Two of the three women who choose pills and two who were undecided, chose a different postpartum contraceptive because of prior negative experiences with previous methods. Participants' future pregnancy intentions were shaped by their pregnancy experiences. About half of the participants indicated that their recent pregnancy was planned, with no distinct racial differences among women who described their pregnancy as planned. All the women who received a bilateral tubal ligation and two of the women who selected the Mirena IUD were clear that they did not want to be pregnant again. While a few women attributed this decision to severe physical health challenges during pregnancy; others stated their age and spacing of their children as a reason they did not want to be pregnant again. The remaining women did not state their future pregnancy intentions or were open to a future pregnancy. Discussion Our findings suggest that there are racial differences within the Medicaid program affecting Black women's and White women's experiences with prenatal and immediate postpartum contraception counseling. Postpartum contraception counseling can help some Medicaid beneficiaries choose methods consistent with their reproductive goals; however, counseling that fails to consider women's reproductive goals may be less helpful and potentially harmful, especially for Black women. Although all women in this study received prenatal counseling, only half received active, ongoing prenatal contraception counseling. Patient-centered contraception counseling has emerged as a key strategy to help all women achieve their reproductive goals. 24,25 Patient-centered contraception counseling involves building trust through an interpersonal relationship; eliciting and appropriately responding to patients' preferences; providing accurate information about utilization and side-effects; helping patients make contraception decisions based on their preferences and options available to them; and recognizing root causes of disparities. 24,26,25 Consistent with previous literature, 24,26,27 counseling reflecting a patient-centered approach that included elements of shared decision-making, prioritized women's preferences, provided accurate information, and offered guidance was especially useful to women considering multiple conception options. Of note, three Black women reported receiving unsupportive/coercive counseling and feeling forced to make contraceptive choices they did not want. There is growing evidence that Black women feel pressured to make certain contraceptive choices, including using LARC 17,28,29 and that providers, in fact, make different recommendations for contraception based on patients' race. 19 This type of dismissive and coercive treatment is reflective of obstetric racism. 15,16 Unsupportive/coercive counseling is inconsistent with patient-centered counseling, as well as reproductive justice. 30 Reproductive justice recognizes the right to bodily autonomy, deciding when and if to have children, and parent children in ways that are reflective of individuals' goals and values. 31 Postpartum contraception counseling should be grounded in the reproductive justice framework and acknowledge the systems and history of injustices that contribute to disparities. 25,32 Coercive practices erode patients' trust in providers and hinder equitable reproductive health outcomes. Although there is evidence that immediate postpartum contraception counseling is associated with postpartum contraception use, 6 our findings suggest discussions during this period have less impact on most women's postpartum contraception decision-making. This is consistent with previous research showing that contraception counseling before discharge was brief and "unhelpful." 33 The immediate postpartum setting may be suboptimal for contraception counseling because providers and new parents are focused on other postpartum issues. 33 Conversely, immediate postpartum counseling may be beneficial to women who desire LARC. 34 Although only one woman who was considering an LARC was advised about immediate postpartum LARC in this study, prior research finds women who are considering immediate postpartum LARC prefer frequent, quality prenatal and immediate postpartum discussions, where their autonomy and valid information are prioritized. 35,36 Counseling women about immediate postpartum LARC during their inpatient postpartum stay is important and should be coupled with prenatal counseling. In our study, we found that women's prior contraception use and future pregnancy intentions were relevant to their postpartum contraception decision-making. This is similar to previous work suggesting that contraceptive choice and use are associated with both pregnancy and method-related experiences (e.g. prior unintended pregnancies, dislike of other contraceptive methods, future pregnancy intentions). 37 The women also believed that planned pregnancies were more desirable than unplanned pregnancies. Unlike some prior research suggesting that women with low incomes may have less clear pregnancy intentions, 38 half the women reported planning their recent pregnancy, and many described clear future pregnancy intentions. Consistent with prior literature, 6,39 about half of the women reported that the cost of contraception under Medicaid contributed to their choice of contraception. Limits on Medicaid coverage for pregnant women diminish women's access to comprehensive postpartum contraception counseling and options. 7,40 In addition, women relied on other sources of information, including family and friends, for contraception information. This aligns with research that suggests that social networks influence women's postpartum contraception decisions. 41 There is also evidence that among women with low incomes, their desire for contraception is influenced by their partners and families. 42 Acknowledging these sources of information may be important to women and help them feel more confident with their postpartum contraception decision. In this study, we explore women's experiences with contraception decision-making, before they lose their Medicaid coverage. This represents a unique time frame to collect information about perceptions of contraception counseling and reproductive health care. We also used an existing conceptual model to better understand the factors that influence women's postpartum contraception decisions. This model allowed us to explore various factors, better illuminating the influence of those factors on contraception selection. Despite these strengths, this study has some limitations. First, participants were non-Hispanic Black and non-Hispanic White women and may not reflect the experiences of women from other racial/ ethnic groups. Although Black women and White women make up more than 83% of all Medicaid births in North Carolina, 43 future research should explore the experiences of other historically marginalized populations. Our participants also had uncomplicated deliveries and healthy infants which may have also influenced their experiences with contraception counseling. Second, the study was conducted in one academic medical center in North Carolina. Given our study sample we may have limited the factors influencing postpartum contraception we were able to uncover or the role of hospital environments on these outcomes. Finally, we did not account for provider characteristics (e.g. race, years of licensing, type, location). Future research should explore how such characteristics may influence prenatal and immediate postpartum contraception counseling. Conclusion In our study, although women insured by Medicaid are receiving prenatal and immediate postpartum contraception counseling, many are not receiving patient-centered contraception counseling. Notably, some Black women received counseling that is unsupportive and/or coercive, which could lead to women failing to achieve their reproductive goals and worsen racial inequities in maternal health outcomes within the Medicaid program. These findings are important for state Medicaid agencies and other stakeholders invested in developing equity-focused solutions to address maternal health disparities in Medicaid. Expanded Medicaid coverage beyond 60-days postpartum is one, fundamental solution to resolve disparities. Additional efforts should focus on addressing obstetric racism, including creating structures that incentivize providers to deliver patient-centered counseling and data measures that assess the quality of contraception counseling for women with Medicaid. Providers and facilities responsible for delivering postpartum contraception counseling should practice counseling is grounded in the reproductive justice framework and prioritizes patients' individual previous experiences and pregnancy intentions. Developing these types of solutions requires Medicaid agencies to meaningfully engage women with lived experience to help inform programmatic changes to Medicaid. It is also important to acknowledge and understand the policies that limit women with low incomes access to contraception as well as how the changing |
landscape of women's access to abortion care may affect women's decisions about contraception. Addressing inequities in prenatal and immediate post partum contraception counseling can help reduce racial disparities in maternal health outcomes. Ethics approval and consent to participate This study was approved and exempted by the University of North Carolina at Chapel Hill Institutional Review Board (study no.: 19-0798). All eligible and interested participants received a written copy of the consent. All participants completed the verbal informed consent process before the interview that was recorded as part of the interview procedures. Consent for publication The written and verbal consent included consent to publish. All non-essential identifying details have been omitted. The effect of vertical centering and scout direction on automatic tube voltage selection in chest CT: a preliminary phantom study on two different CT equipments Highlights • Patient vertical off-centering affects the function of ATVS and TCM.• Patient vertical off-centering has impacts on radiation dose and image quality.• CTDIvol was increased up to 91% with improper patient positioning.• Scout direction affects the function of ATVS and TCM.• The effect of vertical off-centering on ATVS and TCM differ between CT vendors. Introduction Medical radiation exposure to patients in diagnostics has increased mostly because of the growing use of computed tomography (CT) [1][2][3]. The increased use of CT in medical imaging has driven optimization efforts, including both technical and clinical approaches, to decrease radiation dose for the patients while maintaining the sufficient image quality for diagnosis. Technical optimization tools include, for example, tube current modulation (TCM) techniques, automatic tube voltage selection (ATVS), adaptive beam collimation, and iterative reconstruction methods [4][5][6][7][8][9][10][11][12][13][14][15][16][17][18][19]. Despite the effective technical CT optimization tools, the role of user remains important to achieve optimal results both in terms of image quality and radiation dose. Several studies have previously shown remarkable effects of patient off-centering on patient radiation dose and image quality due to function of beam shaping filters and geometric magnification/minification resulted in the scout images (planning radiographs) [20][21][22][23][24][25][26][27][28][29][30][31][32][33][34][35][36]. The function of a bowtie filter is to allow maximum x-ray intensity to the thickest part of a patient and to reduce x-ray intensity in peripheral areas with less attenuation, thereby reducing x-ray scatter and radiation dose of surface tissues [37]. The optimal function of the bowtie filter and TCM techniques assume that the patient is centered on the scan isocenter. Recently developed ATVS approaches aim to maintain a constant contrast-to-noise ratio between the examinations performed for different sized patients with as low radiation exposure as possible [14,19,38,39]. ATVS methods use scout images to determine the net attenuation profile of the patient. Based on the attenuation, ATVS algorithms determine the most dose-efficient combination of tube voltage and tube current settings to provide the needed image quality on a CT scan. Therefore, the ATVS presents a more general optimization tool which includes the TCM functionality to select the most optimal scan settings. While doing so, the ATVS also accounts for the examination type (e.g. CT angiography, contrast-enhanced scan, soft tissue or bone without contrast administration) in order to reach the optimal clinical image quality considering the availability and level of iodine contrast enhancement [19,38]. The impact of off-centering and scout direction on TCM have previously been extensively studied [20][21][22][23][24][25][26][27][28][29][31][32][33][34][35][36]. However, the effect of off-centering on ATVS have only been studied for dose and not for image quality [30]. The effect of scout direction on ATVS has not been investigated. ATVS and TCM methods are strongly interconnected and their technical implementations vary between manufacturers. Therefore, it is necessary to study the impacts of these optimization tools in a comprehensive manner with clinical scan protocols. The aim of the current study was to determine the combined effect of the scout direction and patient's vertical off-centering on ATVS and related TCM in chest CT examinations. Both radiation dose and image quality in these scans were investigated. The measurements with an anthropomorphic chest phantom were performed using two CT systems from different vendors. Phantom measurements An anthropomorphic chest phantom (IMRT Thorax Phantom Model 002LFC, CIRS, Norfolk, USA) was scanned in a supine position on Siemens SOMATOM Definition Flash (Siemens Healthcare, Erlangen, Germany) and GE Revolution HD (GE Healthcare, Milwaukee, Wisconsin, USA) CT systems using clinical protocols for chest CT examinations. The four scanning protocols were 1) a chest CT protocol with ATVS and with contrast administration, 2) a chest CT protocol with ATVS and without contrast administration, 3) a chest CT protocol for pulmonary embolism with ATVS, and 4) a routine chest CT scan protocol with a fixed 120 kVp tube voltage. All protocols utilized TCM and were imaged without contrast agent. The scanning parameters and settings of the protocols are presented in Table 1. The phantom had an elliptical cross section and approximated human torso in proportion, density, and two-dimensional structure (Fig. 1). The phantom dimensions were 30 cm x 20 cm, and consisted of axial slabs of proprietary epoxy materials. According to the manufacturer, linear attenuations of the simulated tissues in the phantom were within 1% of the actual attenuation for water and bone, and within 3% for lung in the diagnostic X-ray energy range. The structure of the phantom did not vary in zdirection. The scan length was adjusted to 27.4 cm, leaving the phantom support parts out of the scan range. Large ("body") bowtie filters were used for the scans. The phantom was scanned in a helical mode at five different vertical levels of the patient table (from 6 cm below to 6 cm above the scan isocenter in 3 cm steps, see Fig. 1). The reference height position (0level) was set using the lasers and landmarks on the phantom. To evaluate the effect of scout direction on ATVS and TCM, the phantom was scanned three times with each protocol and table height combination: using anterior-to-posterior (AP), posterior-to-anterior (PA), and lateral (LAT) scouts. For each scanning protocol and scout direction, the tube voltages selected by ATVS and the volume CT dose index (CTDI vol ) values from the scanner console were recorded. Additionally, the apparent phantom size, measured as the projected width from the scout image, was determined in each vertical height position. Therefore, the apparent phantom size changed according to projected width magnification in the scout image. Image analysis The relative changes in image quality was evaluated from 0.75 mm (Siemens) and 0.625 mm (GE) thick axial reconstructions (512 × 512 pixels) of the phantom. The scans were reconstructed with clinically used reconstruction kernels/filters: pulmonary embolism scans were reconstructed with i26f kernel, whereas i31f kernel was used for other protocols on the Siemens CT system, and a standard reconstruction filter ("STND") was used for all datasets on the GE CT system. Iterative reconstruction was used in all reconstructions: Safire level 2 for Siemens, and 40% ASIR-filtered back projection blending for GE. The axial display field of view (DFOV) was adjusted to 38 cm resulting in a pixel size of 0.742 × 0.742 mm 2 in all the images. The image noise was estimated by calculating CT number standard deviations (SDs) in six regions-of-interest (ROIs) shown in Fig. 2. ROI 1 was placed inside the spine and ROIs 2-6 were placed in the soft-tissue equivalent material. The SDs were calculated for 10 slices with 10 mm intervals (i.e. slabs 1-10 in Fig. 1B comprising a total z-direction coverage of 10 cm) to avoid the small gaps between phantom slabs. For each ROI, the mean of the ten SDs was reported as the noise value and the SD (of the 10 SDs) as the noise error. Additionally, the effect of vertical centering was visualized for the fixed 120-kVp protocol. Absolute differences in CT numbers and relative noise differences between the off-centered and properly positioned phantom images (used as a reference) were calculated. The absolute CT numbers were calculated as the voxel-wise means and the noise maps as the voxel-wise SDs over the aforementioned 10 slices. Tables 2-5 show the effects of phantom's vertical off-centering on the function of ATVS and the CTDI vol values in the four investigated protocols. There were differences in the behavior of ATVS techniques between the Siemens and GE CT systems for the studied geometry. The Siemens' ATVS technique selected systematically higher tube voltages after AP and PA scouts compared to lateral scout, whereas the GE's ATVS technique tended to select more likely a lower tube voltage after AP and PA scouts compared to lateral scout. Moreover, GE's ATVS technique presented less voltage variance compared to the Siemens' ATVS approach. In the GE CT system, the tube voltage varied only in the pulmonary embolism scans, whereas in the Siemens CT system, variance was seen in all the protocols utilizing ATVS. In pulmonary embolism scan, the magnification or minification of the spine caused changes in the tube voltage selection on the Siemens system in AP scout (off-centering -3 cm and -6 cm) compared to PA scout (off-centering +3 cm and +6 cm). ATVS chose 100 kVp tube voltage in the former, and 120 kVp in the latter cases. Apparent width, radiation dose, and the function of ATVS As could be expected based on the projected magnification of the scout image, the CTDI vol values were the greatest with PA scout when the phantom was centered at the lowest table height position, and the lowest when the phantom was at the highest table height position (as a combined effect of ATVS and TCM functions, respectively). Conversely, in the case of AP scout, the CTDI vol values were the greatest when aligning the phantom at the highest table position and lowest when the phantom was centered at the lowest table position. When the lateral scout was used for ATVS and TCM, the changes in CTDI vol values were more subtle than when using AP or PA scouts. Fig. 4 shows the relative doses at each table height position and scout direction compared to reference table positions in pulmonary embolism protocol utilizing ATVS (A), and routine chest CT protocol with fixed 120 kVp tube voltage (B). Notable increase in CTDI vol was observed whenever the tube voltage was increased in pulmonary embolism scans. Similar steps in CTDI vol values were also seen in CT scans Fig. 1. A 0.6 mm thick axial CT image (left) and a photograph (right) of the chest phantom. The phantom represented an average human torso in proportion, density, and two dimensional structures, and was constructed of three specific epoxy materials simulating lung, soft, and bone tissues. Black crosses on the left indicate the scan isocenter locations at the five studied vertical levels. These (from the bottom to the top) are referred in the article as: +6 cm (i.e. phantom is positioned too high), +3 cm, 0 cm, -3 cm, and -6 cm (i.e. phantom is positioned too low). performed with the non-contrast and contrast-enhanced chest CT scan protocols using ATVS. The highest change in relative dose (91%) was observed with the GE scanner when performing a CT scan for pulmonary embolism after PA scout and when the phantom was set 6 cm below the scanner isocenter. This was related to the larger magnification of the spine structure with higher attenuation compared to softtissue and lung areas, thereby contributing more to the automatically adjusted dose level. In the scans performed with the fixed 120 kVp tube voltage, the highest change in relative dose (50%) was seen in the GE system after PA scout and when the phantom was positioned 6 cm below the scan isocenter, whereas the largest change with Siemens system was 18%. Overall, the CTDI vol alterations were higher with the GE scanner than the Siemens system, probably due to the differences in beam shaping filters of the scanners and differences in the TCM systems. GE noise index model involves stronger variability in tube currents pursuing to constant noise statistics in the images, whereas Siemens CARE Dose 4D has smoother response according to patient attenuation [31]. Image analysis Figs. 5 and 6 show the image noises in different phantom regions when using pulmonary embolism protocols with ATVS and routine chest CT protocols with fixed 120 kVp tube voltage, respectively. The image noise values measured in the chest CT protocols with and without the defined contrast administration, both using ATVS, are given in the Supplementary materials Figs. A.1 and A.2, respectively. Clear steps in the image |
noise can be seen in the pulmonary embolism scans (Fig. 5) whenever the tube voltage was changed. Lower tube voltages yielded higher image noise with both scanners. Image noise was higher in bone tissue (ROI 1) compared to soft tissues (ROIs 2-6). Moreover, the vertical centering had greater impact on the noise in the peripheral regions compared to the central region (ROIs 3-4). Beam hardening and streak artefacts contributed to additional local and structural noise component which caused the overall image noise to vary more in ROI 2 than in ROI 3. This can be seen in wider error bars in the Figs. 5 and 6. Noise behavior differences between the scanners, especially in the posterior parts of the phantom, are most probably caused by bowtie filter and TCM method differences. Siemens is using a real-time TCM in addition to scout-based TCM while GE system is only using the last scout image for TCM. Additionally, Siemens system allows higher image noise levels for obese patients and lower noise levels for small patients. GE system, on the other hand, tries to deliver the same noise level, regardless of the patient body size. Figs. 7 and 8 show the difference maps for the GE and Siemens CT scanners, respectively. As the noise difference maps show, with the fixed 120-kVp protocol and AP scout direction, image noise on the posterior side was the greatest when the phantom was positioned 6 cm below the isocenter, and the smallest when the phantom was on the highest table position. On the contrary, image noise on the anterior side was the greatest when the phantom was positioned 6 cm above the isocenter of the CT scanners, and the smallest with the lowest table position. The image noise changes in the GE scanner were higher than with the Siemens scanner for the studied geometry. The PA and LAT scout directions resulted in similar noise behavior as the AP scout direction. Phantom's vertical off-centering affected not only the image noise but also the image contrast. Figs. 7 and 8 show that the relative CT numbers altered the most in bone tissue but also to some extent in the soft and lung tissues when the phantom was vertically off-centered. There were visible differences between the vendors in the studied geometry. For example, with +6 cm centering, the CT numbers in the phantom's anterior side deviated more on the GE system compared to the Siemens system, whereas the overall variations in the bone CT numbers were smaller in the GE system. 7. Axial image data from the GE chest CT scan with a fixed 120 kVp tube voltage after an AP scout. The reference axial CT image with mean CT numbers (A) and corresponding 1SD noise map (B) were calculated from ten slices comprising a total z-direction volume coverage of 10 cm. Absolute CT number differences (C) and relative noise difference maps (D) at table heights +6, +3, -3, and -6 cm were calculated relative to the reference centering (0 cm). Discussion Several studies have shown that patient off-centering is a common and serious problem in CT with detrimental effects on patient dose and image quality [20][21][22][23][24][25][26][27][28][29][30][31][32][33][34][35][36]. According to these publications, patients are typically positioned 1.7 cm-3.5 cm below the scan isocenter. The previous patient positioning studies have mainly investigated the function of TCM and only in a single limited study has the effect of ATVS on radiation dose been investigated [30]. No study to date has systematically assessed the effect of patient off-centering on the function of ATVS with different scout directions and different CT vendors. Due to divergent x-ray fan beam geometry with relatively short focus-to-patient distance in CT, patients are projected larger or smaller to the detector surface depending of the patient alignment in the focusdetector axis (off-centering). This has an effect not only on the selected tube current in the TCM but also the function of ATVS because both techniques utilize scout images in the calculation of the patient's net attenuation. Body regions with higher attenuation or patients with larger cross-sectional size are then scanned with using higher tube currents or tube voltages compared to the body regions with lower xray attenuation or patients with smaller cross-sectional size. As Fig. 3 showed, a vertical off-centering of the phantom causes geometric magnification or minification in the acquired scout images, especially in the AP or PA directions. Similarly, lateral off-centering, if present, would cause geometric magnification or minification for the lateral scout images. The results of the current phantom study revealed clear differences in CTDI vol values and image noise levels when the phantom was scanned in off-centered vertical positions compared to properly positioned scans on the scan isocenter. There were also differences observed between the manufacturers in the function of ATVS. The GE's ATVS worked with less voltage variance than the Siemens' ATVS technique in the studied geometry. The Siemens system tended to select higher tube voltage after the AP and PA scouts compared to the lateral scout, whereas in the GE CT scanner, either higher or the same tube voltage was selected after the lateral scout compared to the scans performed after AP or PA scouts. The difference was most evident in the pulmonary embolism scans. In practice, if ATVS behaves inconsistently or systematically results in suboptimal noise levels, the tube voltage range, patient centering practices, or protocol scout directions might have to be reconsidered. However, if the scout direction is changed, the noise indices or quality reference mAs levels may have to be readjusted to assure consistent image quality according to indication. The scanner reported radiation dose, measured in CTDI vol , increased substantially in the highest or lowest table positions compared to the reference table height. Furthermore, the dose absorbed by sensitive surface tissues such as the breast and thyroid gland will also increase if the patient in supine position is positioned too low and the PA scout direction is used for the TCM and ATVS. These anterior sensitive tissues are then located closer to the scan isocenter and thereby projected on the thinnest and least attenuating part of the bowtie filter during the scan rotation [25,32]. Image analysis results of different ROIs (see e.g. Fig. 6 for the fixed 120 kVp scans) demonstrated distinct noise trends. Higher observed noise values of the posterior ROIs 1-3 while the phantom was in the lower positions (centered 6 cm and 3 cm below scan isocenter) were due to a lower photon flux on these locations and being farther away from the scan isocenter. In this respect, the lower photon flux is partly due to the bowtie filter being thinner at the center, allowing more xrays through, and thicker at the edge, attenuating more x-rays in these parts of the fan beam. The same effect of higher noise values was seen on the anterior ROIs (ROIs 5 and 6) while the phantom was in the Fig. 8. Axial image data from the Siemens chest CT scan with a fixed 120 kVp tube voltage after an AP scout. The reference axial CT image with mean CT numbers (A) and corresponding 1SD noise map (B) were calculated from ten slices comprising a total z-direction volume coverage of 10 cm. Absolute CT number differences (C) and relative noise difference maps (D) at table heights +6, +3, -3, and -6 cm were calculated relative to the reference centering (0 cm). higher positions (centered 6 cm and 3 cm above scan isocenter). The effect depends also on the automatic TCM and scout direction. There was a trend of lower noise levels with increasing centering heights with AP scout, and vice versa with the PA scout, with the GE scanner and most clearly seen in the central ROI 4. This was mostly due to the scout direction-based minification or magnification of the phantom structures, especially the spine, which influence accordingly to the TCM. Therefore, the phantom attenuation appears larger in the PA scout when the phantom is deviated into lower position and closer to the xray tube, causing TCM to use higher tube currents and leading to decreased noise levels in the CT images, respectively. This effect was not seen the same way in the Siemens scanner, probably due to a different TCM technique which utilizes scout image in longitudinal modulation and on-line projection data in angular modulation during the helical CT scan. Therefore, the significance of the scout direction on noise levels was not as pronounced with the Siemens system. Additionally, the effect of scout direction on noise levels was not as pronounced with the PA scout for posterior ROIs, and vice versa with the AP scout direction for anterior ROIs. This was due to the combined effect of automatic TCM, ROI location with respect to the scan, and bowtie differences (Siemens' bowtie filter being flatter than in GE) between the two vendors. Patient centering affected not only the radiation dose and image noise, but also image contrast. A change in tube voltage affects the x-ray spectra, and thus the x-ray attenuation-based CT-numbers. With lower photon energies, photoelectric effect is more prominent and less Compton scattering infers the detected signal, which is a clear benefit especially in CT angiography and contrast-enhanced CT imaging where iodine k-edge plays an important role in the contrast formation. Thus, as a result of using lower tube voltages, higher contrast between the iodine and soft tissue can be achieved [40]. The change in image contrast is also affected by the bowtie filters of the CT systems which modify the x-ray spectra assuming the patient as a cylindrical attenuation target for compensation. With patient off-centering, the bowtie compensating effects which assume isocenter positioning are misplaced with regards to the actual off-centered patient attenuation. Therefore, there will also be corresponding deviations in the reconstructed image noise and contrast (based on the relative x-ray attenuation) distribution across the axial scan plane direction [25,33]. In addition to the effects caused by vendor-specific TCM function and bowtie geometry, the observed image quality and radiation dose differences between the vendors may also be related to the slight differences in the focus-to-isocenter distances between the two scanner models (Siemens 595 mm and GE 541 mm). This also supports the flatter contrast and noise difference maps observed in the Siemens images. The results of this study are consistent with the findings of Filev et al. [30] who found that the function of ATVS and TCM depend on patient centering. They observed that geometric magnification on PA scout images might result in higher tube voltages and tube current values, and thus also higher radiation doses if ATVS was used. However, they only used a Siemens CT system and PA scout direction whereas the current study included a GE CT system as well as AP and LAT scouts. Moreover, the results with the fixed 120 kVp tube voltage support the dosimetric and image noise findings of previous studies [20][21][22][23][24][25][26][27]29,[31][32][33]35]. When using the AP scout for TCM and positioning the phantom 6 cm above the isocenter, 18% and 44% increase in CTDI vol values were observed in the Siemens and GE CT systems, respectively. The PA scout resulted in comparable dose increases (around 13% and 50%) when the phantom was centered 6 cm below the scan isocenter. Habibzadeh et al. [24] reported dose penalties of 13%, 33%, and 51% with 2 cm, 4 cm, and 6 cm vertical position errors for a phantom, respectively. Toth et al. [21] observed that off-centering the phantom by 3 cm and 6 cm increased the surface dose of a cylindrical 32 cm body CTDI phantom by 18% and 41% and the image noise by 6% and 22%, respectively. Kaasalainen et al. [25] found a 16% increase in the absorbed dose of breast tissue when a five-year-old pediatric phantom was scanned with the fixed mAs values and positioned 6 cm below the scan isocenter. Saltybaeva and Alkadhi [32] reported up to 38% increased thyroid dose when off-centering the anthropomorphic phantom by 5 cm. Moreover, Kaasalainen et al. [27] reported 38%, 21%, and 12% increased doses for the adult, five-year-old pediatric, and newborn anthropomorphic phantoms, respectively, when positioning the phantoms 6 cm below the isocenter and using PA scout for TCM in GE CT system. The present study has certain limitations. Firstly, the study was performed using only one anthropomorphic chest phantom. For |
more extensive study, the effect of patient's vertical off-centering should also be investigated on varying chest anatomy and with other anatomical body regions as well. Secondly, only two CT scanners from two vendors were studied, and thereby, the results may not be valid for other ATVS implementations. Thirdly, only CTDI vol values were used to measure radiation doses, and further investigations to measure organ doses, for example, with MOSFET dosimeters and Monte Carlo simulations could be performed. Conclusions The scout direction and vertically off-centering the patient have complex and mixed effects on the ATVS and TCM operation, propagating further on the measured radiation dose and image quality. This emphasizes the importance of proper centering of the patient with the modern CT scanner models with new optimization tools. The greatest variation in the selected tube voltage by ATVS was seen in the pulmonary embolism scans. Furthermore, there were notable differences in the ATVS behavior between the CT vendors for the studied geometry. This emphasizes the importance of adequate in-depth knowledge of the users on the functionality of their scanner model and how optimization tools, centering and scan parameters affect both radiation dose and image quality. Declaration of interest None. Confilict of interests The authors do not have any competing interests to declare. Funding This study was supported by the State Subsidy for University Hospitals in Finland and a research grant from the Radiological Society of Finland. Dice-XMBD: Deep Learning-Based Cell Segmentation for Imaging Mass Cytometry Highly multiplexed imaging technology is a powerful tool to facilitate understanding the composition and interactions of cells in tumor microenvironments at subcellular resolution, which is crucial for both basic research and clinical applications. Imaging mass cytometry (IMC), a multiplex imaging method recently introduced, can measure up to 100 markers simultaneously in one tissue section by using a high-resolution laser with a mass cytometer. However, due to its high resolution and large number of channels, how to process and interpret the image data from IMC remains a key challenge to its further applications. Accurate and reliable single cell segmentation is the first and a critical step to process IMC image data. Unfortunately, existing segmentation pipelines either produce inaccurate cell segmentation results or require manual annotation, which is very time consuming. Here, we developed Dice-XMBD1, a Deep learnIng-based Cell sEgmentation algorithm for tissue multiplexed imaging data. In comparison with other state-of-the-art cell segmentation methods currently used for IMC images, Dice-XMBD generates more accurate single cell masks efficiently on IMC images produced with different nuclear, membrane, and cytoplasm markers. All codes and datasets are available at https://github.com/xmuyulab/Dice-XMBD. INTRODUCTION Analysis of the heterogeneity of cells is critical to discover the complexity and factuality of life system. Recently, single-cell sequencing technologies have been increasingly used in the research of developmental physiology and disease (Stubbington et al., 2017;Papalexi and Satija, 2018;Potter, 2018;Lähnemann et al., 2020), but the spatial context of individual cells in the tissue is lost due to tissue dissociation in these technologies. On the other hand, traditional immunohistochemistry (IHC) and immunofluorescence (IF) preserve spatial context but the number of biomarkers is limited. The development of multiplex IHC/IF (mIHC/mIF) technologies has enabled the simultaneous detection of multiple biomarkers and preserves spatial information, such as cyclic IHC/IF and metal-based multiplex imaging technologies (Zrazhevskiy and Gao, 2013;Angelo et al., 2014;Giesen et al., 2014;Tan et al., 2020). Imaging mass cytometry (IMC) (Giesen et al., 2014;Chang et al., 2017), one of metal-based mIHC technologies, uses a high-resolution laser with a mass cytometer and makes the measurement of 100 markers possible. IMC has been utilized in studies of cancer and autoimmune disorders (Giesen et al., 2014;Damond et al., 2019;Ramaglia et al., 2019;Wang et al., 2019;Böttcher et al., 2020). Due to its high resolution and large number of concurrent marker channels available, IMC has been proven to be highly effective in identifying the complex cell phenotypes and interactions coupled with spatial locations. Thus, it has become a powerful tool to study tumor microenvironments and discover the underlying disease-relevant mechanisms (Brähler et al., 2018;Ali et al., 2020;Aoki et al., 2020;de Vries et al., 2020;Dey et al., 2020;Jackson et al., 2020;Zhang et al., 2020;Schwabenland et al., 2021). Apart from using IMC techniques alone, several other technologies, such as RNA detection in situ and 3D imaging, have been combined with IMC to expand its applicability and utility (Schulz et al., 2018;Bouzekri et al., 2019;Catena et al., 2020;Flint et al., 2020). The IMC data analysis pipeline typically starts with single cell segmentation followed by tissue/cell type identification (Carpenter et al., 2006;Sommer et al., 2011;Liu et al., 2019). As the first step of an IMC data processing pipeline, the accuracy of single cell segmentation plays a significant role in determining the quality and the reliability of the biological results from an IMC study. Existing IMC cell segmentation methods include both unsupervised and supervised algorithms. Unsupervised cell segmentation, such as the watershed algorithm implemented in CellProfiler (Carpenter et al., 2006), does not require user inputs for model training. However, the segmentation results are not precise in particular when cells are packed closely or they are in complicated shapes. To achieve better segmentation results, supervised methods use a set of images annotated with pixellevel cell masks to train a segmentation classifier. However, the manual annotation task is very time consuming and expensive as well since it is normally done by pathologists or experienced staff with necessary knowledge in cell annotation. Particularly, for multiplexing cellular imaging methods such as IMC, their channel configurations including the total number of markers and markers selection are typically study dependent. Therefore, manual annotation may need to be performed repeatedly for each study to adapt the segmentation model to different IMC channel configurations, which can be impractical. To overcome this limitation, a hybrid workflow combining unsupervised and supervised learning methods for cell segmentation was proposed . This hybrid workflow uses Ilastik (Sommer et al., 2011), an interactive image processing tool, to generate a probability map based on multiple rounds of user inputs and adjustments. In each round, a user only needs to perform a limited number of annotations on regions where the probability map generated based on previous annotations is not satisfactory. CellProfiler is then used to perform the single cell segmentation based on the probability map once the result from Ilastik is acceptable. This hybrid workflow significantly reduces manual annotation workload and has gained popularity in many recent IMC studies (Damond et al., 2019;Böttcher et al., 2020;de Vries et al., 2020;Jackson et al., 2020;Schwabenland et al., 2021). However, the annotation process still needs to be performed by experienced staff repeatedly for each IMC study, which is very inconvenient. In addition, the reproducibility of the experimental results obtained from this approach can be an issue due to the per-study, interactive training process used in creating the single cell masks. Hence, a more efficient, fully automated single cell segmentation method for IMC data without compromising the segmentation accuracy is necessary for IMC to gain broader applications in biomedical studies. Convolutional neural networks (CNNs) have been successfully used for natural image segmentation and recently applied in biomedical image applications (Shen et al., 2017;Zhang et al., 2018;Andrade et al., 2019;Vicar et al., 2019). CNNbased U-Net was developed for pixel-wise cell segmentation of mammalian cells (Ronneberger et al., 2015). It has been demonstrated that the U-Net architecture and its variants such as Unet++ (Zhou et al., 2018), 3D Unet (Çiçek et al., 2016), and V-Net (Milletari et al., 2016) can obtain high segmentation accuracy. Motivated by the good performance of U-Nets in cell segmentation (Van Valen et al., 2016;Hollandi et al., 2020;Salem et al., 2020), we developed Dice-XMBD, a deep neural network (DNN)-based cell segmentation method for multichannel IMC images. Dice-XMBD is marker agnostic and can perform cell segmentation for IMC images of different channel configurations without modification. To achieve this goal, Dice-XMBD first merges multiple-channel IMC images into two channels, namely, a nuclear channel containing proteins originated from cell nucleus, and a cell channel containing proteins originated from cytoplasm and cell membrane. Channels of proteins with ambiguous locations are ignored by Dice-XMBD for segmentation as they contribute little to the segmentation results. Furthermore, to mitigate the annotation workload, we adopted the knowledge distillation learning framework (Hinton et al., 2015) in training Dice-XMBD, where the training labels were generated using Ilastik with interactive manual annotations as a teacher model. We used four IMC datasets of different channel configurations to evaluate the performance of Dice-XMBD and the results show that it can generate highly accurate cell segmentation results that are comparable to those from manual annotation for IMC images from both the same and different datasets to the training dataset, validating its applicability for generic IMC image segmentation tasks. Overview of the Pipeline In Dice-XMBD, we used a U-Net-based pixel classification model to classify individual pixels of an IMC image to their cellular origins, namely, nuclei, cytoplasm/membrane, or background. The classification model outputs pixel-level probability values for each class, which were then input to CellProfiler (version 3.1.0) to produce the final cell segmentation masks (Figure 1). The ground truth cell segmentation of IMC images is in general not available. To obtain the training labels, we generated pixel probability maps using an iterative manual annotation process with Ilastik on the training IMC dataset. Furthermore, the same iterative manual annotation process was performed on the testing IMC datasets to produce the ground FIGURE 1 | Dice-XMBD workflow. Imaging mass cytometry (IMC) images are combined into 2-channel images containing nuclear and membrane/cytoplasm proteins expression information. In stage 1, the pixel probability maps of the input 2-channel images are predicted using a semi-supervised learning model based on U-Net architecture. The training data were generated from Ilastik by an iterative interactive annotation process. In stage 2, the cell segmentation masks are generated from the pixel probability maps using the propagation method in CellProfiler. truth pixel probability maps, which were used by CellProfiler to produce the ground truth cell segmentation masks for performance evaluation. Note that to obtain a generic pixel classifier that can be used across IMC datasets of different channel configurations, channels of different proteins were combined based on their cellular origins into two channels, namely, nuclear and cell (membrane/cytoplasmic) channels. Channels of proteins without specific cellular locations were ignored by Dice-XMBD. The pixel classification model was trained using the combined two-channel images as input. Likewise, the same preprocessing was used at the prediction stage to produce the two-channel (nuclear/cell) images as input to the pixel classification model. Of note, although the prediction may be performed on images with different markers, the channels were always combined based on their origins so that pixel classification was performed based on the two channels of putative protein locations rather than channels of individual proteins. Training and Evaluation Datasets We used four IMC image datasets in this study. BRCA1 and BRCA2 contain 839 and 754 images from patients with type I diabetes with 34 markers. Dice-XMBD was trained on a subset of BRCA1 dataset (n = 348) with 200 held-out images reserved for validation and testing. To test the generalization ability of Dice-XMBD, we also tested the trained model on the other three independent IMC datasets (BRCA2, T1D1, and T1D2). Generating Ground Truth Cell Masks The ground truth pixel probability maps and the cell masks used for model training and evaluation were generated using Ilastik and CellProfiler. We used the smallest brush size (1 pixel) in annotating the image to avoid annotating a group of neighboring pixels of different classes. To mitigate the manual workload, the annotation was performed in an interactive manner, where the random forest prediction model of Ilastik was updated regularly during annotation to produce an uncertainty map indicating the confidence level of the classification results produced by the prediction model. The annotation was then guided by the uncertainty map to focus on the regions with high uncertainty iteratively, until the overall uncertainty values were low except for regions of which the boundaries were visually indistinguishable. The initial annotation was performed on a randomly selected subset of the dataset. After the initial annotation, we loaded all the images from the dataset into Ilastik to calculate their uncertainty maps, and then selected those with the highest average uncertainty values for further annotation. This process was iterated until the uncertainty values of all images converged, that is, the average uncertainty |
value over all images did not decrease significantly for three consecutive iterations. In the end, we annotated 49 images in BRCA1 to train the model in Ilastik. We then imported all the images of the BRCA1 dataset into Ilastik for batch processing and export their corresponding pixel classification probability maps for training Dice-XMBD. The probability maps were further input to CellProfiler to produce the ground truth cell segmentation. In CellProfiler, we used the "IdentifyPrimaryObjects" module to segment the cell nuclei and used the "IdentifySecondaryObjects" to segment the cell membranes using the propagation method. The output masks from CellProfiler are regarded as ground truth cell segmentation of the dataset for performance evaluation. We also generated the ground truth cell masks of the other three datasets by the same iterative procedure separately for testing the generalization ability of Dice-XMBD. During the process, 72 images in BRCA2, 39 images in T1D1, and 67 images in T1D2 were manually annotated. Image Preprocessing The multiplexed IMC images were first merged into two channels by averaging the per-pixel values from the selected membrane and nuclear channels. After merging channels, the input IMC images were then preprocessed by hot pixel removal, dynamic range conversion, normalization, and image cropping/padding into fix-sized patches. First, we applied a 5 × 5 low-pass filter on the image to remove hot pixels. If the difference between an image pixel value and the corresponding filtered value was larger than a preset threshold (50 in our experiments), the pixel would be regarded as a hot pixel and its value would be replaced by the filtered value. As the dynamic range of pixels values differs among IMC images of different batches and different channels, we further min-max normalized all images to [0,255] to remove such batch effect as: where x ij denotes the pixel value in one channel, and X max and X min denote the maximum and minimum values in the channel. Of note, as the pixel values in IMC images have a high dynamic range, transforming the pixel values from its dynamic range to [0, 255] would suffer from detail suppression by one or few extremely large values. Therefore, we thresholded the image pixel values at 99.7% percentile for each image before normalization. Finally, we merged all the nuclear channels into one consolidated nuclear channel, and membrane/cytoplasmic channels into one cell channel, by averaging on all channel images with pre-selected sets of protein markers, respectively. We converted the merged two-channel images into patches of 512 × 512 pixels. Image boundary patches that are smaller than the target patch size are padded to target size. For the padded pixels, we set the pixel values of both channels to 0 and the pixel type as background. Data Augmentation Data augmentation is an effective strategy to reduce overfitting and enhance the robustness of the trained models, especially when training data are insufficient. We applied the following data augmentation methods on the input images before feeding to our U-Net-based pixel classification network. First, photometric transformations including contrast stretching and intensity adjustments were used. For contrast stretching, we changed the level of contrast by multiplication with a factor randomly drawn from the range of [0.5, 1.5]. Similarly, for intensity adjustments we changed the level of intensities by multiplication with a factor randomly drawn from the range of [0.5, 1.5]. Geometric transformations including image flipping and rotation were used. For flipping, we implemented random horizontal or vertical flipping. For rotation, the rotating angle is randomly distributed in the range of [−180, 180]. Note that geometric transformations were applied to pairs of input and output images of the network. We also injected random Gaussian noise to the two input channels of the input images. Examples of data augmentation are shown in Supplementary Figures 1, 2. Constructing a Pixel Classification Model The U-Net pixel classification network is an end-to-end fully convolutional network and contains two paths. The contracting path (or the encoder) uses a typical CNN architecture. Each block in the contracting path consists of two successive 3 × 3 convolution layers followed by a Rectified Linear Unit (ReLU) activation and a 2 × 2 max-pooling layer. This block is repeated four times. In the symmetric expansive path (or the decoder), at each stage the feature map is upsampled using 2 × 2 upconvolution. To enable precise localization, the feature map from the corresponding layer in the contracting path is cropped and concatenated onto the upsampled feature map, followed by two successive 3 × 3 convolutions and ReLU activation. At the final stage, an additional 1 × 1 convolution is applied to reduce the feature map to the required number of output channels. Three output channels are used in our case for nuclei, membrane, and background, respectively. As we output the probability map, the values are converted into the range of [0, 1] using the Sigmoid function. Loss Function We take the binary cross-entropy (BCE) as the loss function, which is defined as: where N represents the total number of pixels in an image, y i denotes the ground truth pixel probability, andŷ i denotes the predicted pixel probability. The cross-entropy loss compares the predicted probabilities with the ground truth values. The loss is minimized during the training process. Model Evaluation In a binary cell mask, "1" represents cell boundary and "0" denotes cell interior or exterior. For every pixel in an image, true positive (TP) and true negative (TN) mean that the predicted pixel classification is the same as its label in the labeled (i.e., the ground truth) mask, while false positive (FP) and false negative (FN) mean that a pixel is misclassified. To evaluate pixel-level accuracy, we calculated the number of TP pixels and FP pixels based on the predicted and labeled binary masks. We further evaluated model performance at the cell level. We calculated the intersection over union (IOU) on cells from predicted and labeled cell masks to determine if they are the same cell, and then counted the TP and FP cells. First, we filtered out all cells with IOU below 0.1 from the predicted cells. These cells are identified as FPs. The other cells from the predicted cell mask could be either TP or FP. If a predicted cell only overlaps with one true cell (i.e., a cell from the labeled cell mask), we assume that the cell is segmented accurately (TP). If a true cell cannot find a predicted cell, the "missing" cell is denoted as FN. When multiple predicted cells are assigned to the same true cell, we consider this as a split error. If multiple true cells are matched to the same predicted cell, we consider those predicted cells as merge errors. For simplicity, split errors and merge errors are counted as FPs. Four standard indices are measured as follows: To investigate the effect of different segmentation methods on downstream analysis, an unsupervised clustering method (Phenograph Levine et al., 2015, Python package, v1.5.7) was applied to the high-dimensional single cell expression data processed from each different method under comparison, and the labeled ground truth cell mask, separately. Prior to clustering, single cell protein expressions were quantified by the mean pixel values, and then these values were censored at 99th percentile and transformed with arcsinh function. Scaled high-dimensional single cells were clustered into several groups based on selected markers as from the original publication of each individual dataset. Based on the expressions of cell-specific markers, the cell types of the clusters were identified among T cells (CD3), CD4 T cell (CD4), CD8 T cell (CD8a), B cell (CD20), macrophage (CD68), endothelial cell (CD31), and so on. By comparing the cell annotation from different segmentation methods (predicted cell mask) and the labeled cell mask, the cell annotation accuracy was calculated as n same /N total . Here, n same is the number of correctly predicted cells, which are cells that correctly overlapped with the corresponding cells in the labeled mask (i.e., TP cells), and annotated to the same cell types, N total is the total number of cells from the predicted mask. Dice-XMBD Enables Automatic Cell Segmentation We trained our U-Net cell segmentation model using the BRCA1 dataset with 348 images as the training set and 100 images as the validation set. A complete held out test set with 100 images was used to test model performance within one dataset. We further applied the trained model directly on the other three IMC image datasets to evaluate the cross-dataset performance of the model. For performance evaluation, we computed standard indices (Recall, Precision, F1-score, and Jaccard index) for both pixel-level and cell-level accuracies (see section 2). We compared Dice-XMBD with a generic whole-cell segmentation method across six imaging platforms, Mesmer (Greenwald et al., 2021), which used a deep learning-based algorithm trained on a large, annotated image dataset to segment single cells and nuclei separately. A trained Mesmer model was tested with combined nuclear and cell channels, which is the same as the input to Dice-XMBD. Meanwhile, we compared with three commonly used segmentation methods implemented in CellProfiler with default parameters: distance, watershed, and propagation. These methods first locate nuclei as primary objects, and then the membrane proteins are added together into an image as input to recognize cells. The distance method does not use any membrane proteins information and simply defines cell membrane by expanding several pixels around nuclei. The watershed method computes intensity gradients on the Sobel transformed image to identify boundaries between cells (Vincent and Soille, 1991), while the propagation method defines cell boundaries by combining the distance of the nearest primary object and the intensity gradients of cell membrane image (Jones et al., 2005). Hereafter, we refer to these three CellProfile-based methods as CP_distance, CP_watershed, and CP_propagation, respectively. Results show that Dice-XMBD outperformed all other benchmarked methods with highest accuracy on pixel level (F1 score = 0.92, Jaccard index = 0.85) (Figure 2A). We also observed that CP_distance obtained the highest recall (Recall = 0.95) but lowest precision (Precision = 0.66), which means that it can identify almost every pixel correctly in the labeled mask but only 66% of predicted pixels were accurate. In terms of cell-level performance, we first counted cells per image from predicted and labeled cell masks. The prediction result from Dice-XMBD showed highest correlation with the ground truth (Pearson correlation = 0.998) among all methods tested. Mesmer (Pearson correlation = 0.955) and CellProfiler (Pearson correlation = 0.981) also achieved high correlation with the ground truth. However, Mesmer tended to predict less cells while CellProfiler was more likely to over-split cells, as shown in Figures 2B,C. Moreover, Figure 2C shows that Dice-XMBD had the best prediction performance (F1-score = 0.867) considering precision (Precision = 0.856, percent of cells that were correctly predicted) and recall (Recall = 0.880, percent of true cells that are predicted) than Mesmer (F1-score = 0.557) and CellProfiler (F1score = 0.567, 0.563, and 0.561 for CP_distance, CP_watershed, and CP_propagation, respectively). We further checked the IOU distribution of all one-to-one cell pairs (predicted and true cells), Figure 2D demonstrates that most matched cell pairs predicted from Dice-XMBD were highly overlapping (mean = 0.815, median = 0.821), followed by Mesmer where most matched pairs are only half area of overlap (mean = 0.579, median = 0.595). An example of BRCA1 shown in Figure 2E demonstrates that Dice-XMBD prediction was far superior to other benchmarked methods since it contained most cells with high matched values. Dice-XMBD Enables Generic IMC Image Segmentation The key idea of this study was to generate an IMC-specific single cell segmentation model across different datasets with multiple proteins. We selected three independent IMC datasets generated from different labs to test the generalization ability of Dice-XMBD. Apart from the benchmarked methods mentioned above, we also included the Ilastik model trained from BRCA1 annotations in our comparison. Figure 3A shows that Dice-XMBD outperformed all the other methods, followed by Ilastik. Moreover, the performance of cells prediction from Dice-XMBD was the best and the most stable for all three datasets, while Ilastik and Mesmer tended to under-predict cells. CellProfiler predicted less cells in BRCA2 and over-predicted cells in two T1D datasets, as shown in Figures 3B,C. Furthermore, Dice-XMBD predictions contained most of the cells with IOU value higher than 0.8 ( Figure 3D and Supplementary Figure 3). Dice-XMBD Enables Accurate Downstream Biological Analysis To investigate the influence of |
segmentation accuracy on downstream analysis, we clustered single cells resulting from different segmentation methods separately using Phenograph and compared the clustering results. Taking the result from single cells obtained from Dice-XMBD segmentation on BRCA1 dataset as an example, these cells can be clustered into 26 distinct clusters [ Figure 4A, t-distributed stochastic neighbor embedding (t-SNE) visualization in Figure 4B]. Based on the scaled mean expression for each cluster, we were able to annotate Cluster 3 as T cells, Cluster 18 as B cells, Cluster 16 as macrophage, and the remaining clusters to other cell types which may include tumor cells, stromal cells, or endothelial cells (Figure 4C). We performed the same clustering and annotation process on single cells obtained from other segmentation methods and the ground truth segmentation on all three datasets separately as well. For two T1D datasets [T1D1 (Damond et al., 2019) and T1D2 (Wang et al., 2019)], CD4 T cells, CD8 T cells, and CD31+ endothelial cells were identified based on their selected markers. We compared the concordance of cell fractions based on annotations from different segmentation methods (prediction) versus those from ground truth segmentation (ground truth) ( Figure 4D and Supplementary Figures 4A-7A). On BRCA1 dataset, Dice-XMBD performed better compared with all other segmentation methods on overall results and results of certain cell types ( Figure 4D). Significantly, two CellProfiler-based methods (CP_watershed, R 2 = 0.85 and CP_propagation, R 2 = 0.85) showed inferior performance in reproducing cell fraction results in macrophage while Dice-XMBD still achieved an R 2 = 0.99 in this cell types. CP_distance delivered reasonable performance in macrophage, but was still inferior to Dice-XMBD on T cell. Similar results can be observed on other datasets as well. For example, for the T1D1 dataset, CD4 T cells were poorly predicted by Ilastik (R 2 = 0.043) and CP_distance (R 2 = 0.055) (Supplementary Figure 6A). For the T1D2 dataset, endothelial cells were poorly predicted by Ilastik (R 2 = 0.58) and macrophage cells were poorly predicted by Mesmer (R 2 = 0.033). On the other hand, Dice-XMBD delivered highly consistent prediction results across all cell types in all datasets except for T cell in BRCA2 dataset, where all methods did not perform well. In addition to cell fraction, we also evaluated the annotation accuracy of individual cells for each method (Figure 4E and Supplementary Figures 4B-7B), which is important for spatially related analysis of single cell data such as neighborhood analysis. Dice-XMBD achieved the highest cell annotation accuracies among all segmentation methods on overall results (Figure 4E), and performed as well as or better than other methods on all individual cell types in all datasets (Supplementary Figures 4B-7B). Generalization Ability of Dice-XMBD To investigate the impact of the training data on the segmentation performance of Dice-XMBD, we trained Dice-XMBD using different training datasets, and evaluated the performance of the resulting models on other IMC datasets used in this study. Results show that segmentation performance in terms of pixel-level accuracy were in fact very similar among these models (Supplementary Tables 1-4). We further asked if the performance of Dice-XMBD could be improved by training on multiple datasets. Interestingly, the model did not consistently perform better when more than one datasets were combined as the training set (Supplementary Tables 1-4). All together, these results suggest that by using location specific channels, Dice-XMBD were highly robust to different training datasets, and a Dice-XMBD model trained on one dataset can be well generalized to segmentation tasks on other IMC datasets. Of note, in our approach, the channels of same locations were simply averaged without applying any weighting scheme to produce the location specific channels. We tried to minmax-normalize the selected channels before averaging so that all selected channels contributed equally to the combined channels. However, the pixel-level accuracy dropped on all datasets, albeit at different levels of degradation on different datasets (Supplementary Tables 1-4). As different channels may contain different levels of information to the final segmentation results, combining them with equal weights may not be the optimal approach. However, how to find the optimal weighting combination of different channels remains an open question that deserves further exploration. DISCUSSION Highly multiplexed single cell imaging technologies such as IMC are becoming increasingly important tools for both basic biomedical and clinical research. These tools can unveil complex single-cell phenotypes and their spatial context at unprecedented details, providing a solid base for further exploration in cancer, diabetes, and other complex diseases. Nevertheless, cell segmentation has become a major bottleneck in analyzing multiplexed images. Conventional approaches rely on intensities of protein markers to identify different cellular structures such as nuclei, cytoplasm, and membrane. Unfortunately, the intensity values of these markers are strongly cell type-specific and may vary from cells to cells. In addition, the staining also shows variability across images or datasets. As a result, the accuracy and robustness of the segmentation results are far from optimal. On the other hand, high-order visual features including spatial distribution of markers, textures, and gradients are relevant to visually identify subcellular structures by human. However, these features are not considered in conventional methods to improve the cell segmentation results. The DNN-based image segmentation approaches provide an opportunity to leverage high-order visual features at cellular level for better segmentation results. Unfortunately, they require a significant amount of annotation data that are in general difficult to acquire. In addition, the highly variable channel configurations of multiplexed images impose another important obstacle to (C) tSNE map representing single cells colored by cell-type-specific markers expression (CD68 for macrophage, CD45 and CD3 for T cells, CD45 and CD20 for B cells). Single cells on (A-C) were from BRCA1 dataset and segmented by Dice-XMBD. (D) Scatter plots of cell fraction obtained from ground truth (x-axis) and five segmentation methods (y-axis), colored by different cell types identified from BRCA1 dataset. (E) Cell annotation accuracy from Dice-XMBD and other benchmarked methods in four datasets. Pairwise comparisons of Dice-XMBD and other methods: *P < 0.05; ****P < 0.0001; n.s., not significant (Student's t-test). the usability of these methods as most of them lack the ability to adapt to different channel configurations after models are trained. In this study, we develop Dice-XMBD, a generic solution for IMC image segmentation based on U-Net. Dice-XMBD overcomes the limitation of training data scarcity and achieves human-level accuracy by distilling expert knowledge from Ilastik with manual input of human as a teacher model. Moreover, by consolidating multiple channels of different proteins into two cellular structure-aware channels, Dice-XMBD provides an effective off-the-shelf solution for cell segmentation tasks across different studies without retraining that can lead to significant delay in analysis. Importantly, our evaluation results further demonstrate Dice-XMBD's good generalization ability to predict single cells for different IMC image datasets with minimum impact to downstream analysis, suggesting its values as an generic tool for hassle-free large-scale IMC data analysis. Finally, to facilitate the analysis of large amount of IMC data currently being generated around the world, we made Dice-XMBD publicly available as an open-source software on GitHub (https://github. com/xmuyulab/Dice-XMBD). AUTHOR CONTRIBUTIONS WY, LW, RY, and JH discussed the ideas and supervised the study. YQ and YJ implemented and conducted experiments in deep network cell segmentation. XX performed the model evaluation and biological analysis on segmentation results. XX, WY, and RY wrote the manuscript. All authors discussed and commented on the manuscript. FUNDING This study was funded by National Natural Science Foundation of China (grant no. 81788101 to JH). Flagellar Synchronization Is a Simple Alternative to Cell Cycle Synchronization for Ciliary and Flagellar Studies Cilia and flagella are highly conserved antenna-like organelles that found in nearly all mammalian cell types. They perform sensory and motile functions contributing to numerous physiological and developmental processes. Defects in their assembly and function are implicated in a wide range of human diseases ranging from retinal degeneration to cancer. Chlamydomonas reinhardtii is an algal model system for studying mammalian cilium formation and function. Here, we report a simple synchronization method that allows detection of small changes in ciliary length by minimizing variability in the population. We find that this method alters the key relationship between cell size and the amount of protein accumulated for flagellar growth. This provides a rapid alternative to traditional methods of cell synchronization for uncovering novel regulators of cilia. T he unicellular, biflagellate alga Chlamydomonas reinhardtii is extensively used as a model organism for studying fundamental processes such as photosynthesis, cell motility, cell signaling, cell-cell recognition, and regulation of ciliary assemblydisassembly (1). This organism offers many advantages for molecular and biochemical studies of eukaryotic flagella, as their flagellar structure and function are well conserved (2). Chlamydomonas cells can be chemically or mechanically induced to shed their flagella (termed "deflagellation"). After amputation, they can regenerate flagella to predeflagellation lengths rapidly (within 2 h). Flagellar assembly and disassembly are precisely controlled throughout cell cycle progression and cell division (3,4). During cell division, flagella are disassembled naturally. Flagellar resorption starts at the preprophase stage and continues about 30 min prior to mitotic cell division (5). New flagella begin to form in the daughter cell after division (6,7). During the sexual cycle, flagella begin to resorb a few hours after the fusion of gametes, and the process proceeds gradually as in vegetative growth (8). As cell division plays a critical role in flagellar growth and resorption, cultures with a heterogeneous population of cells in different divisional stages differ in flagellar length (F-L). In contrast, synchronous cultures, which contain cells that are in the same growth stage, have a comparatively homogeneous distribution of flagellar lengths. Thus, synchronous cultures provide advantages over nonsynchronous cells for studying cellular morphology and the effects of various chemical or genetic perturbations on flagellar length. A wide range of physical and chemical methods have been applied to achieve synchronization for different cells or tissue types. Synchronization of bacteria can be carried out by single or multiple changes of temperature or light, single or multiple cycles of nutritional starvation, cell cycle inhibitor block, and size selection by filtration or centrifugation (9)(10)(11)(12). Fission yeast can be synchronized either by separating a subpopulation from an asynchronous culture using specialized centrifugation or by selecting cells from a lactose gradient (13). Temperature-sensitive cell cycle mutations or inhibitors are also used to block the cell cycle at different stages of growth, allowing cells to grow synchronously upon withdrawal of the block (14). Common methods for mammalian cell cycle synchronization are inhibition of DNA replication (15) and inhibition of mitotic spindle formation using different chemical inhibitors (16)(17)(18). Nonchemical methods for cell cycle synchronization include amino acid and serum starvation (19). Cells can also be mechanically separated by physical methods such as flow cytometry, mitotic shake-off, and countercurrent centrifugal elutriation (18). Hypoxic shock and hyperthermic shock have been used to synchronize cells of the ciliate Tetrahymena pyriformis (20). Photosynthetic algal cells are typically exposed to alternative light/dark (L-D) cycles for synchronization (21,22). In Chlamydomonas reinhardtii, as in other photoautotrophic cells, the most common method used for cell synchronization is alternating light/dark cycles (12 h/12 h or 14 h/10 h) in minimal medium (23,24), though other conditions and other methods such as periodic hypothermic conditions (25), selection by size (26), and variable wavelengths of light (27) have been applied. Synchronization can also be achieved by incubating Chlamydomonas reinhardtii cells in low-nitrogen medium for at least 15 h (28). Chlamydomonas cells undergo gametogenesis upon nitrogen deprivation (using nitrogen-free minimal medium [M-N]). After induced gametogenesis, culture contains mostly new-born cells with smaller sizes (29)(30)(31). During L-D synchronization, cells can grow during the light phase to many times their original size (32). In the dark phase, cells can undergo consecutive divisions to produce 2, 4, 8, 16, or even 32 daughter cells depending on the cell size (33). Cells divide in the middle of the dark cycle in Chlamydomonas reinhardtii (23), whereas the division occurs in Chlamydomonas moewusii cells during the late phase of darkness (34). Although cell division is restricted to each dark phase, the starting times of individual cell divisions differ from cell to cell. Thus, consecutive cell divisions take place throughout several hours of the dark period. As a result, the cells are always partially asynchronous in their division at any point in time (8). In addition, cultures are maximally synchronized only after the third iteration of light-dark cycling since some |
populations of the cells divide during the first and second iterations of the light phase (23). Different factors such as light duration and intensity, temperature, and culture density also have an effect on the degree of homogeneity (32,35). For example, in Chlamydomonas eugametos, L-D synchronization can be achieved only if the culture is static without aeration (36). While synchronization of mammalian cells can be optimized, it is not possible to synchronize entire cell populations by any of the methods or techniques described above (37). Traditional L-D synchronization or nitrogen starvation methods can only make partially synchronized Chlamydomonas cultures. As cells are not truly synchronized using these methods, high variabilities of flagellar length are still observed within the population. If the culture contains too much heterogeneity, it can be difficult to detect effects of flagellar length perturbations. As synchronized cells are ideal for assaying length-related flagellar dynamics, here we have outlined and characterized a method in which 100% of cells are synchronized with respect to their flagellar length but are not synchronized with respect to the cell cycle. We tested the utility of this method in evaluating flagellar length after chemical and genetic perturbations. Finally, we probed the basis of flagellar length synchronization by probing the synthesized but unassembled pool of flagellar protein. RESULTS Flagellar length synchronization narrows the steady-state flagellar length distribution. Synchronous culture provides a better way to address cell cycle and related flagellar dynamics. However, cell cycle synchronization methods provide only partial synchronization and thus show high variance in flagellar length. To obtain 100% flagellar length (F-L) synchronization, we exploited an inherent property of Chlamydomonas, which is their ability to regenerate the flagella after amputation (38). We first performed different synchronization methods and then compared their steady-state flagellar lengths to those of nonsynchronized cells (Fig. 1a, Fig. S1 in the supplemental material, and Table S1 in the supplemental material). For F-L synchronization, we tested different regeneration time durations following deflagellation to determine the time point at which flagellar length variability is minimized. We found that the predeflagellation flagellar length distribution was broad, which was expected as our starting culture was nonsynchronous and contained a heterogeneous population of cells (Fig. 1b, red). After deflagellation, all flagella started to grow synchronously but the length distribution still remained broad at 2 and 2.5 h, when some cells were still in the 8-to 9-m size range and did not reach their original length (Fig. 1b, light green). The distribution narrowed and became maximally homogeneous at 3 h (Fig. 1b, medium green) (Table S2). However, the length distribution remained narrow for only a short time, expanding again within 30 min and increasing with time (Fig. 1b, dark Distribution of steady-state flagellar lengths after the use of different synchronization methods. Nonsynchronous cells were used as a control, and steady-state flagellar lengths were measured after the use of each synchronization method as described in the text (n ϭ 100/each). The F test was performed for comparing variance levels, and two-tailed P values were determined. Asterisks indicate significant differences (****, P Յ 0.0001; ***, P Յ 0.001; **, P Յ 0.01). Standard deviations of each distribution are shown below the individually plotted values. (b) Wild-type flagellar length distribution at various time intervals during the regeneration after amputation. Predeflagellation nonsynchronous cells (pre) are shown in red. Regeneration was carried out for the indicated times after deflagellation by pH shock (green). Lighter green and darker green indicate before and after the time of F-L synchronization, respectively (n ϭ 50/each). The F test was performed for comparing variance levels (control ϭ predeflagellation nonsynchronous cells). Bonferroni corrected ␣ altered ϭ 0.007. The asterisks indicate significant difference below ␣ altered (**, P ϭ 0.004). Standard deviations are expressed as bar graphs in the bottom half of the panels. The filled standard deviation bar represents the condition with the smallest variance. Flagellar Synchronization Facilitates Ciliary Studies March/April 2017 Volume 2 Issue 2 e00003-17 msphere.asm.org 3 green). This timing was highly reproducible, and the data shown in Fig. S2 represent the combined results of three independent experiments. On the basis of the standard deviations (SD) of these distributions (Fig. 1b, lower panel) (Fig. S2, lower panel), we selected a regeneration time of 3 h postdeflagellation as the flagellar length synchronization time and the time at which to initiate further experiments. Likewise, after determining conditions of minimal variability for each synchronization method, we measured the steady-state flagellar lengths for comparative analysis. The results revealed that the mean flagellar lengths at steady state were almost equivalent for all methods (Table S1). As expected, nonsynchronous cells had larger variability than all synchronized cells (Fig. 1a, red). The F-L synchronization method shows a remarkably narrow spread of measurements around the mean and the lowest variability of length across all synchronization methods (Fig. 1a, green, and standard deviation bar graph below) (Table S1). In contrast, conventional L-D synchronization and M-N synchronization have comparatively wider distributions (Fig. 1a, blue and purple, respectively) ( Table S1). Our findings suggest that the F-L synchronization is the most effective method for achieving maximum flagellar length homogeneity. Increased effects of chemical perturbations using F-L synchronization. Flagellar length can be perturbed chemically. If the perturbation has a small effect on flagellar length, high variance in the system may mask observed phenotypes. Our observations demonstrate that F-L synchronized cells have reduced variance in flagellar length compared to nonsynchronized cells synchronized using other methods ( Fig. 1a) (Fig. S1). Therefore, we tested the effects of several known flagellar length-altering agents after reducing variability in the initial population through F-L synchronization and compared the results with those obtained by other synchronization methods. When flagellar shortening was induced with 3-isobutyl-1-methylxanthine (IBMX) (39), latrunculin B (LatB) (40), and sodium pyrophosphate (NaPPi) (41), we observed more severe shortening in F-L synchronized cells than in nonsynchronized cells or cells synchronized using all the other methods ( Fig. 2a to c, green) (Table S3). Only L-D synchronization, which is a more time-consuming synchronization method, demonstrated length reduction comparable to that seen with F-L synchronized cells ( Fig. 2a to c, blue) (Table S3). Effects on flagellar length were the most extreme in the case of NaPPimediated length resorption (Fig. 2c). After the treatment, flagellar length distribution was significantly reduced in F-L synchronized cells compared to the others and produced a more dramatic shortening of the mean flagellar length ( Fig. 2c) (Table S3). L-D synchronized cells showed a reduced effect (~41% shortening) compared to F-L synchronized cells (~46% shortening) (Fig. 2c, lower panel). In addition to testing flagellum-shortening compounds, we also tested the effects of lithium chloride (LiCl), which is known to lengthen flagella (6). The flagella were longest in F-L synchronized cells after LiCl treatment. While the levels of variance and percent change in mean flagellar length were comparable between L-D and F-L synchronized cells, each of the LiCl-treated F-L synchronized cells had flagellar length greater than 13.5 m, with an average of 17 m (Fig. 2d, green) (Table S3). Broad distributions of lengths ranging from 9 m to 20 m were observed in both nonsynchronized and M-N synchronized cells (Fig. 2d, red and purple, respectively). As a result, the effect on flagellar length was less apparent. Taken together, all of these data demonstrate that the effect of each chemical is more prominent and detectable in F-L synchronized cells when the variance in starting flagellar length is minimized. Synchronization time varies in flagellar length mutants. Flagellar length mutants with both long and short flagella have been previously isolated in Chlamydomonas (42)(43)(44). Length distributions are reportedly wider in flagellar length mutants than in wild-type cells (43). Therefore, we asked if F-L synchronization would increase our ability to detect length differences in these populations. Some long-flagella mutants have defective regeneration kinetics after amputation (43) and therefore are not suitable for F-L synchronization. However, other mutants have regeneration kinetics comparable to that of wild-type cells (43). We first considered lf4-7 mutant cells, which have the longest flagella of all identified long-flagella (lf) mutants but can regenerate their flagella with wild-type regeneration kinetics after amputation (44). As expected, prior to deflagellation, the flagellar lengths in the initial population of lf4-7 cells were distributed very widely (12 m to 28 m) (Fig. 3a, upper and lower panels) (Table S4). Following amputation, flagellar length variabilities were reduced at the 2-and 3-h time points, but the mean lengths had not yet achieved the predeflagellation lengths (Fig. 3a). We found that a duration of at least 4 h was required to regenerate flagella to their predeflagellation length. As seen with the wild-type cells, the flagella took extra time after reaching their original length (6 h of regeneration in this case) to become homogeneously distributed (Fig. 3a). Also like wild-type cells, a narrow distribution could be maintained for only a short period of time. For a mutant with short flagella, the shf1-253 mutant (42), flagella reached their predeflagellation lengths within 1.5 h following deflagellation but took an additional 1 h (2.5 h total) to distribute more narrowly (Fig. 3b, upper and lower panels) (Table S4). Finally, we studied a mutant with only slightly longer flagella than the wild type (45), the cnk2-1 mutant. Like other flagellar length mutants, these cells regenerated to normal length after 3.5 h following amputation but achieved the narrowest flagellar length distribution at 5 h of regeneration (Fig. 3c, upper and lower panels) (Table S4). These findings suggest that all cells exhibit their tightest flagellar length distribution at a time after they initially reach predeflagellation mean lengths. Therefore, mutants with longer flagella take more time and mutants with shorter flagella take less time to achieve their most narrow flagellar length distribution. Ideally, the optimal time for flagellar synchronization should be adjusted for individual strains. F-L synchronization may mask important outliers. Some flagellar length mutants have a mean flagellar length comparable to that of wild-type cells but have a positively skewed distribution that includes small numbers of mutants with extremely long flagella (43). As F-L synchronization reduces length variance, we asked if the informative long-flagella outliers would be lost after minimizing variability and would thereby decrease our ability to appropriately phenotype this class of mutant. To test this, we chose two long-flagella mutants, mutants lf2-5 and lf3-2, which were able to regenerate their flagella normally and had a large number of flagella in the wild-type range (43). When we induced regeneration for these two mutants for up to 8 h, we found that the F-L synchronization time for both the lf2-5 mutant and the lf3-2 mutant was 4 h (Fig. S3). The flagellar length distributions of lf3-2 cells demonstrated that synchronized cells had a narrow distribution, with flagella no longer than 20 m and no shorter than 10 m, which was expected. The synchronized distribution had a negative kurtosis (Ϫ0.3738), i.e., a distribution with short tails, compared to the nonsynchronized distribution, which had a positive kurtosis (ϩ0.0358) and relatively long tails. The average length seen with the synchronized population compared to the nonsynchronized population was not changed significantly (Fig. 4a). The mode changed from 12.5 to 16, with the number of cells containing 13 m to 16 m long flagella increasing remarkably in the case of F-L synchronized cells (Fig. 4a). Therefore, for the lf3-2 mutant, F-L synchronization did not affect our ability to identify a mutant phenotype despite eliminating long outliers. In the case of the lf2-5 mutant, F-L synchronization also removed outliers from the mutant population (Fig. 4b), but this time the average flagellar lengths of nonsynchronized and synchronized cells differed markedly (Fig. 4c) by left-shifting the distribution (Fig. 4b). The mode changed from 16 to 13, and the lf2-5 mutant showed an average flagellar length value of~14 m after F-L synchronization (Fig. 4b), a value which is sometimes seen in wild-type populations (Fig. S4). Also, the nonsynchronized and synchronized kurtosis values (Ϫ0.6492 and Ϫ0.6753, respectively) were not significantly different. While F-L synchronization of lf3-2 and lf2-5 mutants maintained our ability to discriminate between wild-type and mutant phenotypes by reducing the variance (Fig. 4c), the two mutants behave differently with respect to changes in descriptive statistics. Therefore, losing outliers during F-L synchronization has the potential to obscure important information following genetic perturbation. We therefore recommend testing both synchronized and nonsynchronized cells when characterizing new mutants. Flagellar length variability is related to precursor pool |
variability. As flagellar synchronization time is highly reproducible within wild-type populations, we hypothesized that there might be an internal regulator which is responsible for the narrow distribution pattern seen after 3 h of regeneration. Chlamydomonas cells have a synthesized pool of unassembled flagellar proteins or at least a preexisting pool of some protein that limits the rate of flagellar assembly (termed the precursor pool). The size of this pool is sufficient to assemble flagella to half of their normal length if new protein synthesis is inhibited (46). Limiting-precursor models of flagellar length control have been previously considered, but flagellar length appears to be maintained independently of pool size or concentration (42). However, completely blocking new protein synthesis can limit flagellar length, so we asked if imposing constraints on protein synthesis and incorporation might narrow the resulting flagellar length distribution. In such a case, reduced variability during F-L synchronization would be due to synchronizing flagellar protein synthesis through deflagellation and time-limiting flagellar protein incorporation. In order to test our hypothesis, we determined the variance in the synthesized precursor pool size after the use of different synchronization methods by deflagellating cells and allowing them to regenerate in the presence of cycloheximide (46). This allowed existing flagellar protein to be incorporated into flagella but prevented the synthesis of new protein. In these experiments, flagellar length is a proxy for the amount of limiting protein available for incorporation into flagella without protein synthesis. Precursor pool size is therefore reported in units of micrometers of flagellar length. To evaluate the relationship between flagellar variability and precursor pool variability, we compared flagella that had undergone 3 h of regeneration (tightest distribution) to flagella at two other time points corresponding to increased variability: 2 h and 5 h (Fig. 1b and S2). We performed the same comparison for nonsynchronized, L-D synchronized, and M-N synchronized cells. As expected, the results seen with all cells after synchronization but prior to deflagellation for cycloheximide treatment recapitulated our previous findings ( Fig. S5 and Table S5). When we compared precursor pool variance after different regeneration time intervals, we found a narrow distribution of pool sizes at 3 h of regeneration (medium green) but not at 2 h (light green) or at 5 h (dark green) (Fig. 5a and lower half of 5b) (Table S5). Nonsynchronous cells have a precursor pool distribution comparable to that seen at 2 and 5 h of regeneration in F-L synchronized cells (Fig. 5a, red). For all F-L synchronized cells, we observed that the variance seen with the precursor pool (Fig. 5b, lower panel) echoes the postsynchronization length variance (Fig. 5b, upper panel). We also saw that (Table S5). Because both L-D synchronization and M-N synchronization are cell cycle synchronization methods, we asked next if the narrow precursor pool variability in these cells correlated with a narrow cell size distribution and if we were circumventing the cell size dependence of the flagellar precursor pool during F-L synchronization by time-limiting protein synthesis and incorporation. Flagellar length synchronization changes the relationship between cell size and precursor pool size. It was previously reported that there is no simple relationship between cell size, flagellar length, and precursor pool size (47). However, we observed that the relative variability of flagellar length across synchronization methods is preserved when considering the variability in pool size (Fig. 5b). Given the general scaling of protein quantity with cell size (48)(49)(50)(51)(52), we tested the relationship between cell size and precursor pool size to better understand the factors influencing precursor pool and flagellar length variance across synchronization methods. We regenerated the flagella for 2 h in the presence of cycloheximide after the use of different synchronization methods and measured flagellar length to determine the preexisting precursor pool as before. This time, we also measured the corresponding cell volume. In nonsynchronized cells, cell volumes had a broad distribution (~100 m 3 to 900 m 3 ), as did the precursor pool size (~2.5 m to~9 m of flagellar length), with a significant correlation between cell size and precursor pool size (r ϭ 0.73, two-tailed P Ͻ 0.0001) (Fig. 6a). Expectedly, we found that cell volumes were very restricted, ranging from~100 m 3 to~400 m 3 , in both L-D synchronized cells and M-N synchronized cells (Fig. 6b and c, respectively), as they are cell cycle synchronized. Since smaller cells generally produce less protein (48), the restricted cell volumes of cell cycle synchronized populations also limited the protein precursor pool size to within a very narrow range (~4.5 m of flagellar length) (Fig. 6b and c). On the other hand, like the nonsynchronous cells, F-L synchronized cells had a large cell size range (~100 m 3 to 900 m 3 ). However, unlike nonsynchronous cells, which had a precursor pool size range of~6.5 m of flagellar length (Fig. 6a), F-L synchronized cells had a narrow precursor pool range (~5 m flagellar length) more comparable to the precursor pool size in cell cycle synchronized cells (Fig. 6d). While a correlation between cell size and precursor pool size was still maintained in F-L synchronized cells (r ϭ 0.71, two-tailed P Ͻ 0.0001), the slopes of the regression lines for nonsynchronous cells and F-L synchronized cells were significantly different (P ϭ 0.0012) (53-55) (Fig. 6e, red line and green line, respectively). In other words, the relationship between cell size and available precursor pool was altered upon F-L synchronization. Presumably, F-L synchronization can limit the precursor pool sizes by limiting the amount of time during which the precursor pool can accumulate without the need to limit pool size by restricting cell size through cell cycle synchronization. DISCUSSION Here we have shown a powerful new approach to improve understanding of ciliary length-related biology by characterizing a synchronization method that minimizes flagellar length variability. It is well established that flagellar length variability can be controlled by restricting the cell size (the basis of cell cycle synchronization). Our data suggest that the size of the precursor pool (the existing pool of flagellar protein not assembled into flagella) is also related to the cell size. Limiting flagellar protein is not currently considered a major factor controlling flagellar length because the amount of flagellar proteins in Chlamydomonas clearly exceeds the amount assembled into flagella, the size of the pool of unassembled flagellar precursors does not correlate with flagellar length during assembly, and flagellar length does not appear to be strongly dependent upon the number of flagella in mutants with variable flagellar numbers (41,47,56). However, while the precursor pool size is correlated with cell size, cell cycle synchronization methods severely limit both cell size and precursor pool variability. We further found that we can circumvent cell size restrictions for minimizing flagellar length variability by limiting the amount of time that the flagellar precursor pool can accumulate and incorporate into flagella. It is well known that mammalian cell ciliary studies often initiate ciliogenesis by serum starvation of confluent cells (5,19). By standardizing plating density and limiting serum starvation time prior to subsequent experimentation (thereby limiting the time window of assembly), the F-L synchronization method may be applicable to studies in mammalian cells. Assembly of full-length flagella requires a preexisting precursor pool, de novo synthesis of flagellar precursor proteins, and also incorporation of those proteins into the flagellar structure (57). The expression of genes encoding ciliary proteins is dramatically upregulated after flagellar amputation to replenish the precursor pool and to provide the proteins required for flagellar assembly (58). We propose a model for F-L synchronization (Fig. 7) where F-L synchronization via deflagellation works by stimulating a highly regulated program of gene expression and flagellar protein incorporation so that all cells can regenerate their flagella synchronously regardless of their divisional phase. Synchronization is then achieved by limiting the time window (Fig. 7, red dotted line) during which the cells are allowed to regenerate their flagella (3 h). Combining a simultaneous induction of the regeneration program and a restriction of the amount of protein synthesis and incorporation results in a tighter length distribution pattern. If time limiting protein synthesis and incorporation results in a narrow distribution of flagellar lengths, why was there increased variability of flagellar lengths at an earlier flagellar synchronization interval (2 h)? Rates of flagellar regeneration differ from cell to cell; some flagella are fast growing (Fig. 7, medium green line) and attain their original length within 2 h of amputation, while slow-growing flagella (Fig. 7, light green line) can reach only 80% of their length within that period. We saw in measurements of unassembled flagellar protein (Fig. 5b, lower panel) (see Table S5 in the supplemental material) that cells that have undergone 2 h of F-L synchronization have a smaller precursor pool (mean, 5.4 m of flagellar length) than those that have undergone 3 h of synchronization (mean, 6.2 m of flagellar length). The slow growth of some cells at 2 h postdeflagellation may therefore be due to reduced protein synthesis and accumulation. We propose that, as we extended the regeneration time beyond 2 h, slow-growing flagella finally reached their original length and fast-growing flagella approached their maximum length by reducing their assembly rate and reaching equilibrium with continuous disassembly (Fig. 7). To confirm this, we would need data at the individual-cell level rather than at the population level, which will be obtained in future studies by trapping individual motile cells in a microfluidic chamber (59). While all cells must initiate a regeneration program upon deflagellation, with increasing time, the deflagellation-induced protein synthesis and incorporation program (which decreases as a function of time and flagellar length) may be overcome by other regulating factors such as disassembly, mechanical force, proteosomal degradation, feedback control, and autophagy (60-64) (Fig. 7, dark green lines). Also, when the regeneration program no longer drives flagellar length after 3 h, cell cycle regulation may dominate, resulting in the heterogeneous flagellar length and precursor pool size distributions seen in nonsynchronized cells. In addition to maximizing our ability to detect effects in inhibitor studies, we observed that F-L synchronization can be readily applied to genetically perturbed length mutants to reduce their length heterogeneity. All mutants have different genetic defects, and we showed that several mutants responded differently to synchronization, highlighting that both synchronized cells and nonsynchronized cells should be tested when phenotyping newly identified mutants. Interestingly, we were able to discriminate long-flagella mutants from wild-type cells on the basis of the synchronization time alone, regardless of mean length. In other words, when F-L synchronization eliminated important outliers and reduced the ability to discriminate on the basis of mean flagellar length, cells still showed a flagellar synchronization profile more similar to that of long-flagella mutants (4 to 5 h synchronization time) than to that of wild-type cells (3 h synchronization time). This suggests that flagellar synchronization time itself can be a useful phenotyping parameter. Currently, the most commonly used method of reducing flagellar length variability in Chlamydomonas is cell cycle synchronization using L-D cycling. However, in L-D synchronized cells, natural variance in cells (8,35) prevent 100% synchronization of msphere.asm.org 11 flagellar length. Using F-L synchronization, we can synchronize 100% of the population through deflagellation and produce a homogeneous distribution of length by 3 h. Conventional L-D synchronization, in contrast, requires at least 3 days to achieve comparable levels of homogeneity. F-L synchronization does not require a dark chamber with automated light switching. Moreover, the entire experiment can be performed in a rich medium such as Tris-acetate-phosphate (TAP) medium instead of minimal medium, which is very sensitive to changes in pH. This facilitates the use of inhibitors that would otherwise dramatically affect the pH of the medium. F-L synchronization showed effects of length-altering chemicals on flagellar length that were equivalent to or stronger than those seen with L-D synchronization, demonstrating its utility in addition to its uniformity and simplicity. The results presented here facilitate identification of ciliary length-related defects (65-68) by increasing our ability to detect small changes in ciliary size but, more broadly, help us understand factors affecting ciliary length regulation. By inducing a fully synchronous cellular program (regeneration) that temporarily dominates multiple other factors to minimize flagellar heterogeneity, F-L synchronization also has the strong potential to benefit studies of ciliary motility or ciliary signaling. MATERIALS AND METHODS Strains and length-altering chemical treatments. Chlamydomonas reinhardtii wild-type 137c mtϩ (CC125), lf4-7 mtϪ (CC4534), shf1-253 mtϪ (CC2348), cnk2-1 (CC4689), lf2-5 |
mtϪ (CC2287), and lf3-2 mtϪ (CC2289) strains were obtained from the Chlamydomonas Resource Center at the University of Minnesota. All chemicals were purchased from Sigma (St. Louis, MO), and final concentrations of 0.4 mM IBMX, 10 mM NaPPi, 10 M LatB, 25 mM LiCl, and 10 g·ml Ϫ1 cycloheximide were used. Compounds were diluted to the indicated doses either with TAP medium or with 100% dimethyl sulfoxide (DMSO). For the chemical treatment, 1 or 2 ml of cells was treated with the indicated concentration of chemicals with the indicated controls and placed on a rotator for 90 min or 120 minutes as indicated in the text. Culture condition and different synchronization methods. All cells were maintained on TAP plates containing 1.5% agar (Difco Laboratories, Detroit, MI) (24). For liquid cultures, cells were inoculated from TAP plates at less than 2 weeks of age. Nonsynchronous culture. For nonsynchronous culture, cells were grown in liquid TAP medium for 24 h on a culture rotator drum at 25°C under conditions of continuous illumination with a light-emitting diode (LED) LumiBar with independent red and blue light control (LumiGrow, Inc.). Light/dark (L-D) synchronization. Cells were inoculated in minimal medium (M1 medium) from the TAP plates and kept in light for 12 h and then in dark for 12 h, alternating at 25°C for at least 3 days. After each L/D cycle (12 h/12 h), cultures were diluted to 2 ϫ 10 5 cells·ml Ϫ1 with fresh M1 medium. On the fourth day, after growing in the light phase for 5 h, cultures were immediately transferred to TAP medium prior to chemical treatment. Synchronization by nitrogen starvation (M-N synchronization). M-N synchronization was attained by inducing gametogenesis in nitrogen-free minimal medium for 18 to 20 h in continuous light at 25°C under a LumiBar. These cells were then transferred to TAP medium for 4 h prior to chemical treatment. Flagellar length synchronization (F-L synchronization). For F-L synchronization, Chlamydomonas cells were grown in liquid TAP medium and then induced to regenerate flagella after acid-mediated flagellar excision (69). Acetic acid (60 l of 0.5 N) was added to 1 ml of cells for deflagellation (pH ϭ 4.5). Immediately after 45 s, 70 l of 0.5 N KOH was added to neutralize the medium, which ultimately induced the flagellar regeneration. Wild-type cells were grown for 3 h for flagellar regeneration under conditions of continuous illumination with a LumiBar on a rotator drum. Flagellar length and cell volume measurement. For measurements of flagella, cells were fixed in 1% glutaraldehyde and kept at 4°C. Cells were then centrifuged at 1,000 ϫ g for 1 min and mounted between a glass slide and coverslip. Imaging was performed using a Zeiss Axioscope differential interference contrast (DIC) microscope with a 40ϫ objective lens and a Zeiss AxioCam 105 color camera. Flagellar length measurements were done by analysis of line segments and spline fitting using ImageJ software (NIH, USA). All flagella in a particular field were considered, and at least 50 flagella were measured at each time point. For cell size determination, cell volumes were calculated using the ellipsoid equation 4/3 (L/2)(W/2) 2 , where L is cell length and W is cell width (70). Flagellar length distributions and cell volumes were plotted using GraphPad Prism software version 6 (GraphPad, USA). Flagellar precursor pool determination. For flagellar precursor pool determination, cells were allowed to regenerate their flagella in the presence of 10 g·ml Ϫ1 cycloheximide following deflagellation (46). Cells processed using different synchronization methods were induced to regenerate flagella after acidic shock and were then returned to neutral pH by addition of KOH as described above and subjected to cycloheximide treatment immediately. For precursor pool determination in F-L synchronized cells, cells were allowed to regenerate their flagella for 2 h, 3 h, and 5 h after the first deflagellation and then subjected to a second deflagellation prior to cycloheximide treatment. For all cases, cells were centrifuged at 1,000 ϫ g for 2 min after neutralization and were then resuspended in TAP medium containing cycloheximide. Cells were placed on a rotator for 120 min, and flagellar length measurements were carried out to determine the amount of unassembled limiting flagellar protein. Statistical analysis. Statistical analyses were performed using GraphPad Prism software version 6 and Microsoft Excel-2010. Descriptive statistics were expressed as means and standard deviations. F tests were performed in Excel to compare the levels of variance of the data in the data set and to determine the P values (by two-tailed test). The nonparametric Mann-Whitney U test was performed for comparing two means. One-way analysis of variance (ANOVA) and Bonferroni's post hoc tests were performed for multiple comparisons and to determine P values. For all data sets, P values of Ͻ0.05 were considered statistically significant. However, Bonferroni's correction was applied when multiple pairwise comparisons were performed on a single set of data and P values were adjusted accordingly. Frequency distributions and column statistics were used to determine mode and kurtosis, respectively. For determining Pearson r values, we performed correlation analyses in GraphPad Prism. Slopes and associated standard errors (SE) were determined using linear regression (least-squares fit), also in GraphPad Prism. We also determined the difference between the two slopes in Fig. 6e using an available online calculator (53,54). Punicalagin Attenuates Disturbed Flow-Induced Vascular Dysfunction by Inhibiting Force-Specific Activation of Smad1/5 Background Pathophysiological vascular remodeling in response to disturbed flow with low and oscillatory shear stress (OSS) plays important roles in atherosclerosis progression. Pomegranate extraction (PE) was reported having anti-atherogenic effects. However, whether it can exert a beneficial effect against disturbed flow-induced pathophysiological vascular remodeling to inhibit atherosclerosis remains unclear. The present study aims at investigating the anti-atherogenic effects of pomegranate peel polyphenols (PPP) extraction and its purified compound punicalagin (PU), as well as their protective effects on disturbed flow-induced vascular dysfunction and their underlying molecular mechanisms. Methods The anti-atherogenic effects of PPP/PU were examined on low-density lipoprotein receptor knockout mice fed with a high fat diet. The vaso-protective effects of PPP/PU were examined in rat aortas using myograph assay. A combination of in vivo experiments on rats and in vitro flow system with human endothelial cells (ECs) was used to investigate the pharmacological actions of PPP/PU on EC dysfunction induced by disturbed flow. In addition, the effects of PPP/PU on vascular smooth muscle cell (VSMC) dysfunction were also examined. Results PU is the effective component in PPP against atherosclerosis. PPP/PU evoked endothelium-dependent relaxation in rat aortas. PPP/PU inhibited the activation of Smad1/5 in the EC layers at post-stenotic regions of rat aortas exposed to disturbed flow with OSS. PPP/PU suppressed OSS-induced expression of cell cycle regulatory and pro-inflammatory genes in ECs. Moreover, PPP/PU inhibited inflammation-induced VSMC dysfunction. Conclusion PPP/PU protect against OSS-induced vascular remodeling through inhibiting force-specific activation of Smad1/5 in ECs and this mechanism contributes to their anti-atherogenic effects. INTRODUCTION Atherosclerosis is a chronic inflammatory vascular disorder highly associated with endothelial cell (EC) dysfunction, vascular smooth muscle cell (VSMC) proliferation and migration, inflammatory monocyte infiltration, lipid deposition, and vascular wall remodeling (Gimbrone and Garcia-Cardena, 2016). The non-random distribution of atherosclerotic lesions is related to different patterns of blood flow and hemodynamic forces acting on the vascular wall. As an important signal transduction medium between blood flow and arterial wall, vascular ECs are constantly exposed to different flow patterns and shear stresses, including disturbed flow with low and oscillatory shear stress (OSS) and pulsatile flow with relatively high shear stress (PSS), leading to distinct impacts on the vascular wall (Zhou et al., 2014). Plaques preferentially occur at arterial branches and curvatures where the local flow is disturbed with OSS (Chiu et al., 2009). By contrast, arterial regions exposed to pulsatile flow with PSS are relatively lesion-free (Sorescu et al., 2003). Pulsatile flow with PSS generally is anti-inflammatory and anti-atherogenic, whereas disturbed flow with OSS promotes the formation and progression of atherosclerosis (Gimbrone and Garcia-Cardena, 2016). Our previous studies and others have shown that ECs are capable of perceiving OSS as a mechanical signal to induce force-specific activation of bone morphogenetic protein receptor (BMPR)-associated Smad1/5, leading to upregulation of cyclin A and downregulation of p21 and p27, thereby increasing EC cell cycle progression and proliferation (Sorescu et al., 2003;Chang et al., 2007;Zhou et al., 2012Zhou et al., , 2013. OSS-activated Smad1/5 can further promote the activation of nuclear factor-κB (NF-κB) pathways and release of pro-inflammatory cytokines interleukin-1β (IL-1β) and tumor necrosis factor-α (TNF-α) to aggravate EC injury (Zhou et al., 2012(Zhou et al., , 2013. OSS-induced pro-inflammatory responses in ECs can elicit chemotaxis and adhesion of monocytes to the EC layers mediated by intercellular adhesion molecule-1 (ICAM-1), vascular cell adhesion molecule-1 (VCAM-1), and monocyte chemotactic protein-1 (MCP-1), which further promotes atherosclerosis progression (Chiu et al., 2004). On the Abbreviations: OSS, oscillatory shear stress; PE, pomegranate extraction; PU, purified punicalagin; PPP, pomegranate peel polyphenols extraction; EC, endothelial cell; VSMC, vascular smooth muscle cell; HFD, high fat diet; BMPR, bone morphogenetic protein receptor; ICAM-1, intercellular adhesion molecule-1; VCAM-1, vascular cell adhesion molecule-1; TNF-α, tumor necrosis factor-α; MCP, monocyte chemotactic protein; TC, total cholesterol; TG, total triglyceride; iNOS, inducible nitric oxide synthase; eNOS, endothelial nitric oxide synthase; L-NAME, L-N(G)-nitroarginine methyl ester; IL-6, interleukin-6; HPLC, high performance liquid chromatography. other hand, the injured ECs produce pro-inflammatory cytokines such as TNF-α to promote phenotypic modulation, proliferation and migration of VSMCs to further aggravate atherosclerosis (Sorescu et al., 2003). Taken together, all of these sequential events suggest that force-specific activation of Smads may be a promising hemodynamic-based molecular target for intervention against disturbed flow-associated vascular disorders, such as atherosclerosis. Pomegranate peel is rich in polyphenols with the main component of punicalagin. Pomegranate extraction (PE) was reported to exert anti-atherogenic effects via lowering circulating levels of low-density lipoprotein and formation of macrophagederived foam cells (Al-Jarallah et al., 2013;Rosenblat et al., 2013;Atrahimovich et al., 2016Atrahimovich et al., , 2018. Prolonged PE supplementation inhibited OSS-related atherosclerosis by upregulating endothelial nitric oxide synthase (eNOS) expression and modulating oxidation-sensitive gene expression in ECs (de Nigris et al., 2007). However, whether the extraction of pomegranate peel polyphenols (PPP) and its purified compound punicalagin (PU) exert ameliorative effects on OSS-induced vascular dysfunction and hence protect against atherosclerosis remain unclear. Anti-atherogenic Studies of PPP/PU Healthy 6-week-old male low-density lipoprotein receptor knockout (LDLR −/− ) mice (Jackson Laboratory, United States) were housed with free access to water and standard laboratory chow diet. Atherosclerosis was induced by feeding LDLR −/− mice with a high fat diet (HFD) containing 20% lard and 0.5% cholesterol for 12 weeks. LDLR −/− mice were randomly assigned to the treatment groups. Vehicle group mice were administrated with phosphate-buffered saline (PBS) by gavage. In drug treatment group, PPP or PU was dissolved in PBS and given daily to mice by gavage at different doses. After 12 weeks, all mice were euthanized with an overdose of sodium pentobarbital. The blood samples were taken from the mice after 12-h fasting and centrifuged (Eppendorf 5418R, Eppendorf Corporation, Hamburg, Germany) at 1,400 × g for 10 min at 4 • C. And the plasma was taken for measure of plasma triglycerides (TG) and total cholesterol (TC) according to manufacturer's instructions. The aortas were isolated for en face immunostaining and aortic roots were used for Oil Red O and immunohistochemical staining. Briefly, mouse aortas were fixed with 4% paraformaldehyde for 2 h and dehydrated in 20% sucrose solution overnight and aortic span sections were stained by 0.5% Oil Red O for detecting lipid deposition in plaques. Positive areas of Oil Red O staining in the lesions were quantified using Image ProPlus 6.0 image analysis software. Mouse aortic roots were cut into serial frozen sections containing 300 cross-sections in 7 µm thickness, and ten cross-sections obtained from an interval of 30 sections were used for Oil Red O staining to analyze plaque areas. Atherosclerosis sections were blinded per individual to avoid the bias. Representative images of immunohistochemical staining of Mac-2 indicated the number of macrophages in atherosclerotic lesions. Endothelium-Dependent Vaso-Protective Assay Effects of PPP and PU on eNOS activity were detected by myograph. Briefly, Sprague Dawley male rats were euthanized by an overdose of carbon dioxide. The aorta was dissected and excised quickly and placed in ice-cold physiological saline solution (PSS) containing (in mmol/l) 119 NaCl, 4.7 KCl, 25 NaHCO 3 , 1.17 KH 2 PO 4 , 1.17 MgSO 4 , 1.6 CaCl, |
and 5.5 dextrose, gassed by 95% O 2 -5% CO 2 . The aorta was then cut into 3 mm vessel rings and the EC layer was stripped off in -endo group. The vessel rings were incubated with PPP or PU at different concentrations (from 1 to 75 µg/mL for PPP and 1-50 µg/mL for PU) in the presence and absence of nitric oxide synthase inhibitor L-N(G)-nitroarginine methyl ester (L-NAME, 100 µM). Vessel viability was detected by the stimulation of phenylephrine (PE) and acetylcholine. PE stimulation would constrict the integrated vessel ring and acetylcholine stimulation could relax the PEconstricted vessel. When the EC layer was removed or treated with L-NAME, acetylcholine could not make the constricted vessel ring relaxation. Aortic Stenosis Studies in Rats Aortic stenosis was induced in the rat by using a U-shaped titanium clip to constrict its abdominal aorta for 2 weeks as described (Miao et al., 2005;Zhou et al., 2012Zhou et al., , 2013. Briefly, following anesthetization with isoflurane, the rat was laid supine and a lower midline abdomen incision was made. Thereafter, the part of the intestine was gently lifted out of abdomen and kept moist with saline throughout the surgical procedure. The aorta, left and right common iliac artery were exposed and the accompanying vein was carefully separated. The clip was held with a pair of forceps and placed around the isolated segment (approximate 1 cm from the arterial bifurcation) to partially constrict rat's abdominal aorta. The extent of clipping was controlled by placing a stopper of given size between the two arms of the forceps. Our previous study using ultrasonography indicated that placement of the U-clip resulted in a 65% constriction of the aorta diameter, which induced an accelerated forward laminar flow in the constricted region, followed by a pronounced oscillating flow with the existence of retrograde velocities downstream in the region of poststenotic dilatation (Zhou et al., 2012). The flow patterns and wall shear stress distributions in the constricted rat abdominal aorta were further characterized by computational fluid dynamic modeling using the Comsol Multiphysics software, which confirmed the existence of recirculation eddies with retrograde velocities downstream to the constricted sites (Zhou et al., 2012). PPP (750 mg/kg/day) or PU (500 mg/kg/day) was given daily by gavage 3 days before operation and the daily treatment lasted for additional 2 weeks. All rats were sacrificed by the end of treatment and aortas were perfusion-fixed with 4% paraformaldehyde at 120 mmHg. The fixed aortas were embedded in paraffin blocks for immunohistochemical studies. Isolation and Culture of Primary ECs ECs were isolated from fresh human umbilical cords with collagenase perfusion technique (Gimbrone, 1976). The cell pellets were resuspended in a culture medium consisting of medium 199 (M199, Gibco, Grand Island, NY, United States) supplemented with 20% fetal bovine serum (FBS, Gibco) and 1% penicillin/streptomycin (Gibco). ECs were grown in Petri dishes for 3 days and then seeded onto glass slides (75 by 38 mm, Corning, NY, United States) pre-coated with fibronectin (Sigma) to reach confluence. The culture medium was then replaced by the identical medium that contained only 2% FBS, and the cells were incubated for 24 h before use. Oscillatory Flow Apparatus The cultured ECs were subjected to shear stress in a parallel-plate flow chamber, as previously described (Zhou et al., 2012(Zhou et al., , 2013. The chamber containing the cell-seeded glass slide fastened with the gasket was connected to a perfusion loop system, kept in a constant-temperature controlled enclosure, and maintained at pH 7.4 by continuous gassing with a humidified mixture of 5% CO 2 in air. The fluid shear stress on the ECs can be estimated as τ = 6µQ/wh 2 , where τ is shear stress, Q is the flow rate, and µ is dynamic viscosity of the perfusate. The oscillatory flow is composed of a low level of mean flow with shear stress at 0.5 dynes/cm 2 supplied by a hydrostatic flow system and the superimposition of a sinusoidal oscillation using a piston pump with a frequency of 1 Hz and a peak-to-peak amplitude of 4 dynes/cm 2 . In the in vitro experiments, ECs were pretreated with PPP (50 µg/mL) or PU (50 µg/mL) for 30 min and then subjected to OSS (0.5 ± 4 dynes/cm 2 ) in a parallelplate flow chamber for 4 or 24 h, or stimulated with BMP (100 ng/mL) for 30 min or TNF-α (100 U/mL) for 4 h in the presence of PPP or PU. Immunofluorescence Assay (IFA) ECs were seeded on coverslips in the culture plate wells and subjected to OSS. IFA was performed using antibodies against ki67 (Abcam), as described (Wang et al., 2016). The coverslips were mounted onto the slides using Fluoromount-G clear mounting medium containing DAPI (Southern Biotech, Birmingham, AL, United States). The fluorescence signals were observed via fluorescence microscopy (Nikon ECLIPSE Ti), and images were taken using NIS-Elements F software (Nikon). Isolation and Culture of Primary VSMCs Primary VSMCs were obtained from rat thoracic arteries, as described (Wang et al., 2015;Shen et al., 2019). Briefly, male SD rats weighing about 100 g were anesthetized and thoracic arteries were carefully excised, and the surrounding perivascular adipose tissues and connective tissues were trimmed off. The arteries were washed in 0.01 M PBS containing 100 g/mL streptomycin and 100 IU/mL penicillin. Arterial ectoderm was removed and sliced with an ophthalmic scissor, and the vascular endothelium was scratched gently using curved dissection forceps. Vascular tissues were washed and cut into small pieces. They were placed at the bottom of a 100 mm culture dish filled with 1 mL DMEM with 10% FBS, 100 g/mL streptomycin and 100 IU/mL penicillin and incubated at 37 • C for 6 h until they all stick to the bottom of the dish. Thereafter, culture medium was replenished every 4 days. The cells between passages 4 and 7 were used. VSMC Migration Assay The wound healing assay was performed to test migratory capability of VSMCs. Briefly, VSMCs were seeded into 24well tissue culture plates. The cells were incubated in DMEM supplemented with 0.5% FBS for 24 h to reach 70-80% confluence. The cell layer was scratched gently with a new 200 µL pipette tip across the center of the well. The cells were washed three times in 1 × PBS to remove the detached cells and incubated in DMEM supplemented with 0.5% FBS. The cells were treated with TNF-α (100 U/mL) in the absence and presence of PPP or PU for additional 24 h, and then washed twice in 1 × PBS and finally stained with 1% crystal violet in 2% ethanol for 30 min. Photos were taken for the stained monolayer on a microscope. Multiple views of each well were documented, and each experimental group was repeated 5 times. RNA Isolation and Quantitative Real-Time PCR (RT-PCR) Total RNA was extracted using Trizol reagent (Invitrogen, United States) and the first-strand cDNA was generated using an RT kit (Invitrogen, United States). Quantitative real-time PCR was performed using primers shown in Supplementary Table 2. Amplifications were performed using an Opticon-Continuous Fluorescence Detection System (MJ Research) with Eva Green fluorescence dye (Molecular Probes, Eugene, United States). All samples were quantitated by using the comparative CT method for relative quantitation of gene expression, normalized to GAPDH levels. Statistical Analysis All statistical analyses were performed using GraphPad Prism for Windows (Version 4, San Diego, CA, United States). Values were expressed as mean ± standard error of the mean (SEM). All data sets were tested for normal distribution. For normally distributed data, unpaired t-test, one-way ANOVA with Tukey post-test or paired t-test were used as most appropriate. All results were considered significantly as p < 0.05. RESULTS The Anti-atherogenic Effect of PPP LDLR −/− mice in 6 weeks old were fed with a HFD and treated daily with PPP for 12 weeks (Figure 1A). The En face immunostaining revealed that treatment with PPP at 750 mg/kg reduced the areas of Oil Red O-stained plaques by 56% in mouse aortas, whereas PPP at 250 mg/kg and 500 mg/kg decreased the plaque areas by 23 and 37%, respectively. These data indicate that Frontiers in Cell and Developmental Biology | www.frontiersin.org PPP exerts anti-atherogenic effects in vivo in a dose-dependent manner (Figures 1B,C). The lesion area of the aortic root is an important parameter of aortic stenosis and correlates with the severity of atherosclerotic plaques. Thus, plaque areas were further measured in crosssections of aortic roots. The Oil Red O-staining results showed that treatment with PPP at 750 mg/kg reduced the plaque area by 49% compared to the vehicle control, and plaque area in the aortic roots was also reduced in the other two PPP-treated groups, with 500 mg/kg being more effective than 250 mg/kg (Figures 1D,E). Immmunohistochemical staining of the aortic roots showed dramatic reduction in macrophage content stained by Mac-2 in the lesion areas from the PPP-treated LDLR −/− mice, indicating that PPP alleviated infiltration of inflammatory macrophages in the plaques (Figure 1D). Disorder of plasma lipid metabolism has been deemed to play an important role in the progression of atherosclerosis. Treatment of LDLR −/− mice with 750 mg/kg or 500 mg/kg PPP significantly lowered plasma levels of TC and TG (Figures 1F,G). No difference was observed in body weight among all groups (Figure 1). PU Is an Active Anti-atherogenic Compound in PPP Since PU is the major bioactive compound in PPP, we investigated whether PU treatment produces the similar protective effects on atherosclerosis as PPP. PU was administered daily to LDLR −/− mice on HFD for 12 weeks (Figure 2A). The results showed that treatment with PU at 500 or 250 mg/kg reduced Oil Red O-stained atherosclerotic plaque areas in mouse aortas by 57 and 42%, respectively (Figures 2B,C). The Oil Red O staining of the plaque areas in aortic roots showed that PU treatment also decreased the lesion areas by 75 and 60%, respectively (Figures 2D,E). The number of infiltrated macrophages in aortic roots was reduced by PU (Figure 2D), and this effect was similar to that of PPP. In addition, PU treatment lowered plasma concentrations of TC and TG in HFD-fed LDLR −/− mice (Figures 2F,G) without affecting body weight ( Figure 2H). These results indicate that PU is the major anti-atherogenic component in PPP. Next, we examined whether PPP and PU could inhibit macrophages to uptake lipids and their transformation into foam cells. The Oil red O-staining results showed that both PPP and PU reduced ox-LDL uptake into macrophages and thus inhibited foam cell formation (Supplementary Figure 2A). Meanwhile, PPP and PU decreased the lipopolysaccharide (LPS, 1 µg/mL)-induced expression of pro-inflammatory cytokines or chemokines including IL-1β, IL-6, TNF-α, and MCP-1, and inducible nitric oxide synthase (iNOS) (Supplementary Figures 2B-F). The results showed that PPP and PU inhibited macrophagedirected foam cell formation and further blocked the inflammatory cascade of macrophages to retard atherosclerosis progression, which were in agreement with the results of previous reports (Kaplan et al., 2001;Rosenblat et al., 2013;Aharoni et al., 2015). PPP and PU Exert Endothelium-Dependent Vaso-Protective Effects Then we examined whether PPP and PU can exert atheroprotective effects beyond the above effects but through improving vascular function. It is known that ECs are exposed to regular laminar shear stress in the normal physiological condition, which stimulates the release of nitric oxide (NO) by the sustained activation of eNOS (Buga et al., 1991). However, eNOS activity is reduced at sites of perturbed shear stress with OSS (Wang et al., 2019). There was a study showing that pomegranate juice enhances the biological actions of NO (Ignarro et al., 2006). Therefore, we investigated the effects of PPP and PU on the vascular reactivity. Firstly, the aorta vessel rings were incubated with PPP or PU at the designated concentrations with or without eNOS inhibitor L-NAME. Myograph system was used to detect the relaxation on pre-constricted vessel rings with endothelium (control) or without endothelium (-endo). The results showed that PPP and PU had a relaxation effect on PE-induced vasoconstriction, and this effect was abolished when the aortas were treated with L-NAME or the endothelia on the aortas were stripped off, demonstrating that the vasodilation effects of PPP and PU are endothelium-dependent (Figures 3A-D). These results indicate that PPP and PU may exert atheroprotective effects through improvement of endothelial function. PPP and PU Inhibit Activation of Smad1/5 in Vascular Endothelium Induced by |
Disturbed Flow in vivo Our previous studies demonstrated that disturbed flow can induce force-specific activation of phospho-Smad1/5 in the post-stenotic sites, where the local flow is disturbed with OSS (Zhou et al., 2012). Thus, we examined whether PPP and PU could modulate this force-specific activation of Smad1/5 in post-stenotic sites in vivo. Rat abdominal aorta was subjected to a constriction by using a U-clip (Figure 4A), which can produce a disturbed flow region downstream to the constricted site, as described (Zhou et al., 2012(Zhou et al., , 2013. Either PPP or PU was administrated by gavage daily to the rats. Immunohistochemical examination of serial sections of the constricted aortas showed that the post-stenotic sites exhibited a high detection level of phospho-Smad1/5 in the luminal EC layer. By contrast, there was virtually no detectable staining of phospho-Smad1/5 in the upstream and middle point of constriction ( Figure 4B). Treatment with PPP or PU dramatically decreased the elevation of phospho-Smad1/5 induced by disturbed flow at post-stenotic sites (Figures 4C,D). These results indicate that PPP and PU are effective in inhibiting force-specific activation of Smad1/5 in ECs induced by disturbed flow in vivo. PPP and PU Inhibit OSS-Induced Smad1/5 Phosphorylation in ECs in vitro We next further tested the effects of PPP or PU on Smad1/5 activation in ECs in response to disturbed flow with OSS. Application of OSS to ECs induced a sustained phosphorylation of Smad1/5 in ECs over the 24-h period. The phosphorylation of Smad1/5 in ECs was induced rapidly (4 h) by OSS and remained elevated after 24 h, as compared with static control ECs. This OSS-induced phosphorylation of Smad1/5 in ECs was normalized to the basal level after PPP or PU treatment (Figures 5A,B). Previous studies showed that OSS-induced activation of Smad1/5 in ECs is specifically through the activation of BMPRs (Zhou et al., 2012(Zhou et al., , 2013. Thus, we treated ECs with BMP for 30 min to active BMPR and its downstream Smads as a control experiment to explore the effects of PPP and PU on the BMPelicited signaling in ECs. The results showed that PPP and PU significantly suppressed BMP-induced phosphorylation of Smad1/5 in ECs (Figure 5C). PPP and PU Inhibit OSS-Induced Pro-inflammatory Responses of ECs EC inflammation is the early event in the pathogenesis of atherosclerosis, and OSS can activate EC pro-inflammatory responses by triggering the release of signaling molecules such as TNF-α to promote atherosclerosis progression (Zeiher et al., 1991;Sun et al., 2019). Thus, ECs were stimulated with OSS for 4 h or TNF-α for 4 h to stimulate the pro-inflammatory conditions of ECs and the effects of PPP or PU were examined. As expected, both OSS and TNF-α augmented the expressions of ICAM-1, VCAM-1 and E-selectin at both mRNA and protein Figure 3 and Figure 6). Pretreatment with PPP or PU abolished either OSS or TNF-αinduced protein expression of these molecules (Figures 6A,B). Likewise, mRNA expression of these genes was also reversed by PPP or PU treatment (Supplementary Figures 3A,B). These results indicate that PPP and PU exert inhibitory effects on OSS-induced pro-inflammatory responses in ECs. PPP and PU Inhibit OSS-Induced Proliferation of ECs OSS-induced inflammation promotes proliferation of ECs. PPP or PU treatment significantly decreased protein expression of the pro-inflammatory molecules ICAM-1 and E-selectin in ECs after exposing to OSS for 24 h (Figures 7A,B). In addition, OSS-induced upregulation of cell cycle regulatory proteins in ECs, including Cyclin A and pRb, was all repressed by treating the ECs with either PPP or PU (Figures 7A,B). Results of immunofluorescence staining of Ki67, a marker of EC proliferation, showed that PPP or PU treatment significantly suppressed OSS-induced EC proliferation ( Figure 7C). These data demonstrated that PPP and PU could inhibit force-specific activation of Smad1/5, which consequently attenuated proinflammatory responses and proliferation of ECs induced by disturbed flow with OSS. PPP and PU Inhibit TNF-α-Induced VSMC Migration, Phenotypic Modulation, and Pro-inflammatory Responses Injured ECs and activated monocytes in plaques can release various growth factors, such as platelet-derived growth factor (PDGF), to promote migration and phenotypic modulation of VSMCs (Schachter, 1997). Since migration, proliferation, and phenotypic modulation of VSMCs are the critical factors contributing to the progression of atherosclerosis (Gomez and Owens, 2012), we first examined the effects of PPP and PU on TNF-α-induced migration of VSMCs using the wound-healing assay. The results showed that PPP and PU inhibited TNFα-induced migration of VSMCs ( Figure 8A). To further explore the effects of PU on phenotypic modulation of VSMCs, the gene expression of the contractile VSMC marker, i.e., SMα-actin, was determined ( Figure 8B). The results showed that SMα-actin expression was reduced by TNF-α and this effect was reversed by PU treatment. The synthetic phenotype of VSMCs can release pro-inflammatory IL-6, and c-fos was shown to promote VSMC proliferation (Sylvester et al., 1998). We found that PU inhibited TNF-α-induced expression of IL-6 and c-fos in VSMCs ( Figure 8C). Taken together, these results indicate that PPP and PU can protect against TNF-α-induced proliferation, migration, inflammation, and phenotypic modulation in VSMCs. All of these responses participate in the progression of atherosclerosis. DISCUSSION Accumulating evidence show that polyphenols from pomegranate fruit, juice and PE are beneficial to human health (Al-Jarallah et al., 2013;Kalaycioglu and Erim, 2017). Extracts from natural plants may have potential to be developed as new drugs for prevention and treatment of cardiovascular diseases (AlMatar et al., 2018). Previous studies have shown that PE exerts an antiatherogenic effect (Aviram et al., 2000;Kaplan et al., 2001;Rosenblat et al., 2006;Estrada-Luna et al., 2019), and lipidlowering effect may account for part of its anti-atherogenic properties (Hou et al., 2019). In the present study, we used a mouse model of atherosclerosis through feeding LDLR −/− mice with HFD for 12 weeks. The results showed that both PPP and PU can lower the lipids and reduce plaque formation. The effect of PU was greater than PPP, suggesting that PU is the effective anti-atherogenic agent in PPP. These results are consistent with the previous report showing that PE reduces plasma lipid levels in SR-B1/apoE double knockout mice (Al-Jarallah et al., 2013). In addition, other researchers reported that PE protects against atherosclerosis through inhibiting foam cell formation in the lesion areas (Aharoni et al., 2015;Bi et al., 2019). Our results provide the first line of evidence to show that treatment of macrophages with PPP or PU reduces ox-LDL uptake and inhibits LPS-induced pro-inflammatory responses, manifested by reduced mRNA expression of several pro-inflammatory factors, including IL-6, IL-1β, iNOS, MCP-1, and TNF-α. Endothelial dysfunction is the initial step for the development of atherosclerosis. Disturbed flow with OSS is highly recognized to be the initial cause of EC dysfunction and pathogenesis of atherosclerosis (Buga et al., 1991;Lee et al., 2015). Under normal physiological condition, the hemodynamic forces stimulate NO release through expressing eNOS in ECs (Silacci et al., 2001), whereas eNOS activity is shown to be reduced at sites of perturbed shear stress (Go et al., 2014;Marchio et al., 2019). Previous studies showed that PE ameliorates perturbed shear stress-related atherosclerosis by regulating the expression of eNOS and oxidation-related genes in ECs (de Nigris et al., 2005;de Nigris et al., 2007). In addition, pomegranate juice was shown to enhance the biological actions of NO (Ignarro et al., 2006). Our results advanced the new notion to demonstrate that the vasoprotective effects of PPP and PU are endothelium-dependent. All these results suggest that the anti-atherogenic effects of PPP and PU may be attributable to their EC protective effects under disturbed flow with OSS. Earlier studies suggest that laminar blood flow with PSS in the straight part of the arterial tree modulates cellular signaling and EC function, and is anti-atherogenic (Qin et al., 2007). By contrast, disturbed flow and its associated OSS in the branches and curvatures of the arterial tree promote EC dysfunction and thus aggravate atherosclerosis (Chiu and Chien, 2011). Our previous studies (Zhou et al., 2012(Zhou et al., , 2013 investigated the role of OSS in modulating EC mechanotransduction and hence the development of atherosclerosis, and demonstrated that EC layer expressed high levels of phosphorylation of BMPRspecific Smad 1/5 in the lesion areas exposed to disturbed flow with OSS. In this study we further generated an in vivo disturbed flow/OSS model in rats by using an U-shaped clip to constrict the rat abdominal aorta, and the results showed that high levels of phospho-Smad 1/5 were detected in the EC layer at post-stenotic regions of rat aortas, where disturbed flow with OSS occurred. PPP or PU administration significantly inhibited the OSS-induction of phospho-Smad 1/5 in the stenosed areas. In vitro studies on the ECs subjected to OSS also showed that PPP and PU attenuated OSSinduced Smad1/5 activation. In addition, PPP and PU could also reverse BMP-induced Smad1/5 activation. Our previous studies have demonstrated that disturbed flow with OSS can induce Smad1/5 phosphorylation through the sustained induction of bone morphogenetic protein receptor (BMPR) type IB (BMPRIB)-αvβ3 integrin association in ECs (Zhou et al., 2012(Zhou et al., , 2013. This OSS-induced sustained association of BMPRIB-αvβ3 integrin was mediated by the intracytoplasmic kinase domain of BMPRII and subsequently activated the Shc/focal adhesion kinase (FAK)/extracellular signal-regulated kinase (ERK) cascade, leading to Smad1/5 activation. Thus, it is possible that PPP and PU may share common pathways to exert regulatory effects on Smad1/5 activation through the BMPRII/BMPRIB-αvβ3 integrin signaling cascade in ECs in response to disturbed flow or cytokines/growth factors. The detailed mechanisms by which PPP and PU exert protective effects on ECs remain an interesting issue that warrants further investigation. Previous studies indicated that OSS promotes inflammation and proliferation of vascular endothelium to aggravate EC dysfunction (Ajami et al., 2017;Sun et al., 2019). Application of OSS to ECs activated Smad 1/5 and led to up-regulation of cyclin A and down-regulation of p21 and p27 in ECs and hence EC proliferation (Zhou et al., 2012(Zhou et al., , 2013. Our results showed that PPP or PU treatment greatly retards cell cycle progression and proliferation of ECs. Besides, Gimbrone and Garcia-Cardena (2016) reported that OSS up-regulated the expression of adhesion molecules such as ICAM-1 and VCAM-1, which increased the recruitment of monocytes to endothelium, thus promoting vascular remodeling. Indeed, up-regulations of ICAM-1, VCAM-1, and E-selection in ECs after OSS exposure or TNF-α induction were also observed in the present study, and PPP or PU treatment inhibited the pro-inflammatory responses of ECs. We also found that the effect of PU on ICAM-1/VCAM-1 is more effective than PPP. This result is probably due to the reason that PU is a purified, effective compound of PPP, which is only a mixture. To sum up, our study advanced the new notion that treatment with PPP or PU can inhibit disturbed flow/OSS-induced activation of Smad1/5 and strongly suppressed the pro-inflammatory responses and proliferation of ECs induced by disturbed flow with OSS. There is considerable evidence that laminar flow and associated shear stress significantly inhibit EC cell cycle progression and proliferation and enhance EC migration, and hence are atheroprotective. In contrast, disturbed flow with OSS can promote EC proliferation and inhibit EC migration, and hence is thought to be atherogenic (Chiu and Chien, 2011). Our previous studies also demonstrated that disturbed flow-induced activation of Smad1/5 can promote EC cell cycle progression and proliferation, which may contribute to the formation and progression of atherosclerosis (Zhou et al., 2012(Zhou et al., , 2013. In addition, there has been considerable evidence that disturbed flow with OSS or reduction in blood blow and shear stress can induce EC apoptosis and death (Chiu and Chien, 2011). Whether PPP/PU can modulate EC migration and exert protective effects on ECs to inhibit disturbed flow-induced EC apoptosis and death warrant further investigation. In addition to ECs, VSMCs also play a vital role in maintaining vascular integrity and normal physiological function (Ross, 1975;Chistiakov et al., 2015;Lao et al., 2015). The aberrant interaction between ECs and VSMCs under pathological conditions promotes the occurrence and development of cardiovascular diseases, including atherosclerosis (Lao et al., 2015). It is widely accepted that VSMCs undergo phenotypic modulation during the progression of atherosclerosis, and the contractile phenotype of VSMCs converts to the synthetic phenotype, which triggers release of many pro-inflammatory factors, such as IL-6. Growth factors and inflammatory cytokines released from the injured ECs or other types of inflammatory cells in the plaques promote migration and pathological phenotypic modulation of VSMCs. Moreover, c-fos was |
shown to promote VSMC pathological proliferation (Hsieh et al., 1998;Sylvester et al., 1998;Fang et al., 2004). OSS-induced EC injury may promote abnormal activation of VSMCs, thereby accelerating pathological remodeling of blood vessels and promoting atherosclerosis. Here we reported for the first time that PU may inhibit migration, proliferation and phenotypic modulation of VSMCs induced by TNF-α, suggesting that PU has protective effects on VSMCs in response to inflammation. In addition to the direct effects of PPP/PU on VSMCs, it is also possible that the protective effects of PPP/PU on VSMCs might come from the anti-inflammatory effects of PPP/PU on ECs which may reduce the release of cytokines, such as TNF-α. In summary, the present study demonstrates for the first time that PPP and PU protect EC dysfunction by inhibiting OSS-induced proliferation and inflammation. These protective effects of PPP and PU may be attributable to their inhibition of force-specific activation of Smad1/5 in ECs. Furthermore, the present results show that PPP and PU can inhibit inflammation, migration and phenotypic modulation of VSMCs under pro-inflammatory stimulation. Our findings provide new insights into the mechanisms by which PPP and PU inhibit disturbed flow/inflammation-induced EC dysfunction and VSMC proliferation, migration, and phenotypic modulation, with the consequent inhibition in atherosclerosis. DATA AVAILABILITY STATEMENT The original contributions presented in the study are included in the article/Supplementary Material, further inquiries can be directed to the corresponding author/s. ETHICS STATEMENT The studies involving human participants were reviewed and approved by the experiments for the use of human umbilical cords were approved by the Hospital Human Subjects Review Committee (IRB Approval Nos. CGH-P101088 and CGH-CT9672) of Hsinchu Cathay General Hospital and were conducted under the guidelines established by the Ethics Review Board of National Health Research Institutes, Taiwan. Written informed consent was obtained from all individuals. The patients/participants provided their written informed consent to participate in this study. The animal study was reviewed and approved by This study was carried out in accordance with the principles of the Basel Declaration and Recommendations of Animal Care and Use Committee and PU IRB Laboratory Animal Welfare Committee in Peking University; the latter committee approved the protocol in this study (Approval No. LA2017193). AUTHOR CONTRIBUTIONS RQ and J-JC designed the study and revised the manuscript. X-LG and G-ZM prepared PPP extracts, purified PU, and analyzed the contents of PPP and PU by HPLC. Z-YC helped preparation and analysis of PPP and PU. GA and GL analyzed the data, prepared the figures, and wrote the manuscript. GL and W-LS performed in vivo studies. GA, C-IL, and P-LL performed in vitro experiments. Y-XW and X-YT performed endotheliumdependent vaso-protective assay. All authors read and approved the final manuscript. Expression of ID2 and Cyclooxygenase 2 Genes in Endometrial Tissues and Their Clinical Significance in the Pathogenesis of Endometrial Polyps Reproductive Medicine Center, Department of Obstetrics and Gynecology, The First Affiliated Hospital of Anhui Medical University; Anhui Province Key Laboratory of Reproductive Health and Genetics; Biopreservation and Artificial Organs, Anhui Provincial Engineering Research Center, Hefei, Anhui 230088, 1Department of Cardiovascular medicine, The First Affiliated Hospital of USTC, Division of Life Sciences and Medicine, University of Science and Technology of China, Hefei, Anhui 230001, China Exclusion criteria: Patients with endometrial tumors, uterine fibroids and other diseases; Patients with severe liver and kidney dysfunction; Patients with severe endocrine, respiratory, digestive and other system diseases; Patients with incomplete clinical data. There was no statistically significant difference in age between the two groups of patients (p>0.05), and they were comparable. The messenger ribonucleic acid (mRNA) expressions of ID2 and COX-2 in the endometrial tissues of the two groups were detected by Quantitative reverse transcription polymerase chain reaction (qRT-PCR): Endometrial tissues of the two groups of patients were selected and two samples were retained. The first specimen was washed with phosphate-buffered saline (PBS) and stored in liquid nitrogen tank. The relative mRNA expression levels of ID2 and COX-2 were detected by real-time fluorescence quantitative PCR. Total RNA was extracted with Trizol kit. The complementary Deoxyribonucleic acid (cDNA) was synthesized with reference to the reverse transcription kit. The reaction system was 20 µl: 2x SYBR (Synergy Brands, Inc.) mix 10 µl, cDNA 1 µl and double-distilled water (ddH 2 O) 8 µl. The upstream and downstream primers were 0.5 µl each, with Beta-Actin (β-actin) as internal reference. The primers were synthesized by Wuhan Biomisp Co., Ltd. And the design of primers is as follows ( Table 1). The reaction set-up procedure was as follows: pre-denaturation at 95º for 1 min; 95º for 15 s, 95º for 15 s, 60º for 15 s, 72º for 2 min, with a total of 40 cycles. The expression levels of ID2 and COX-2 mRNA in endometrium were quantitatively analyzed by 2 -ΔΔ CT . The expressions of ID2 and COX-2 were detected by western blot assay: The second endometrial tissue samples retained in 1.2.1 were ground and added into protein lysis buffer containing protease inhibitor to extract the total protein of endometrial tissue and determine the total protein. The levels of ID2 and COX-2 protein in endometrium were detected by western blot with β-actin as internal reference. The detection was carried out strictly in accordance with the operation instructions. The pathological process is similar to the proliferation and migration of tumor tissue. At present, it is believed that the factors related to cell biology are involved in the development of endometrial polyposis. Studies have shown [3,4] that Deoxyribonucleic acid binding inhibitor 2 (ID2) is related to the proliferation and migration of tumor cells. Cytochrome P450 aromatase (P450arom) is a speed limiting enzyme in estrogen synthesis. Many studies have confirmed that in endometrial polyps and endometrial cancer, the expression of P450arom in local lesions is positively correlated with cyclooxygenase 2 (COX-2) [5][6][7] . At present, there are relatively few studies on the expression of ID2 and COX-2 in endometrial tissues of patients with endometrial polyposis. Therefore, this study is to explore the abnormal expression of ID2 and COX-2 in endometrial tissue of patients and their clinical application value. MATERIALS AND METHODS 38 patients with endometrial polyps diagnosed and treated in gynecology department of our hospital from October 2020 to December 2020 were selected as the experimental group. All patients were pathologically diagnosed as endometrial polyp, and the corresponding polyp tissues and endometrial tissues around the polyp were taken. In addition, 36 cases of women who underwent insertion of intrauterine contraceptive device in our hospital during the same period were scraped a little endometrium before placing the intrauterine device (IUD). The samples were taken in the first half of the menstrual cycle, and were confirmed as normal endometrial tissue by pathology. All the women did not receive any hormone therapy 3 mo before operation, and there were no estrogen related diseases such as cervical polyps, uterine fibroids, endometrial polyps, etc. there was no significant difference in age between the two groups (p>0.05). The age of patients in the experimental group was 27-45 (33.42±7.14) y old, and that of the control group was 25-47 (34.76±8.30) y old. This study was approved by the ethics committee of our hospital, and the participants were informed and agreed. Inclusion criteria: Patients who met the diagnostic criteria of endometrial polyps; Patients without hormone therapy; Patients without other serious hormone disorders. images were taken by Tanon software, and the protein expression level was analyzed semi quantitatively Detection and observation indicators: 5 ml of fasting venous blood was taken from two groups and divided into two parts immediately. One sample was centrifuged, and the supernatant was aspirated. The serum levels of vascular endothelial growth factor (VEGF) and tumor necrosis factor-alpha (TNF-α) were determined by enzyme-linked immunosorbent assay (ELISA) according to the instructions of the test kit. The other sample was placed in the anticoagulant tube, and the detection reagents of interleukin 1 beta (IL-1β) and cancer antigen 125 (CA 125) were added. The expression level of IL-1β and CA 125 was detected by colloidal gold method. The levels of VEGF, TNF-α, IL-1β and CA 125 in the two groups were observed, and the correlation between the expressions of ID2 and COX-2 and the levels of VEGF, TNF-α, IL-1β and CA 125 were analyzed. The correlation between ID2 and COX-2 expression and clinical indicators of patients was compared and analyzed. Analysis and Statistics: Statistical software, Statistical Package for the Social Sciences SPSS 22.0 was used to analyze the data. The quantitative data were represented as mean value±standard deviation, and the comparison between the two groups was performed by t test. In addition, the qualitative data were described by the number of cases (%), while the comparison between groups was conducted by F test (statistical test). Moreover, the correlation analysis between variables was performed by person correlation analysis. p<0.05 was significant difference. RESULTS AND DISCUSSION The results of PCR showed that compared with the control group, the expressions of ID2 and COX-2 mRNA in the experimental group were significantly increased, and the difference was statistically significant (p<0.05), as shown in fig. 1. According to the results of western blot, the protein levels of ID2 and COX-2 in endometrial tissue of the experimental group were significantly higher than those of the control group (p<0.05), and the difference was statistically significant (p<0.05), as depicted in fig. 2. Compared with the control group, the levels of VEGF, TNF-α, IL-1β and CA 125 in the experimental group were significantly higher than those in the control group (p<0.05), as shown in fig. 3. In the endometrial tissues of the experimental group, ID2 was positively correlated with VEGF, IL-1β and CA 125, while COX-2 was positively correlated with IL-1 and CA 125, with statistically significant differences (p<0.05), as shown in Table 2. The results showed that ID2 and COX-2 in endometrial tissue had high predictive value for clinical diagnosis of endometrial polyps, among which the sensitivity of ID2 was 87.32 %, specificity was 64.88 % and truncation value was 1.68. The sensitivity and specificity of COX-2 were 93.15 %, 84.33 %, and the truncation value was 2.03, as shown in Table 3. Endometrial polyps can cause chronic inflammatory changes in the internal environment of pelvic inflammatory disease, resulting in irregular menstruation or postmenopausal vaginal bleeding. Its occurrence and development can lead to a variety of adverse clinical outcomes [8] . Noninvasive investigations ID2 is a member of the differentiation /DNA binding inhibitor family. It forms heterodimer by binding with basic helix loop helix transcription factors, resulting in ID2 deletion, affecting its binding with target genes, and then inhibiting the expression of related genes [10] . Studies have shown that abnormal expression of ID2 can promote the proliferation and migration of cancer cells, which is closely related to cell adhesion, protein glycosylation, cell invasion and angiogenesis [12] . Due to the similarity between endometrial polyps and cancer, the expression of ID2 mRNA was detected by qRT-PCR in the experimental group and the control group, and the protein expression level was detected by western blot assay. The results showed that the expression of ID2 mRNA and protein in the experimental group were higher than those in the control group, suggesting that ID2 was up-regulated in endometrial tissues of endometrial polyps. The relationship between ID2 in the endometrial tissue of the experimental group and the clinical indicators of the patients was observed, and the results showed that ID2 in the endometrial tissue of the experimental group was positively correlated with VEGF, IL-1β and CA 125, among which the high expression of VEGF can increase the infiltration capacity of heterotopic gland cells [13] . IL-1β and CA 125 are highly expressed in patients with endometrial polyps, which can improve the adhesion ability of glandular cells by damaging mitochondria in endometrial polyps, and eventually damage the epithelial tissue of basement membrane seriously [14][15][16] . It is suggested that the expression of ID2 in endometrial tissue can be used as an index to predict the severity of endometrial polyps. Cyclooxygenase includes COX-1 and COX-2 isozymes. COX-1 catalyses the release of prostaglandin E2 (dinoprostone) by cells at rest, which mainly maintains cell self-stability. COX-2 was almost not expressed under normal conditions. When inflammation and cell proliferation and differentiation appeared, it was induced by cytokines, growth factors and tumor promoters, and its expression could be rapidly increased by 8-10 times. COX-2 is a rate-limiting enzyme of prostaglandin production, which is |
widely considered as one of the important inducible enzymes in inflammatory process [17] . Recent studies have found that the biological function of COX-2 can enhance the angiogenesis of subunit fibroblast growth factor and VEGF by stimulating the mitosis of human endothelial cells [18] . Based on this, the expression of COX-2 in endometrial tissues of patients with endometrial polyps and normal endometrial tissues was detected by qRT-such as transvaginal ultrasonography, with the use of 3-dimensional ultrasonography remain the mainstay of first-line investigation. At present, hysteroscopic surgery is used to remove endometrial polyps, but there is a high recurrence rate after Hysteroscopic resection is the most effective management for endometrial polyps and allows histologic assessment. Preutthipan et al. reported that the recurrence rate can reach 15 % [9] . Studies have shown that the changes of cell biology related factors are related to the pathogenesis of endometrial polyps, which can accelerate the infiltration process of glandular cells by affecting the proliferation and invasion of glandular cells in endometriosis tissue, and eventually lead to the continuous deterioration of the disease [10] . Therefore, it is of great significance to seek specific molecular markers for the diagnosis and treatment of endometrial polyps. Studies have shown that ID2 and COX-2 are related to the pathogenesis of ovarian endometrial tumor, and both of them are directly or indirectly involved in the process of cell migration, invasion and angiogenesis [11] . There are few studies on the expression of ID2 and COX-2 in patients with endometrial polyps. Therefore, this study mainly discusses the expression levels of ID2 and COX-2 in endometrial tissues of patients with endometrial polyps, which can provide theoretical basis for the pathogenesis of endometrial polyps. PCR and western blot assay in this study. The results illustrated that the expression of COX-2 in endometrial tissues of patients with endometrial polyps in the experimental group was higher than that in the control group, indicating that COX-2 was highly expressed in endometrial tissues of patients with endometrial polyps. Further studies showed that COX-2 was positively correlated with IL-1β and CA 125, which suggested that COX-2 might be involved in the occurrence and development of endometrial polyps and accelerate the deterioration of the disease through the interaction with serum related factors. The results of this study indicated that ID2 and COX-2 in endometrial tissue have high predictive value for clinical diagnosis of endometrial polyps, indicating that ID2 and COX-2 in endometrial tissue can be used as predictive factors for clinical diagnosis of endometrial polyps. At the same time, this study analyzed the related factors of endometrial polyps by logistic regression analysis. The results showed that ID2 and COX-2 in endometrial tissue were risk factors for patients with endometrial polyps, indicating that both of them can interact and promote the development of the disease. In summary, the high expression of ID2 and COX-2 in endometrial tissue is closely related to the occurrence and development of endometrial polyps. Both of them may be candidate specific target genes for endometrial polyps, and both of them can be used as markers for clinical diagnosis of endometrial polyps. The Future of Axial Spondyloarthritis Rehabilitation: Lessons Learned From COVID‐19 Supervised physical therapy and rehabilitation are vital for effective long‐term management of axial spondyloarthritis (SpA). However, the unprecedented year of 2020 and the COVID‐19 pandemic has prompted a drastic change in health care provision across all disease areas. In this review, we summarize changes that have been introduced to support rehabilitation in axial SpA during the pandemic and considerations for the future of axial SpA rehabilitation in the wake of COVID‐19. We have witnessed the launch of online virtual physical therapy and education, in addition to an emphasis on remote monitoring. We have been propelled into a new era of digital service provision; not only providing a temporary stop‐gap in treatment for some patients, but in the future, potentially allowing for a wider reach and provision of care and resilience of vital services. Unique collaboration between patients, health care professionals, and researchers will be key to fostering relationships and trust and facilitating wider evaluation and implementation of digital services at each stage in a patient’s journey, which is imperative for relieving pressure from health care providers. Despite the potential of such digital interventions, it is important to highlight the maintained critical need for face‐to‐face services, particularly for vulnerable patients or during diagnosis or a flare of symptoms. It is also vital that we remain vigilant regarding digital exclusion to avoid further widening of existing health inequalities. Optimization of digital infrastructure, staff skills, and digital education alongside promoting accessibility and engagement and building trust among communities will be vital as we enter this new age of blended in‐person and digital service provision. Introduction Physical therapy and rehabilitation are cornerstones of nonpharmacologic treatment for axial spondyloarthritis (SpA) and are critical for adequate long-term disease management (1,2). There is extensive evidence to suggest that physical activity is effective at reducing symptoms and disease activity in axial SpA, with a corresponding increase in spinal mobility, physical function, and cardiorespiratory fitness (1,(3)(4)(5)(6)(7)(8). As such, European treatment guidelines highlight the importance of a combination of nonpharmacologic and pharmacologic treatment modalities, including an emphasis on physical therapy, to optimize management of the condition (9). However, the most effective protocol for physical activity in axial SpA remains unclear (1,10). Recent evidence suggests that physical therapy for axial SpA should be prescribed based on the individual, while covering aerobic, flexibility, resistance, and neuromotor training (1). While unsupervised home-based exercises have been found to be efficacious for patients, supervised physical therapy has been suggested to be more effective (2,(11)(12)(13)(14). Furthermore, recent research has highlighted the potential paradoxical role of biomechanical stress and entheseal microdamage in the radiologic progression of axial SpA through potential development of tissue-specific inflammation and complex interactions between proinflammatory pathways, including the likely role of cytokines, growth factors, and | 45 tissue-resident cells (10). Therefore, evidence-based exercises provided in a one-to-one or group setting guided by a highly experienced, specialized physical therapist may be preferable initially, whereby the specialist can gauge the capabilities of the patient and recommend appropriate stretches and exercise accordingly on a case-by-case basis. This supervised mode of delivery by a specialist has also been identified as important to patients (13). The unprecedented year of 2020, however, and the COV-ID-19 pandemic have prompted a drastic change in health care provision across all disease areas. Patients have been unable to attend face-to-face appointments or supervised physical therapy, and a widening of existing gaps in health care have been highlighted (15). In the international REUMAVID study, of 1,707 patients with rheumatic musculoskeletal diseases (RMDs) surveyed from 15 European countries (47.5% of patients with an axial SpA diagnosis), 45.0% reported worsening health during the pandemic (16). In REUMAVID, patients also reported increased alcohol consumption, smoking, weight gain, and reduced physical activity, including an inability to continue rehabilitation exercises or physical therapy programs (17). Individuals participating in REUMAVID received poor access to care, 60.6% being unable to keep their rheumatologist appointment, 92.5% of which were canceled by their health care provider. More than one-half of participants perceived their health status to be "fair to very bad" and reported poor well-being as indicated by the World Health Organization Five Well-Being Index. Similar results have been reported in the UK specifically, where in a survey of health care professionals and patients conducted by the National Axial Spondyloarthritis Society (NASS), almost one-half of the patients reported a worsening of symptoms and deterioration of both general and mental health during lockdown (15). In the US, a study of 1,692 rheumatology patients from New York demonstrated that difficulties with medication access and flares were common during the peak of the pandemic (18). Furthermore, difficulty with medication access and COVID-related distress were both strongly associated with patient-reported flare and disease activity in this patient group. As described by the NASS in the UK, although the COV-ID-19 pandemic has highlighted existing gaps in service provision for patients with axial SpA, it has also accelerated change, with the introduction of virtual and remote consultations, including care for flares, and an increased interest in digital service provision and the importance of remote monitoring (15). Indeed, it has required a rapid adaptation of both patients and clinicians' practices to embrace new ways of working. The pandemic has also highlighted the need for imminent changes and prioritization of initiatives to revolutionize both the resilience and efficiency of our current health care systems to ultimately provide optimal support and the best possible care for patients with axial SpA (15). In the present article, we discuss changes that have been introduced to support rehabilitation in axial SpA during the pandemic and considerations for the future of axial SpA rehabilitation in the wake of COVID-19. Change in axial SpA rehabilitation services during COVID-19 At the Royal National Hospital for Rheumatic Diseases (RNHRD) in Bath, the unique 2-week inpatient physical therapy rehabilitation program has been integral to axial SpA care since the 1970s. The course provides individuals with the tools that they need to confidently self-manage their condition, placing an emphasis on education, self-management, physical therapy, and hydrotherapy, with input from a multidisciplinary team of physical therapists, a consultant rheumatologist, occupational therapist, counsellors, pharmacist, dietician, and health care assistants. There are no strict entry criteria for program referral. However, it is thought to be particularly beneficial for newly diagnosed patients, those in flare and who are struggling to manage their condition, postsurgery (e.g., following hip replacement), or to maximize outcomes of biologic therapy. To cater to differing levels of disease activity, function, and mobility, the program is delivered at 3 levels of intensity depending on spinal mobility (according to the Bath Ankylosing Spondylitis Metrology Index [BASMI]): fast (BASMI score 0-3), fast/ moderate (BASMI score [3][4][5], and moderate (BASMI score ≥5). Patients may attend the course more than once on an as-and when-appropriate basis. Significant short-and long-term improvements in disease activity, spinal mobility, and function have been observed following course attendance (19,20). The social element of the course, including meeting others with the condition, is also a critical element of the program's success. Participants have been known to forge long-lasting relationships following the course and to form critical support networks of mutual understanding. Although yet to be explored in detail in the context of the course, relatedness indeed forms 1 of the 3 basic psychological needs as detailed in self-determination theory. Self-determination theory proposes that when 3 innate basic psychological needs for autonomy, competence, and relatedness are fulfilled, positive outcomes are achieved, with these 3 factors suggested to be the most predictive and reliable mediators of motivation, engagement, and well-being (21). The impact of the course on such outcomes is currently being explored in ongoing analysis. During the pandemic, the importance of maintaining some form of supervised axial SpA rehabilitation delivery was recognized very early on at the RNHRD. As such, a group of highly skilled specialist physical therapists and rheumatologists, with input from a team of academics and behavioral scientists, was able to develop an online course to be delivered remotely via Zoom. While some services were obviously not available virtually (e.g., hydrotherapy), the core components of the course (education, self-management, and physical therapy) remained or could be reproduced, to an extent, online. Similarly, we have seen organizations such as the NASS migrate from in-person to online educational events, enabling a much wider reach for axial SpA education (22). The NASS has been hosting regular live online self-management sessions, with a wealth of legacy resources now available across its platforms, including recorded physical therapy sessions delivered live by specialist physical therapists. Introduction of remote data collection for axial SpA services At the RNHRD, not only are participants now able to attend the Bath axial SpA rehabilitation course from their own home, but standard patient-reported outcome measures collected pre-and post-course (and at each clinic appointment) have been migrated to an online system called Meridian. This includes measures such as disease activity (Bath Ankylosing Spondylitis Disease Activity Index [BASDAI]), function (Bath Ankylosing Spondylitis Functional Index [BASFI] patient global assessment), quality of life, fatigue (Functional Assessment of Chronic Illness Therapy), anxiety and depression (EuroQol 5-domain instrument), work productivity (Work Productivity and Activity Impairment questionnaire), and sleep (Jenkins sleep scale). Patients enter data via a unique online Meridian |
portal, and these data are then automatically integrated into the hospital system. This has facilitated the previously unforeseen efficiency of data collection both for research and for clinical use in axial SpA. Clinicians can now access individual patient-level graphical representations of, for example, disease activity (BASDAI) over time via Meridian during a clinic appointment, while approved researchers can access anonymized, aggregated data for patients who consented to the Bath Spondyloarthritis Research Biobank. More than 30 years' worth of paper records have also been digitized and integrated into Meridian. This includes additional measures such as spinal mobility (BASMI) and laboratory results such as C-reactive protein level. Furthermore, additional digitized information for research, such as coded Margolis Pain Diagrams (specifying regional or chronic widespread pain) and occurrence of significant life events, is available for a subset of ~200 patients. Although traditional patient-reported outcome measures are critical for understanding overall changes in disease activity and quality of life over time, they are subject to recall bias and may fail to capture a significant proportion of day-to-day disease information. In chronic, inflammatory conditions such as axial SpA where there may be fluctuating periods of disease activity and flare, these subtle daily changes in symptoms could be of critical importance for gaining a better understanding of the condition and for optimizing and personalizing treatments such as physical therapy. In 2017, the European Alliance of Associations for Rheumatology produced a "research roadmap to transform the lives of people with RMDs," often referred to as Rheuma-Map, which highlighted the need to better explore the impact of physical activity and lifestyle on the progression of axial SpA. Implementation of remote monitoring and digital technologies such as wearable devices and smartphones for granular, daily remote monitoring of symptoms and activity could be critical to meet this outlined need. Monitoring of lifestyle and physical activity and symptom data may also allow patients to gain a better understanding of their condition, while allowing them to gauge the level of physical activity that feels good for them and implement lifestyle changes accordingly. Since the start of the pandemic, we have seen an increased interest in remote monitoring both for research and clinical purposes. At the RNHRD, >350 patients are now registered with the RNHRD Project Nightingale study (www.proje ctnig hting ale.org), whereby individuals can use a smartphone app to track daily self-reported data in between clinical appointments, as well as before, during, and after course attendance. This includes variables such as pain, mood, stress, sleep, fatigue, flare, use of antiinflammatory drugs, and recommended exercise in addition to less explored variables such as menstrual cycle, caffeine intake, and screen time. The app can also be linked with an individual's wearable device if they have one to collect data on steps, heart rate, and sleep. Since September 2020, all patients invited to attend the virtual rehabilitation program have been invited to participate in Project Nightingale when referred to the course. This will form a larger piece of validation work to determine the capabilities of smartphone technologies to support both assessment of rehabilitation outcomes and potentially self-management. Indeed, enthusiastic patients at the RNHRD have expressed how Project Nightingale has helped them better self-manage and understand their disease while providing them motivation to exercise independently following intensive, supervised rehabilitation (23). However, until the platform has been evaluated scientifically, we cannot make firm recommendations for its use in health care. Considerations for future axial SpA rehabilitation delivery In terms of rehabilitation specifically, as suggested in feedback from RNHRD patients' post-virtual course, the future will likely involve a blended combination of in-person and online physical therapy with complementary remote data collection pre-and post-course. Online therapy could be implemented either as a "top up" between in-person appointments or as an alternative for patients who may not have the time to commit to an intensive rehabilitation program, such as the 2-week inpatient course delivered at Bath. Indeed, axial SpA often develops in the second or third decade of a patient's life, which is a critical time for establishing relationships and careers. Therefore, some individuals may prefer a shorter online course, whereby they can fit their initial education and physical therapy around their daily routine. This could also potentially be beneficial in terms of incorporating patients' habits into their usual environment, which may be trickier to implement and adjust to if they are coming from an immersive program away from day-to-day life. In Bath, while feedback on the axial SpA virtual rehabilitation program has been overwhelmingly positive, we need further | 47 robust evidence to ensure the acceptability, accessibility, and efficacy of digital rehabilitation interventions, and in particular, their comparative effectiveness alongside in-person rehabilitation. While there is some published evidence to suggest telerehabilitation as a suitable substitution for face-to-face interventions in chronic nonmalignant musculoskeletal pain, including some forms of arthritis, we should be cautious about generalizing these results to axial SpA specifically, and methodologic limitations have been described (e.g., small sample size, short follow-up) (24). Research has been conducted assessing the effectiveness of telerehabilitation in RMDs more broadly. These studies have found that real-time telerehabilitation can improve physical function and pain and is comparable to face-to-face intervention in terms of this improvement (25). A recent systematic review in rheumatoid arthritis identified 5 randomized controlled trials reporting a positive impact of telehealth interventions on factors such as disease activity, medication adherence, physical activity, and self-efficacy (26), although there was high heterogeneity in the interventions described. Similarly, a recent rapid review identified 14 systematic reviews exploring the effectiveness of telerehabilitation in musculoskeletal conditions, whereby, despite contradictory results, telerehabilitation could be comparable with in-person rehabilitation or better than no rehabilitation for conditions such as osteoarthritis, low back pain, and hip and knee replacement (27). These findings suggest that telerehabilitation may be effective in improving symptoms in RMDs. However, evidence is still limited, and there is an imperative need for better quality clinical trials and systematic reviews to provide sufficient evidence on efficacy and effectiveness (27). Analyses of the virtual rehabilitation program for axial SpA are currently ongoing in Bath, while similar web-based physical therapy interventions are also being tested for axial SpA in Glasgow (28). Input and considerations from physical therapists will also be critical when considering implementation of telerehabilitation for axial SpA. Key challenges currently identified are difficulties assessing patient mobility via Zoom or when observing and instructing patients, particularly while monitoring their performance of instructed exercises or if needing to provide discrete, individualized feedback during group activities (which is much easier in person, e.g., taking someone to one side to adjust their movement, and not so feasible in an online setting). Smaller groups of patients were also preferable with remote delivery, as it was harder to monitor multiple patients' movement via a screen. Over time, the format of the digital course can be tweaked based on further feedback from patients and the unique experience and expert knowledge of the contributing health care professionals. Economic evaluations could also be useful to determine the costeffectiveness of digital versus in-patient rehabilitation. Future wider implementation of digital rehabilitation for axial SpA could be critical in terms of relieving pressure from the health services, reducing wait times, and reducing travel burden for patients. However, we foresee that some form of in-person, supervised delivery will still be vital, particularly for those individuals who are newly diagnosed, fearful of movement, or who may have more severe disease and need closer supervision to prevent injury during exercise. Future studies to identify those patients who may most benefit from an in-person versus virtual rehabilitation program will be useful to refine these parameters, as will collaborations between patients, health care professionals, and researchers from multidisciplinary fields (biomechanics, human-computer interaction, health psychology) to assess the impact of such interventions and the best way to implement them. An initial in-person first-contact visit should also be considered to fully triage a patient's capabilities before prescription. The immersive element of the 2-week inpatient program may also have greater benefits in terms of improving or maintaining motivation for exercise in the long term. Spurring or maintaining motivation may be more difficult when being guided over a monitor versus an immersive experience with peers and physical therapists who are living and breathing the rehabilitation together in a socially supportive environment away from other commitments and worries in day-to-day life. Even in terms of the pandemic, many of us have experienced dull motivation and focus, described as languishing (29), when attempting to work from home all day behind a monitor; similar feelings could be experienced with the virtual course. It must therefore be ensured that we do not simply abandon invaluable in-person follow-up visits and rehabilitation completely, as certain aspects simply cannot be replicated virtually. Furthermore, loss of in-person follow-up or initiation of patient-initiated, in-person follow-up may be particularly detrimental to those patients who are more stoic in nature. Indeed, in a clinic, it is not unusual for a physician to notice a sign or symptom that has not otherwise been raised by a patient. In a recent service evaluation in Bath involving interviews with rheumatology patients and clinicians at the RNHRD, the importance of in-person interaction for reassurance was highlighted (both for patients, that they have been assessed holistically, and for staff, that they have not missed key signs of disease progression) to build patient trust in what was going to be a long-term therapeutic relationship. While digital interventions such as virtual rehabilitation potentially offer an array of benefits in terms of accessibility, relieving pressure on health services, and economic implications, digital exclusion is another key factor that must be considered. The term digital exclusion refers to those who lack the access, capacity, skills, motivation and/or trust to confidently go online (30). Indeed, digital exclusion exists at the intersection of multiple inequalities, whereby studies have shown that nonusers of the internet, devices, and online services are increasingly in vulnerable groups and may be older, less educated, and more likely to be unemployed, disabled, or socially isolated (31). In a recent study of 548 rheumatologists from 64 countries, although 82% of rheumatologists had switched to telehealth video during the pandemic, 17% estimated that approximately one-fourth of patients did not have access to telehealth video, especially those patients living below the poverty line (32). Respondents expressed a concern for these more socially and economically vulnerable patients, whereby wide implementation of telehealth could further widen existing health inequalities and differences in health literacy. During the pandemic, interruption of disease-modifying antirheumatic drugs without recommendation by a physician was also shown to be associated with lower socioeconomic status (33). The identification of vulnerable patients at risk of digital exclusion should be considered when beginning to implement telehealth. These patients should perhaps be prioritized for in-person, face-to-face health care delivery. In the context of rehabilitation, however, for individuals who may be more economically vulnerable and unable to take considerable time off work for an immersive rehabilitation program such as the 2-week course at the RNHRD, an online course to complete around other commitments may be preferable if provided with the appropriate resources and support. Other considerations are provision of digital education and optimization of health services, which will be critical for suitable implementation. In a recent survey of patients and clinicians, although >70% of patients and rheumatologists believed that digital health applications were useful in the management of RMDs, patients and rheumatologists respectively highlighted lack of information on suitable applications (58.5% of patients; 41.9% of rheumatologists) and poor usability (42.1% of patients) as key barriers to implementation (34). Rheumatologists also highlighted the importance of research evidence to support the implementation of such digital services. In the UK, a survey study of patients with axial SpA and rheumatologists during the pandemic highlighted some key areas requiring urgent attention, including upskilling of digital service provision (embedding good digital practice) and addressing gaps in digital infrastructure and staff skills (15). For example, in terms of patient coding, just 58% of health care professionals surveyed in the aforementioned study were able to identify the cohort of patients at high-risk of COVID-19 under their care in 2 weeks or less. Furthermore, 10% of respondents were still un able to identify high-risk patients 4 months after shielding guidance was first issued by the UK government. Coding challenges were often the cause of these delays and the huge variation |
in times to identify high-risk patients. Interestingly, similar coding concerns throughout other rheumatology services prompted in Leeds the development of a strategy to communicate with patients online and enable them to self-assess their COVID-19 risk (35). The authors described the flexibility and agility of the NHS in the UK for introducing drastic change rapidly when pressured on such an unprecedented scale, in addition to describing the encouraging level of engagement of patients when it came to self-assessment and self-education. Conclusions Physical therapy and rehabilitation are key in the management of axial SpA. Despite the challenges faced, the pandemic has also fostered an environment for adaptation and development of creative solutions to provide continued care. Indeed, all services have been tested and as such have been propelled into a new era of digital service provision. We have witnessed the launch of online virtual physical therapy and education in addition to an emphasis on remote monitoring. Not only has this provided a temporary stop-gap in treatment for some patients, but in the future, it may allow for a wider reach and provision of care and resilience of vital services. Unique collaboration between patients, health care professionals, and researchers will be key to fostering relationships and trust and facilitating wider evaluation and implementation of digital services at each stage in a patient's journey (from diagnosis to rehabilitation and long-term condition management), which are imperative to relieve pressure from health care providers. Despite the potential of such digital interventions, it is important to highlight the maintained critical need for face-to-face services, particularly during diagnosis or during a flare of symptoms. We must ensure that digital interventions are evaluated rigorously before widespread implementation in clinical practice. It is also vital that we remain vigilant regarding digital exclusion and that we avoid a further widening of existing health inequalities. Optimization of digital infrastructure, staff skills, and digital education alongside promoting accessibility, engagement, and building trust among communities will be vital as we enter this new age of blended in-person and digital service provision. Are we measuring loneliness in the same way in men and women in the general population and in the older population? Two studies of measurement equivalence Background High levels of loneliness are associated with negative health outcomes and there are several different types of interventions targeted at reducing feelings of loneliness. It is therefore important to accurately measure loneliness. A key unresolved debate in the conceptualisation and measurement of loneliness is whether it has a unidimensional or multidimensional structure. The aim of this study was to examine the dimensional structure of the widely used UCLA Loneliness Scale and establish whether this factorial structure is equivalent in men and women. Methods and sample Two online UK-based samples were recruited using Prolific. The participants in Study 1 were 492 adults, selected to be nationally representative by age and gender, whilst the participants in Study 2 were 290 older adults aged over 64. In both studies, participants completed the UCLA Loneliness Scale (Version 3) as part of a larger project. Results In both studies, the best fitting model was one with three factors corresponding to ‘Isolation,’ ‘Relational Connectedness,’ and ‘Collective Connectedness.’ A unidimensional single factor model was a substantially worse fit in both studies. In both studies, there were no meaningful differences between men and women in any of the three factors, suggesting measurement invariance across genders. Conclusion These results are consistent with previous research in supporting a multidimensional, three factor structure to the UCLA scale, rather than a unidimensional structure. Further, the measurement invariance across genders suggests that the UCLA scale can be used to compare levels of loneliness across men and women. Overall the results suggest that loneliness has different facets and thus future research should consider treating the UCLA loneliness scale as a multidimensional scale, or using other scales which are designed to measure the different aspects of loneliness. Introduction Throughout their evolutionary history, humans have lived in social groups and depended on forming long-term relationships with others for survival [1,2]. Thus, humans have a basic and universal need to form strong, stable interpersonal relationships with others-a 'need to belong' [3]. When this need is unmet and people feel disconnected from others, this lack of meaningful social relationships has a profound impact on physical and mental health [4]. Loneliness is defined as an unpleasant subjective state arising from a mismatch between the quantity and quality of social relationships we have and those we would like to have [5]. A large body of research has demonstrated that high levels of loneliness are associated with negative health outcomes in relation to both morbidity and mortality (reviews in [6][7][8][9][10][11]). Loneliness also has a key place on the social and political agenda in countries such as the United Kingdom [12], and the pandemic has further exacerbated the need for policy intervention on this front [13]. It is thus important that we can reliably measure loneliness, in order to accurately measure its prevalence over time, in different parts of the population and to evaluate whether interventions to combat loneliness are effective [14,15]. Over the past five decades, many scales have been developed to measure loneliness, including: the Differential Loneliness Scale [16], the Loneliness Rating Scale [17], the De Jong-Gierveld Loneliness scale [18], and the Social and Emotional Loneliness Scale for Adults (SELSA, [19]). One of the most commonly used measures is the UCLA Loneliness Scale, which has appeared in first [20], second [21] and third [22] versions, and its short form adaptations (e.g., [23][24][25]). The UK Office for National Statistics has recommended that future UK national surveys of loneliness use three items from the UCLA scale [26]. The scale has been translated into many languages (e.g., Russian: [27]) and validated in many countries (e.g., Italy: [28]; Zimbabwe: [29]). UCLA loneliness factor structure: One, two, or three factors? A key unresolved debate in the conceptualisation and measurement of loneliness is whether it has a unidimensional or multidimensional structure [20][21][22][30][31][32]. From its inception, the UCLA Loneliness Scale was argued to tap into a unidimensional construct [20][21][22], with deficits in a variety of relationships producing the same underlying state. Indeed, many studies have found evidence for a unidimensional structure (e.g., [33,34]), or for a unidimensional structure with a subsidiary factor accounting for methodological effects due to wording [35]. Some such studies have used student participants, for example, a sample of over 650 South African students supported a one-factor solution [34]. Yet a one-factor solution is also supported in other samples, such as adolescents (e.g., [36]). Other studies (e.g., [37, 38]) do not conduct factor analyses to establish the factor structure of the UCLA Loneliness Scale, but instead, treat the scale as defining a unitary construct. A synthesis of eighty studies using the UCLA Loneliness Scale as a unidimensional construct revealed an estimate of Cronbach's α of . 87 [39]. The size of this estimate depended on four factors: article type (focussing on measurement or not), scale standard deviation, whether a social support network was measured, and sample composition. Interestingly, in terms of sample composition, adolescent samples tended to yield lower reliabilities than non-adolescent samples. However, whether a sample was composed of older adults or not did not influence the reliability estimate. From its inception, however, the unidimensional nature of the UCLA loneliness scale has been challenged on both theoretical and statistical grounds (e.g., [40,41]). Studies have argued for two (e.g., [29]), three [42] or even four or five factor solutions (e.g., [23,[43][44][45]). There are only a minority of papers reporting four and five factor models respectively, so we restrict our review of the literature to two and three factor models. Whilst some argue loneliness is a unitary state [21,22], other researchers propose that loneliness has two key components: emotional and social isolation (e.g., [32,46]). Thus, Weiss [32,41] argued that the need for the emotional security provided by a single 'attachment figure' is distinct from the need to be connected to a broader social network, and people can be dissatisfied with one aspect (e.g., lack of a long term romantic partner) without being dissatisfied with the other (e.g., having a good network of friends). In line with this proposition, Zakahi and colleagues [47] argued for a two factor solution. Similarly to Zakahi and colleagues [47], Wilson and colleagues [29] recovered a two-dimensional factor structure in a sample from Zimbabwe. These two factors were labelled as "social other" and "intimate other." However, Knight and colleagues [48], while recovering a similar factor structure, attributed this to the framing of items as positive or negative. Accordingly, Russell [22] revised the scale (UCLA Loneliness Scale Version 3) and suggested a two-dimensional structure. Using this Version 3 of the UCLA Scale, some studies have found support for the two-factor structure. For example, Ausín and colleagues [49] found support for a two-factor model in a large sample (n > 400) of adults aged 65 or over. However, other research has argued for a three-factor structure for the UCLA loneliness scale (e.g., [42, 50, 51]). One such three-factor structure is Russell's model [22], which allocates all items to one factor, and then additionally allocates each item to either a "negative items" factor or to a "positive items" factor. This structure has been supported using confirmatory factor analyses in relation to the UCLA Scale Version 3 [22] in two Turkish samples [52], and in a sample of 300 healthy Iranian adults [53]. Similarly, a sample of over 500 respondents from Argentina [54] supported this model using the second version of the UCLA [21]. Given the range of studies supporting the Russell model [22] model, we attempt to fit this model to our data, below. Other three-factor solutions have also been put forward in relation to the second and third versions of the UCLA, and these more conventionally allocate each item to one factor exclusively. These solutions include McWhirter et al.'s model [50] which named the factors "Intimate Others," "Social Others," and "Affiliative Environment"; Boffo and colleagues [28] who named the factors "Isolation," "Relational Connectedness," and "'Trait Loneliness"; and Sancho and colleagues who named the factors "Isolation," "Trait Loneliness," and "Social Connectedness" [55]. Most notably, however, the work by Hawkley and colleagues [40] argued for the following three factors: "Isolation," reflecting feelings of rejection and aloneness; "Relational Connectedness," corresponding to feelings of familiarity; and "Collective Connectedness," which deals with feelings of group identification. This model has received support from large-sample studies, including one of over 1,400 Irish adolescents [56], and another that relied on student samples (n > 500) [57]. Contrastingly, a study using participants from Indonesia, Germany, and the United States, did not find the three factor solution to be a good fit in absolute terms [31], although a three factor solution did perform slightly better than a one or two factor solution. Given this range of support, we test this latter three-factor model [40] in our analysis below, together with the unidimensional model as proposed by Russell Gender differences Research exploring gender differences in loneliness presents mixed findings, with some research suggesting that women report more loneliness than men (e.g., [58,59]), some research indicating that men report more loneliness than women (e.g., [22, 60-63]), and yet other research not finding a robust gender difference (e.g., [64]). In addition, much of this research has tended to rely on scales with a unidimensional approach to loneliness, rather than a multidimensional approach (but see [65]). It is important to establish that the scales used yield the same factorial structure for men and women to enable us to make valid comparisons between men's and women's experiences of loneliness. Such testing across genders is regularly carried out in connection with the development of psychometric instruments [66,67]. Researchers have previously tested the measurement invariance across genders of various loneliness scales, such as the De Jong Gierveld loneliness scale [68] and the Loneliness and Aloneness Scale for Children and Adolescents [69]. Similarly, some studies have examined whether the UCLA Loneliness Scale has the same structure across men and women. Allen and colleagues used a short 7-item version of the UCLA Loneliness Scale [70] and found support for a unidimensional structure, which did not meaningfully differ between men and women. Hawkley and colleagues found support for a three-factor structure in both genders [40], using the 1980 [21] version of the UCLA. Finally, a study that |
was based on a sample of over 1,000 teachers in Canada and that used the second version of the full 20-item UCLA scale found support for a three-factor structure that was invariant between men and women [71]. To our knowledge, however, measurement invariance based on gender has not been established in a representative sample of the population, nor in a sample of older adults for the UCLA Version 3. Our research contributes to the literature by examining measurement invariance of the UCLA Version 3 loneliness scale [22] in two separate samples: a UK-based adult online sample where participant age and gender were nationally representative (Sample 1), and an online sample of UK-based older adults (Sample 2). We examine one, two and three factor models via confirmatory factor analyses, and examine if we can establish whether this factorial structure is equivalent in men and women across our two different samples. Methods Both studies were advertised on Prolific, a crowd sourcing website for scientific studies [72]. In a comparison of online platforms for recruiting participants, participants from Prolific failed fewer attention checks, showed lower levels of dishonest behaviour and were more naive in relation to common psychological research materials, as compared to participants from Amazon MTurk [73]. Potential participants are recruited to Prolific primarily via word-ofmouth (including on social media), following an original recruitment drive when Prolific was founded in 2014, which recruited via social media, flyer distribution at university campuses, and a paid refer-a-friend scheme [74]. Once signed up to the Profiific platform, participants have the opportunity to take part in research in exchange for monetary payment. Sample 1 (nationally representative adults) We used the Prolific settings to request a sample of 500 UK-based adults whose age and gender were nationally representative. We obtained 498 complete responses (self-reported gender: 257 women, 236 men, 2 neither, 3 non-disclosures). Three participants did not provide their age, but for the remaining participants, the ages ranged from 19 to 82 years (M = 49.15, SD = 15.53). 289 out of 498 participants indicated that they had completed at least a Bachelor level degree. Participants who did not report their gender as male or female were excluded from the further analyses, given that we wished to examine measurement equivalence between men and women. One participant did not complete all items and was excluded from the Structural Equation Models (SEM). Thus, the final sample consisted of 492 participants. Participants were paid £3.35 for completing the survey. Sample 2 (older adults) We used the Prolific settings to request a sample of UK-based adults aged 65 years old or older. 290 participants (179 women and 111 men) completed the survey. One participant did not report their age, and one reported an improbable value (66,123). As we did not include age as a factor in any of the analysis, these two participants were retained in the final sample. For the participants who provided their ages, the range was from 64 to 86 years (M = 69.04, SD = 3.88). 146 out of 290 participants indicated that they had completed at least a Bachelor level degree. Participants were paid £2 for completing the survey. Procedure For Sample 1 (nationally representative adults), the UCLA Loneliness Scale was administered as part of a larger online egocentric social network study [75,76]. The full study protocol was preregistered on the Open Science Framework (OSF). In Sample 2 (older adults) the UCLA Loneliness Scale was collected as part of a larger study where participants completed multiple scales on health, psychological well-being, and friendships. The protocol is registered on the OSF. Both studies were approved by the Northumbria University Psychology Department Ethics Committee, and participants recorded their consent within the online survey. Materials Loneliness. In both studies, participants completed the UCLA Loneliness Scale Version 3 [22]. This scale contains 20 items, where 11 of these refer to positively valenced feelings such as feeling part of a group of friends, and 9 of these refer to negatively valenced feelings such as feeling left out, and are conventionally reverse-scored. Participants are asked to respond on a 4-point scale, anchored at 1 = Never and 4 = Always. In version 2 of the UCLA Loneliness Scale [21] a different endpoint was used (4 = Often). It is unclear why this change happened, and correspondingly some papers have used the older anchor (e.g., [56,71]). In our study, Sample 1 used the version 2 anchors (never / often) from [21], and Sample 2 used the version 3 anchors (never / always) from [22]. The negatively valenced items were not reverse-scored for SEM, as this is not necessary. This just implies that there will be negative correlations between a negatively valenced factor and (an)other factor(s) in two and three factor solutions, rather than a positive one (if we had reverse-scored). Data analysis Our analyses consist of Confirmatory Factor Analyses (CFA) and group invariance testing [77]. While there is an active debate about sample sizes in CFA and the use of heuristics to determine sample sizes (e.g., [78,79]), we note that our sample exceeds a common heuristic of N = 200 (e.g., [80]), and is in line with other studies (e.g., [53]). All the analyses were conducted in R 4.0.2 [81] and various R packages (e.g., [82][83][84]). Among these packages, we used 'lavaan' [85] to perform CFA, following the one-factor solution proposed by [22], the two-factor solution proposed by [29], and the three-factor solution proposed by [40] (see Table 1 and [56]). We also attempted Russell's [22's] bifactor model (as supported by [52-54]-see Introduction), where all items load on to a general loneliness factor, and in addition each item is allocated to a "positive items" or a "negative items" factor, but this did not give rise to a reliable solution, and is not discussed further in this paper. Next, we examined measurement invariance [67,[86][87][88]. The Open Science Framework provides free public access to all data, code, and analyses, as well as further analyses and fit metrics not reported in text (e.g., Standardized Root Mean Square Residual, SRMR). Tables 1 and 2 show the descriptive statistics for all items for Sample 1 (nationally representative adults) and Sample 2 (older adults), respectively. These are the raw scores, i.e. not reversescored. Descriptive statistics When using the scale as a unitary construct, the Cronbach αs for the respective samples were . 95 Measurement invariance modelling showed that the model that produced the lowest RMSEA = .089 (Table 3; 'Mean,' Model 5) was the one where the factor loadings, intercepts, residual variances and means were constrained to be equal across groups. There is some loss of fit in terms of CFI moving from configural to mean invariance, but it falls within the suggested -.01 change [89] or -.02 change [90]. We, therefore, conclude that the factor means can be considered equal between groups: i.e. there are no measurable mean differences between men and women as regards these three latent constructs. Fig 1 shows the resulting models for men and women. The labels are based on the model by Hawkley and colleagues [40]. The associations between the three latent constructs are also similar between men and women. Measurement invariance modelling showed that the model where the factor loadings, intercepts, residual variances and means are constrained to be equal across groups produced the lowest RMSEA = .095 (Table 4; 'Mean,' Model 5). There is some loss of fit in terms of CFI PLOS ONE Loneliness equivalence moving from configural to mean invariance; it is close to the suggested -.01 change [89], but below the suggested -.02 change [90]. While the -.02 criterion is more liberal, on the whole Table 4 leads us to conclude that the factor means can be considered equal between groups, i.e. there are no measurable differences between men and women on these three latent constructs. Fig 2 shows the resulting models for men and women in Sample 2 (older adults). The associations between the three constructs are also similar, as in Sample 1 (nationally representative adults). The only exception is that the association between Collective Relatedness and Isolation is somewhat lower in men (r = -.57) than in women (r = -.74) but the 95% confidence intervals still comfortably overlap (-.73 to -.41 and -.83 to -.66, respectively). Discussion In this study, we investigated the factorial structure of the widely used UCLA Loneliness Scale for men and women in two different online samples: an adult UK sample that was nationally representative by age and gender, and a sample of UK older adults. In both samples, a model with three factors proved the best fit. Authors have reported slightly differing ways of allocating the 20 items of the UCLA Loneliness Scale (either the second or third version) to a threefactor structure model, and such solutions have been reported in several studies, including large samples from Argentina, Iran, Ireland, Spain, and Turkey ([52, 55, 56], see Introduction; but see [31] for contrasting results). We did not seek to test each of the slightly differing threefactor models in relation to our data to avoid over-fitting, but instead focussed on the popular Hawkley et al. [40] model (e.g., [56]). We also examined Russell's [22] bifactor structure composed of three factors, but this model was not identified, see OSF. Our findings support the notion that the UCLA Loneliness Scale reflects loneliness as a multidimensional rather than a unidimensional structure, with three factors corresponding to Isolation (feelings of aloneness and rejection), Relational Connectedness (feelings of familiarity, closeness and support) and Collective Connectedness (feeling part of groups that provide a sense of identity and belonging), as suggested by [40,56]. Prolonged periods of loneliness are consistently associated with poorer health outcomes [10], and as such tackling loneliness can be part of a country's political and social agenda [91]. There are several different types of interventions to reduce loneliness [14,15], including social prescribing approaches which are designed to provide a non-medical referral option for General Practitioner doctors to improve health and well-being [92]. In designing and evaluating these interventions, it is important to accurately measure the different facets of loneliness. For example, interventions that promote membership of community groups [92] may be more effective in providing a broader range of social connections (Collective Relatedness), as compared to emotionally close relationships (Relational Connectedness). As many interventions use the UCLA Loneliness Scale as an outcome measure [14,15], if treated as a unitary scale this may miss these more subtle changes in different aspects of loneliness as a result of the intervention. Future work on loneliness should therefore consider treating the UCLA measure as a multidimensional measure, or use the other scales specifically designed to measure the different facets of loneliness (e.g., [93]). The multidimensional nature of loneliness might reflect its differing etiologies, manifestations, and consequences, and thus might in turn be reflected across different questionnaire measures. As an example, the abbreviated Social and Emotional Loneliness Scale for Adults (SELSA) is also reported to have a three-factor structure [65]. Where the UCLA Loneliness Scale focuses perhaps more on the experience of loneliness, the SELSA focuses on its sources, and as such its subscales separate romantic, family, and social loneliness; for instance, an individual could have a strong relationship with a partner (romantic loneliness) and family (family loneliness), but not a strong friendship group (social loneliness). Previous research has shown relationships between people's scores on the SELSA subscales and the UCLA [19,93,94], and we might anticipate further that the scores on the three UCLA factors would differentially predict scores on the SELSA subscales. For instance, we might predict particular overlap between the SELSA's "social loneliness" and the UCLA's "Collective Connectedness," which incorporates items such as feeling part of a group of friends and feeling like you have a lot in common with the people around you. That is, loneliness, or the lack thereof, may depend on having both close and affiliative ties [32]. In addition to examining the overall factor structure of the UCLA scale, we also examined measurement invariance based on gender. We found support for the 'means' model in our analysis. This suggests that there are no meaningful differences between men and women in any of the three constructs. Now that we have established that the UCLA yields the same factorial structure for men and women, this enables researchers to make valid comparisons between men's and |
women's experiences of loneliness. Similarly, we note that the factor loadings, correlations, fit indices, and structure are similar across our two samples (nationally representative adults, and older adults), in line with [39]. Our samples were sourced from adults in the United Kingdom, and relied upon people who were enrolled on Prolific, a crowd-sourcing website for scientific studies. Thus, although our 'nationally representative' sample in Study 1 was representative in terms of age and gender, we would not expect them to be fully nationally representative of the United Kingdom, nor of course of other countries. Equally, adults aged 65 years old or older are less likely than other age groups to use the internet [95], and yet our 'older adults' sample all necessarily used the internet in order to access Prolific. It is important to be wary of assuming invariance in psychological variables across all countries and cultures [96,97]. Having said this, we do not have serious concerns that our findings would be, prima facie, non-replicable in other samples. This is in part because other researchers report similar findings on the factor structure of the UCLA Loneliness scale in countries outside the UK (e.g., [52,55,56], but see [31]), and in part because of the affiliative and sociality requirements that are part of human nature [3], and that are indeed seen in related species [98]. In conclusion, we find support for a multidimensional (three-factor) structure to the UCLA Loneliness Scale, in a nationally-representative UK sample by age and gender, and in a UK sample of older adults. This multidimensional structure is consistent with previous research (e.g., [40,56]), and is in line with the differing etiologies of loneliness (e.g., [32]). We suggest that our findings are broadly generalisable to other samples given the inherent sociality of humans as a species, although of course this awaits testing. We found no meaningful differences between men and women in any of the three constructs, something which supports the usage of the UCLA Loneliness Scale to compare men's and women's experiences of loneliness, and which may help us further tackle this important predictor of individual wellbeing (e.g., [10]). Future studies of loneliness should consider treating the UCLA Loneliness Scale as a multidimensional rather than unidimensional measure, or use other scales which are designed to measure the different facets of loneliness (e.g., [93] Tolerance to coxibs in patients with intolerance to non-steroidal anti-inflammatory drugs (NSAIDs): a systematic structured review of the literature Adverse events triggered by non-steroidal anti-inflammatory drugs (NSAIDs) are among the most common drug-related intolerance reactions in medicine; they are possibly related to inhibition of cyclooxygenase-1. Coxibs, preferentially inhibiting cyclooxygenase-2, may therefore represent safe alternatives in patients with NSAID intolerance. We reviewed the literature in a systematic and structured manner to identify and evaluate studies on the tolerance of coxibs in patients with NSAID intolerance. We searched MEDLINE (1966–2006), the COCHRANE LIBRARY (4th Issue 2006) and EMBASE (1966–2006) up to December 9, 2006, and analysed all publications included using a predefined evaluation sheet. Symptoms and severity of adverse events to coxibs were analysed based on all articles comprising such information. Subsequently, the probability for adverse events triggered by coxibs was determined on analyses of double-blind prospective trials only. Among 3,304 patients with NSAID intolerance, 119 adverse events occurred under coxib medication. All adverse events, except two, have been allergic/urticarial in nature; none was lethal, but two were graded as life-threatening (grade 4). The two non-allergic adverse events were described as a grade 1 upper respiratory tract haemorrhage, and a grade 1 gastrointestinal symptom, respectively. In 13 double-blind prospective studies comprising a total of 591 patients with NSAID intolerance, only 13 adverse reactions to coxib provocations were observed. The triggering coxibs were rofecoxib (2/286), celecoxib (6/208), etoricoxib (4/56), and valdecoxib (1/41). This review documents the good tolerability of coxibs in patients with NSAID intolerance, for whom access to this class of drugs for short-term treatment of pain and inflammation is advantageous. Introduction Non-steroidal anti-inXammatory drugs (NSAIDs) are the most commonly used therapeutics in the outpatient management of pain and inXammation in a wide spectrum of diseases. Their primary mode of action is the blockade of prostaglandin synthesis by cyclooxygenases (COX): Constitutively expressed COX-1 is involved in fundamental mechanisms of homeostasis, whereas the inducible COX-2 mediates inXammation. Therapeutic eVects of NSAIDs are primarily related to their ability to inhibit COX-2, whereas some of their most frequent adverse eVects may be caused by COX-1 inhibition (Fig. 1). In contrast to most "classic" NSAIDs which block both isoforms, the so-called coxibs preferentially inhibit COX-2. This may result in better tolerability, namely reduction of gastrointestinal side eVects [29,85]. Respiratory and cutaneous adverse events triggered by NSAIDs are among the most common drug-related intolerance reactions in medicine. Typically, these manifest as asthma attacks or urticaria. Pathogenesis of these symptoms seems to be related to COX-1 inhibition [76]. Therefore, the hypothesis was put forward that coxibs may safely be used in patients with known NSAID intolerance. However, serious intolerance reactions to coxibs have also been observed, thus cautioning too euphoric expectations [71]. We therefore reviewed the relevant literature in a systematic and structured approach for evidence of coxib tolerance in patients with NSAID intolerance. Search strategy We searched MEDLINE (1966MEDLINE ( -2006, the COCHRANE LIBRARY (4th Issue 2006) and EMBASE up to December 9, 2006. The following search terms were used: rofecoxib OR celecoxib OR valdecoxib OR parecoxib OR etoricoxib, combined with hypersensitivity OR intolerance. Those coxibs used in veterinary medicine (deracoxib, tiracoxib and cimicoxib) were not included in the literature search. No language or publication restriction was predeWned. All publications reporting individual patients with NSAID intolerance and subsequent exposure to a COX-2 Inhibitor were identiWed and the reference lists of these articles were hand-searched for further publications. If articles could not be retrieved in full text, a copy was requested from the corresponding author and/or journal editor. Inclusion criteria Each publication was appraised for inclusion in a stepwise approach (Fig. 2). Only publications describing individual patients and providing a rational medication scheme as well as a suYciently speciWc outcome report were included in this review. In a Wrst step, evaluation focused on clinical symptoms and severity of adverse events in patients with NSAID intolerance. Therefore, all articles on this topic were included for this analysis. Subsequently, the probability for adverse events was analysed based exclusively on publications of double-blind prospective trials. Data extraction and synthesis All articles were analysed using a predeWned evaluation sheet. Uncertainties were resolved by consensus decisions among the investigators. Data synthesis was qualitative and descriptive. The Common Terminology Criteria for Adverse Events version 3.0 (CTCAE) was used to categorize adverse events. Results We identiWed 230 publications on coxibs and/or NSAID intolerance. Hundred and forty-six references not focusing on individual patients with NSAID intolerance were excluded. Unclear medication schemes or outcome reports led to exclusion of two articles. Eighty-four publications were evaluated for severity and type of adverse reactions to coxibs. Thirteen publications on double-blind studies were used to determine the probability of adverse reactions to coxibs. Discussion This systematic review documents the low probability of allergic/pseudo-allergic reactions induced by coxibs in patients with NSAID intolerance. To our knowledge, this is the Wrst comprehensive analysis of data published on this topic. We have searched all three major medical databases available, namely MEDLINE, COCHRANE LIBRARY, and EMBASE using very broad and general search terms. Following identiWcation of relevant publications, these were evaluated by means of a pre-deWned evaluation form. The data available are described in the form of a structured review [19]. NSAIDs are among the most commonly prescribed therapeutics in the world. Although generally considered safe, their wide and frequent use results in these drugs being among the most common causes of drug-related intolerance reactions. This may at least in part be due to their nonselective inhibition of both cyclooxygenase isoforms. In line with this hypothesis, NSAIDs characterized by pronounced COX-1 inhibition bear a high risk to trigger asthma attacks in patients with aspirin-sensitive asthma bronchiale, whereas preferentially COX-2 inhibiting NSA-IDS are better tolerated by these patients [36]. Our analysis of published studies on this issue further supports this notion, as only 13 of 591 NSAID-intolerant patients showed adverse reactions upon provocation with coxibs in double-blind clinical studies; all of these were grade 3 or milder. Still, relatively selective COX-2 inhibitors have been identiWed as triggers of serious intolerance reactions [71]. This implies that our current understanding of NSAID-triggered intolerance is still imperfect, and its path-ogenesis cannot be reduced to cyclooxygenase-mediated eVects alone, but needs to take into account clinically relevant additional NSAID-mediated eVects such as secretion of leukotrienes from mast cells and other leukocytes. It has been suggested that coxibs may confer an elevated risk for acute myocardial infarction and sudden cardiac death, namely after long-term therapy [25]. As a reaction, several coxibs are no longer available despite a recommendation by the participants of an expert meeting organized by the Food and Drug Administration to grant further prescription of rofecoxib, celecoxib and valdecoxib in the US [59]. On the other hand, substantial evidence described here points towards a good tolerability of coxibs in patients with NSAID intolerance. Given the wide use of NSAIDs in the short-term treatment of trivial signs and symptoms, the availability of coxibs for these indications would be advantageous for this relevant subpopulation of patients, since this type of application is unlikely to increase cardiac risk. Urinary D-glucaric acid excretion in the Seveso area, polluted by tetrachloro-dibenzo-p-dioxin (TCDD): five years of experience. On July 10, 1976, an explosion in a factory in Seveso, Italy, located 30 km north of Milan, producing trichlorophenol caused the release of TCDD-containing compounds in the surrounding area. Since extremely small doses of TCDD have been shown to induce hepatic microsomal enzymes in animals, urinary D-glucaric acid excretion (a measurable index of enzyme induction), has been investigated in Seveso in adults and children 6 to 8 years old, in order to clarify whether levels of environmental exposure to TCDD were sufficient to produce an induction in man. Urine samples were collected from 1976 to 1981. As a control group, people living in Cannero (a nonindustrialized village on lake Magiore), in Busto Arsizio (a small industrial town near Milan) and in Lentate (a noncontaminated zone near Seveso) were chosen. In the first period of collection, children with chloracne (which is considered to be a characteristic manifestation of intoxication with chlorinated products) showed significantly increased levels of D-glucaric acid excretion compared to children without chloracne living in the same zone. As far as chronic exposure is concerned, up to 3 years after the accident both adults and children living in the Seveso area showed a statistically significant enhancement of D-glucaric acid elimination compared to the control groups. This study demonstrates that adults and children living in the polluted zones had an increased activity of hepatic microsomal enzymes for some years, since, although the urinary excretion of D-glucaric acid is only an indirect measure of enzyme activity, studies in man have indicated that it is, however, sensitive and quantitative.(ABSTRACT TRUNCATED AT 250 WORDS) Shortly after noon (12:40) on July 10, 1976, in Seveso, Italy (located 30 km north of Milan), a reaction at the Icmesa chemical plant producing trichlorophenol got out of control, and an explosion spread substances to the surrounding area one of which was tetrachloro dibenzop-dioxin (TCDD). TCDD is one of the most toxic synthetic chemicals yet discovered, and there are large differences in species susceptibility. Many organs and systems were demonstrated to be affected after TCDD intoxication (1)(2)(3). A few days after the accident in Seveso, particularly in the area near the Icmesa plant, some people, especially children, developed skin lesions of various kinds, including chloracne, which is considered to be a characteristic, though nonspecific, manifestation of intoxication with chlorinated products (4)(5)(6)(7)(8). The contaminated area was divided into three zones, depending on the TCDD concentration of the soil ( Table 1). The inhabitants of zone A were evacuated 15 days after the explosion to areas without detectable TCDD, whereas those of zones B and R continued to live in their homes. In order to clarify whether there was an increased number of persons with induced hepatic enzymes among the Seveso population, we assayed urinary D-glucaric acid, which is considered to be a convenient index, because although indirect, it is |
simple, sensitive, reproducible and correlates quantitatively with the degree of enzyme induction (27)(28)(29)(30)(31)(32)(33)(34)(35)(36)(37)(38). Various groups of people (adults and especially children, with ages between 6 and 8 years) were studied in different periods; the results of the follow-up of the population, as far as urinary D-glucaric acid excretion is concerned, is presented in this paper. Preliminary results concerning only the children were presented elsewhere (39). Materials and Methods The case material is summarized in Table 2. As far as the selection of the subjects is concerned, some considerations are necessary. Adults 1978 For one week, urine of people living in zone B who were called to undergo blood tests, according to the monitoring plan for Seveso, were collected. The compliance with the request was about 80%. As a control group we collected urine samples from adults living in Cannero (a small town on Lake Maggiore, 130 km from Milan, in a nonindustrialized area, uncontaminated by environmental toxin). Half ofthe population of this town was selected (450 people) and was asked to undergo a clinical visit, plus urine and blood examina tions. The compliance with our request was about 90%. One-third of the collected samples, chosen at random, were utilized for D-glucaric acid estimation. Both in Seveso and in Cannero, people affected by overt renal or liver disease or taking liver enzyme-inducing agents were excluded. Children 1976 This retrospective study was concerned with children known to have been in contact with the poison because they presented with chloracne. The levels of D-glucaric acid in their urine samples collected in the period August-December 1976 and immediately frozen at -30°C were compared with those of children of the same zone without skin lesions, whose urine was collected and frozen in the same period. Sixty names were drawn: 30 children with skin lesions and 30 without skin lesions. Of these it was possible to recover from thousands of test tubes frozen at -30°C 31 samples, i.e., 14 subjects (9 girls, 5 boys) with chloracne and 17 subjects (10 girls, 7 boys) without any skin lesions. Children 1979 Urine samples of children living in zone B and in zone A (evacuees) were collected. In some of these children, D-glucaric acid excretion was also estimated one year later. As control groups we chose children living in Cannero and Busto Arsizio (a town about 30 km from Milan, in a highly industrialized area but not contaminated with TCDD). Children 1981 In May 1981 we compared urinary excretion of Dglucaric acid in children of zone A, zone B, two groups living in zone R and as a control, children living in Lentate, a zone near Seveso not contaminated by TCDD. All the urine samples were collected, after informed consent, from pupils in the first three elementary school grades; the compliance with our request was about 90%. The urine samples were assayed for D-glucaric acid by the method of Simmons et al. (40) and corrected for variations in urine concentrations from the creatinine values. Because in children a portion of creatine is not converted into creatinine, we expressed the urinary Dglucaric acid values in relation to the sum of the creatine converted into creatinine and the creatinine itself. For the assay of urinary creatinine we used the method of Jaff6 without deproteinization, as standardized by Roche Products (41). Creatine was converted into creatinine by acidification and boiling of the urine for 4 hr (40). The values found are then expressed in ,imole saccharo-1,4-lactone/g urinary creatinine. Statistical Methods The relatively small sizes of the samples analyzed were not suitable for a proper analysis of the distributions of the values of D-glucaric acid in urine samples. However, these distributions appeared to be positively skewed and rather irregular; moreover large differences in their dispersion parameters were observed. Therefore, it seemed sensible to process data according to distribution-free methods. Results obtained in 1976 regarding children with and without chloracne were compared by using Kruskal- Wallis (43) nonparametric analysis of variance. The same method was followed to analyze data collected in 1979 from exposed subjects (Seveso: zones A and B) and controls (Cannero and Busto Arsizio). Dunn's techniques (44) were adopted to perform multiple comparisons. The difference between the results obtained in 1979 and in 1980 was tested according to Wilcoxon (45), while the association between the same results was measured by Spearman's rank correlation index (46). Results The D-glucaric acid excretion in urines of adults collected in 1978 was higher in people living in the Seveso area than in those of Cannero, the difference being statistically significant (median 27.1 ,umole/g of creatinine in Seveso and 19.8 in Cannero; lower quartile, 11.9 and 14.8, respectively; upper quartile, 42.1 and 29.3, respectively; p < 0.05) (Fig. 1). The distribution of the values related to sex was similar, while significant dependence of D-glucaric acid levels on age in both centers was found: in all ages exposed subjects showed excretion significantly higher than controls (p < 0.05) (Fig. 2). two groups were therefore used as a single control group in the subsequent comparison (86 subjects overall, median 21.8, lower quartile 16, upper quartile 25.5). Of the children exposed to TCDD, D-glucaric acid concentration in zone A (evacuees) was similar to that in controls (median 23.2, lower quartile 15, upper quartile 27.2); however, in zone B the excretion was significantly higher (median 26 ,umole/g of creatinine, lower quartile 19.1, upper quartile 31.5, p < 0.05). The comparison between D-glucaric acid values observed in 1979 and 1980 in urines of children living in zone B showed significantly lower levels in 1980 ( Fig. 4 and Table 3). Finally Figure 6 and Table 4 show the D-glucaric acid excretion in urine samples collected in children in May 1981. A trend toward an increased excretion in children living in zone B and in zone R (Polo) was observed; however it does not reach a significant difference with respect to the controls (0.10 > p >0.05). Discussion Our study could be divided into two parts. First Dglucaric acid excretion in children with and without chloracne who lived in the most contaminated zone before the evacuation was evaluated. The urine samples were collected in the period August-December 1976, and this could be considered an acute exposure. Chloracneic children had significantly higher urinary D-glucaric acid concentrations than nonchloracneic children from the same zone. It is likely that the Icmesa disaster, responsible for the skin lesion, was also the cause of this change. As far as the responsible substance is concerned, it is only possible to suppose that among the substances released that day (glycols, caustic soda, trichlorophenol and TCDD), the latter is the most indicated, since only this has been demonstrated to be a potent inducer (9)(10)(11)(12)(13)(14)(15)(16)(17)(18)(19)(20)(21)(22)(23)(24)(25)(26). In the second part of the study, estimation of D-glucaric acid excretion was performed in adults and children living in the polluted zones; since the research was carried out 2, 3, 4 and 5 years after the explosion, this could be considered a chronic exposure. Up until 1979 adults and children living in the Seveso area were demonstrated to have significantly higher values of D-glucaric acid excretion, in comparison to those observed in controls (inhabitants of Cannero and Busto Arsizio). Because in children factors such as alcohol, contraceptives or other drugs commonly known to act as enzyme inducers can be ruled out as a cause for the increase in excretion, it is reasonable to surmise that in this, as in the other study, TCDD is responsible for the phenomenon. It is hardly conceivable that the differences observed could have been due to different lifestyles or eating habits and as for nonspecific environmental pollution or the like, since the urinary D-glucaric acid excretion values in the children of Cannero, a lakeside village, and in those of Busto Arsizio, a small but industrialized and heavily polluted town, were practically identical. In zone B, TCDD (but not the other compounds released in the toxic cloud, which were quickly broken down) has been present in the soil for years, and children could have easily come into contact with it. The similar elimination of D-glucaric acid values between children of zone A and controls can be explained by the fact that children from zone A were moved away from the polluted zones a few days after the accident, and it is likely that 3 years are sufficient for complete elimination of any poison absorbed in July 1976. From 1980 the elimination of D-glucaric acid was observed to return toward the normal value. The lower level observed in 1980 compared with 1979 in the majority of children living in zone B could be explained either by major precautions taken by the children whose parents were advised of the results of the test and of its significance, or by a reduction of TCDD present in the soil (as a consequence of the natural degradation and the decontamination). The almost similar level of D-glucaric acid observed in 1981 in children living in different zones of Seveso and in Lentate (uncontaminated town near the Icmesa plant) supports the latter hypothesis. In conclusion, determination of urinary D-glucaric acid, a sensitive index for hepatic microsomal enzyme induction, enabled us to prove the existence of induction in subjects living in the TCDD-polluted zones. Our findings also show that D-glucaric acid estimation could be a useful method for detecting metabolic changes or adaptation, such as enzyme induction, which precede liver alteration caused by poisons like TCDD, even in the absence of clinical disease. Can antler stage represent an activity driver in axis deer Axis axis? The aim of this study was to determine the seasonal activity patterns and asynchrony between different antler stages in male axis deer from the Mediterranean island of Rab in Croatia using camera traps. Nine cameras with an infrared motion detection system were used to track animal activity over a 12-month period, 24 h per day. Stags were divided into two categories of antler development: regeneration stage and hard antler stage. The frequency of detection of each category in the photographs allowed us to investigate seasonal activity patterns. To describe the seasonal activity pattern in each category, we fitted the segmented linear regression and predicted that the ratio of monthly activity of stags in the two antler categories would interchange regularly during the research period. Over the 12-month study period, 36 862 photographs were analysed. A significant difference in frequency was found between the two antler categories (p < 0.01), with a consistently greater presence of hard-antlered stags. The highest frequency of detection of both antler categories was found in the winter season, and the lowest in spring. The segmented linear regression clearly distinguished three break points in April, June and December in the hard antler stage, with a significant difference in activity pattern among the months for each slope. On the other hand, no significant difference was found for the regeneration stage. Therefore, the expected proportional interchange in the number of stags in the two antler stages throughout the year did not occur. This study revealed that the Mediterranean axis deer population showed a unique activity pattern, where antler stage may act as a possible driver regulating stag movements. The antler cycle is closely linked to the testicular cycle and the associated seasonal fluctuations in androgen secretion in male deer (Price et al. 2005, Ramesh et al. 2013. Antler growth occurs during the period of low testosterone concentration, whereas antler mineralization and velvet shedding from antlers are spurred by increasing testosterone levels and corresponding increase in mating activity (Lincoln 1992, Bubenik 2006. Casting of antlers is triggered by a marked decline in testosterone levels, which can occur at various rates, depending on the sex ratio within the respective population. Antlers are the only body appendages in mammals to undergo full regeneration (Goss 1983). Growth of antlers in cervid species from temperate areas is a seasonal event occurring every year in close synchronicity with photoperiod (Goss 1983, Sempéré 1990). Antlers are cast and re-grown from the top of the pedicles, permanent frontal protuberances covered with normal scalp skin (Kierdorf and Kierdorf 2011). The age of first development of the pedicles depends on nutrition status and body mass (Sempéré 1990, Price andAllen 2004). The onset of antler formation on top of the pedicles is indicated by the appearance of velvet, a specialized integument rich in |
sebaceous glands and with hair lacking arrector pili muscles (Goss 1983, Bubenik 1993. After velvet shedding, antlers appear as completely bony structures (Goss 1983, Ramesh et al. 2013. The size and complexity of antlers increase with increasing body size and age of individuals (Bubenik 1990). Accordingly, antlers play a primary role in defining social structure in male deer, as they are used as display weapons in intra-sexual combat (Clutton-Brock 1982), and their size correlates positively with dominance (Lincoln 1972). In fact, upon casting antlers, dominant stags in red deer Cervus elaphus are almost challenged by one or more subordinate stags and often displaced from the group. A similar pattern was shown to occur also in moose Alces alces, elk Cervus canadensis and caribou Rangifer tarandus in a 'dummy antler' experiment (Bubenik 1983). Seasonal fluctuations of climatic and environmental conditions are the key drivers that shape the reproductive cycle (including mating activity) in cervids (Bubenik et al. 1986, Foley et al. 2015. Unlike boreal deer species, which are strictly seasonal breeders, axis deer Axis axis reproductive cycle is closely related to the species' adaptability to a wide range of environments (Willard and Randel 2002) and their tropical origin with comparatively minor annual changes in photoperiod (Loudon and Curlewis 1988). Although photoperiod is the dominant environmental cue that entrains the seasonal antler cycle in deer from temperate regions (i.e. seasonally breeding deer; Feldhamer and McShea 2012), in tropical (axis) deer, Loudon and Curlewis (1988), suggested that photoperiod is not involved in controlling the antler cycle. Reproductive cycle and the antler cycle of tropical deer species such as axis deer appears to be aseasonal in most of their native habitat (tropics) and depends on the availability of food, which is determined by local climatic conditions (Fraser-Stewart 1985). Depending on the environmental conditions, the mean interval from casting to casting in axis deer from tropical regions ranges from 45 to 60 weeks, with the duration of the hard antler stage from 23 to 37 weeks and 17 to 30 weeks from the casting of the old antler to the cleaning of velvet (Loudon and Curlewis 1988, Bubenik et al. 1991, Lincoln et al. 1998). However, in subtropical regions axis deer may exhibit seasonality with relatively good synchronization of the antler cycle (Bubenik et al. 1991). As a result, mating activities that influence monthly and seasonal changes in axis deer group size (Raman 1997) show great variations among populations across various ecological contexts (Schaller 1967, Šprem andZachos 2019), thus affecting activity patterns of male axis deer individuals in different antler development stages. Although seasonal dynamics of mating activity in axis deer remain unexplored, there are few studies describing activity patterns of this species (Schaller 1967, Dave 2008, Centore et al. 2018. By taking into account the above-mentioned, antler stage can allow to predict phase of reproductive cycle. Axis deer originating from the Indian subcontinent have been successfully introduced to Croatia twice, though the exact provenance of the introduced individuals is unknown (Kusak andKrapinec 2010, Frković 2014). In 1974, seven individuals were released on the Kalifront Peninsula of the island of Rab and preserved ever since for hunting purposes (Tomljanović 2016). The axis deer from the Adriatic Islands in Croatia are the only free-ranging European populations (for more details see Šprem and Zachos 2019). Taking advantage of the atypical (i.e. non-native) ecological conditions of the extant axis deer populations in Croatia, we aim to investigate the seasonal asynchrony in the antler cycle and seasonal activity patterns of male axis individuals in different antler stages using camera traps. Owing to the scarce data available on the antler phenology in axis deer in different ecological contexts (Schaller 1967, Mishra and Wemmer 1987, Waring 1996, if the seasonal activity pattern of individuals in different stages is related to the antler cycle, we expect that: 1) seasonal activity pattern will vary between individuals in different antler stages, with the highest occurrence of stags in the hard antler stage expected in the expressed rutting periods: January-February, July-August and October-November (Šprem et al. 2008); 2) the ratio of individuals in different antler stages will be unbalanced throughout the year and should regularly interchange between consecutive months, i.e. the higher occurrence of stags in the hard antler stage during rutting periods will be replaced by a higher occurrence of stags in the regeneration stage in other periods. Study area The study was conducted on the northwest side of the island of Rab, in the northern Adriatic Sea in Croatia ( Fig. 1). Owing to the climate and geological conditions, scrublands and woodlands of Euro-Mediterranean vegetation dominate. The area is a forest ecosystem composed of carbonate rocks with predominant holm oak Quercus ilex, and manna ash Fraxinus ornus cover (Ugarković and Ugarković 2013), at elevations between sea level and 94 m. Forest cover prevails (94%; Tomljanović 2016), with stem density (733-9581 ha -1 ) in the study area (Oršanić et al. 2011). According to the Köppen classification, the study area belongs to Cfa climate group of humid subtropical climates (Seletković et al. 2011), with an average annual temperature of 15.3°C (mean temperature: 23.6°C in summer; 8.0°C in winter). The vicinity of a continental montane region strongly affects the study area, characterized by the maritime rainfall regime (annual precipitation 1102 mm). However, the area is considered one of the sunniest places in Europe, with an average of 2479 sunshine hours per year, coupled with an extremely dry period from mid-June to August. Highly nutritional low vegetation forms (various grasses and herbaceous plants) develop in early spring (April and May). Large herbivores in the study area during the 2015/2016 season were axis deer (n = 78; 9.2 ind. 100 ha -1 ) and European mouflon (n = 103; 12.2 ind. 100 ha -1 ) (Centore et al. 2018). According to the estimates in the present management plan, the sex ratio ranges between 0.86 and 1 in favour of females, and the age structure is: 19% juveniles, 18% yearlings and 63% adults (Tomljanović 2016). Mating and calving in axis deer occurs year round and there is no strictly defined mating season, therefore stags can be found in various stages of antler development regardless of season (Šprem et al. 2008). Most of the time, adult males are separated from the herds consisting of females with calves (3-10 individuals) or young males (3-5 individuals). During the rut, males rarely fight, whereas subadults and males with growing antlers avoid the rutting areas (Šprem and Zachos 2019). Data collection Camera traps are increasingly used to observe the behaviour patterns of individuals, estimate population abundance and identify sex and age where possible (Anile and Devillard 2018). Moreover, daily or seasonal activity patterns of wild animals can be investigated using metadata information (time and date of caption, Frey et al. 2017) alongside the frequency of detections (Green et al. 2015, Bu et al. 2016). Data were obtained using camera-traps (Wildlife Innovation, Spypoint models: Iron, PRO-X, HD-12, Moultrie models: MFH-DGS-D55IR, MCG-12589 and Primos Truth Cam 35 infrared detection game cameras) between 1 April 2015 and 31 March 2016. A systematic net grid of 1000 × 1000 m was set and cameras placed at each grid intersection, covering an area of approximately 90 ha. For this purpose, nine cameras were placed from 50 to 100 cm above ground level, with a 5-min delay between consecutive photos and a default focus distance of 5 m. Cameras were active 24 h per day and checked twice a month to collect photographs and check battery status. Data analysis To obtain actual information on antler status, all photos from the nine cameras were pooled, since the population range is restricted by island size and habitat usage is evenly distributed across the study area. Only photographs clearly showing stags were used. Photographs were categorized by month (from January to December). Since the new set of antlers begins to grow almost immediately after the old ones are shed (Schaller 1967), we combined stags without antlers and those with antlers in velvet to reduce the bias when making conclusions about the activity patterns of stags in the antler growth period. Therefore, the antler stages were divided into two categories: regeneration stage and hard antler stage, taking into account the average duration of these categories in other studies (Loudon andCurlewis 1988, Lincoln et al. 1998). All male age classes were pooled, since it was impossible to properly estimate age based on the antler development (Gee et al. 2013, Ikeda et al. 2013. Frequency of detection of each category in the photographs allowed us to investigate the seasonal activity patterns (Green et al. 2015, Bu et al. 2016. Accordingly, to examine the differences between the frequency of detection of antler categories on camera traps, the Student's t-test was used, where only presence (1) or absence (0) of each category in the same photograph was recorded. The presence of stags without antlers allowed for an estimation of the peak of casting periods. To detect the finer scale of antler transformation, based on irregular growth patterns of axis deer antlers, each individual captured in the same photograph was treated as a single observation and associated to one of the two categories. To describe the activity pattern of each category (i.e. including the total number of individuals counted, belonging to one of the two categories), we fitted segmented linear regression models with month as a predictor variable for each antler category, using the R package 'segmented' (Muggeo 2008). The package offers facilities to estimate and summarize generalized linear models with segmented relationships without any limit on the number of segmented variables or the number of change points. The activity pattern of each category was compared using breakpoints and slopes. Finally, we predicted that the ratio between hard antler stage and regeneration stage will regularly interchange during the study period. All statistical analyses were performed using R ver. 3.3.2 (< www.r-project.org >) in RStudio, ver. 1.1.423 (RStudio Team 2016). All parameter estimates were reported with standard errors (SE). Results During the 12-month study period, 122 082 JPEG photographs were collected, and 36 862 were selected according to the above criteria for analysis. We found significant differences in the frequency of detection of axis deer stags between two antler categories (regeneration stage versus hard antler stage; t-test: t = −6.3208, df = 11, p < 0.001). The highest frequency of detection was found for both antler categories in the colder part of the year (i.e. November, December and January) (Fig. 2), particularly in November, with 2694 counts of stags in the hard antler stage and 1304 counts of stags in the regeneration stage. In contrast, the lowest frequency of detection was recorded in May for stags in the hard antler stage (406 counts), and in June for stags in the regeneration stage (148 counts) (Fig. 2). Furthermore, the highest peak of antler casting occurred in November (546), with two smaller peaks in April (117 counts) and August (241 counts) (Fig. 3). With reference to the total number of recordings in each antler category, the segmented linear regression in the hard antler stage clearly distinguished three break points (Fig. 4), in April (estimate = 4.53), June (estimate = 6.79) and December (estimate = 10.21). The first decrease in the frequency of hard antler stage occurred from December to May (slope 1: estimate = −802.3, SE = 46.612, t = −17.212, p < 0.01), followed by an increase from May to July (slope 2: estimate = 808.0, SE = 147.4, t = 5.4817, p < 0.01) and from October to December (slope 4: estimate = 756.0, SE = 147.4, t = 5.128, p < 0.01). No significant change in frequency was found from July to October (slope 3: p > 0.05) (Fig. 4). Comparatively, in the regeneration stage, segmented linear regression distinguished only two break points in May (estimate = 5.83) and July (estimate = 7.02), without a significant difference in frequency between months for each slope (Fig. 4). Discussion The aim of this study was to provide insight into the antler phenology of a Mediterranean island population of freerange axis deer. To the best of our knowledge, this unique study provides a first review on seasonal activity patterns of axis deer stags in different antler stages in Europe. The camera trap survey distinguished three periods of antler casting, |
with the highest portion of stags without antlers occurring in November (Fig. 3), thus suggesting a conspicuous reproduction shift with an increasing percentage of stags in the regeneration stage from September to November (Fig. 5). In deer, the testosterone cycle is responsible for regulating the antler cycle and synchronizing it with the reproductive season (Rolf and Fischer 1996). As recorded by Šprem et al. (2008), matings of axis deer in the Mediterranean region can be observed in January-February, July-August and October-November. Hence, looking at the frequency of detection of stags in the hard antler stage in this study, the results indicate a similar mating pattern in line with those proposed in our hypotheses, though it is evident that mating occurs year round. In Nepal, for example, the peak period of hard antlers occurs once a year in May and coincides with the peak period of the rut when the sexes mixed (Mishra and Wemmer 1987), which is remarkably different than found in this study. Information about the activity patterns of axis deer antlered males is lacking, and findings concerning antler cycles vary. Observations conducted in India, for exam- ple, showed contrasting patterns, with a high frequency of hard-antlered males (82.2%, Miura 1981) or a prevalence of velvet stags from February to April (Waring 1996), December to February (Mishra and Wemmer 1987), and the occurrence of a transition phase from velvet to hard antlers in March (Schaller 1967). The occurrence of stags in both antler stages was lowest in the early spring, suggesting an apparent decrease in the level of activity. Similarly, a recent study by Centore et al. (2018) recorded the lowest activity of axis deer (all age and sex categories) in spring on the island of Rab. This may relate to the peak of productivity of low vegetation layers, especially grasses in April and May, which constrained foraging activity to pastures. This constraint supports previous findings that showed how grasslands were used preferentially by axis deer in lowland Nepal in February, March and April (Moe and Wegge 1994). Unlike continental areas, the Mediterranean region has unfavourable conditions for ungulates in summer due to high temperatures and drought. Consequently, in this period animals are forced to move and increase their search for food. In late June and early July (after the grazing period in spring), axis deer on the island of Rab preferred one-year shoots of woody species (Krapinec et al. 2000). Indeed, the present study found a gradual increase in the frequency of both categories, starting in June and reaching its peak in November, suggesting a marked change in the activity pattern across seasons (Fig. 2). In particular, the results indicate that antler stage may be an important driver of deer activity during the year, as confirmed by the consistently lower frequency of detection of stags in the antler regeneration stage. Consequently, the expected proportional interchange in the number of stags in different antler stages throughout the year did not occur (Fig. 5). In contrast, in their native habitat in Nepal, axis deer showed a clear interchange during the season in the percentage of stags in the velvet and hard antler stages (Mishra and Wemmer 1987). We acknowledge that the duration of the period between the hard antler stage and the regeneration stage should be accounted for in the analysis, as it showed conspicuous variations among individuals in previous studies (23-37 weeks hard antlers, 17-30 weeks regeneration phase; Loudon and Curlewis 1988). Unfortunately, such information is lacking for this island population, though the higher frequency of detection of individuals in the hard antler stage during the season could be partially explained by a longer duration of this stage. On the other hand, sexual competition in deer species during the breeding season can cause dispersal among males (Jarnemo 2011) and it is possible that the hard-antlered axis stags showed increased mobility while searching for females in heat, which was may be reflected in the higher frequency of captions. However, breeding dispersal is primarily driven by strong competition for mating and overt aggression (cf. Rosenberry et al. 2001), which are not expressed in the sexual behavioural patterns of axis deer (Mishra andWemmer 1987, Šprem andZachos 2019). Growing antlers are very sensitive and the slightest pressure may cause pain, bleeding and even disfigurement of the (still) soft bone surface (Caton 1877, Whitehead 1972. Consequently, deer in velvet tend to avoid contact between antlers and other objects (Whitehead 1972). In turn, velvet stags living in dense forest habitats may show a decrease in activity to avoid contact with dense vegetation (Tomljanović 2016). Moreover, since the period of rut is not fixed, and mating occurs year round, it is possible that stags in the regeneration stage avoid clashing antlers with hard-antlered stags, thereby restricting their movement. In conclusion, the present study showed that the Mediterranean island population of axis deer display a unique activity pattern, with antler stage as a possible driver regulating stag movements. Certainly, restricted habitat use raised the possibility of direct contact between individuals and shaped their movement patterns. However, the effect of environmental conditions and related food availability should not be neglected as one of the foremost activity drivers. To support our claims, future analysis of movement and habitat use from telemetry data in relation to habitat conditions is required. Utilization, Contributions, and Perceptions of Paid Home Care Workers Among Households in New York State Abstract Background and Objectives While family caregivers have traditionally provided care for older adults with chronic conditions and disabilities, the demand for paid home care workers has increased in the last decade. Although typically thought to assist with personal care, emerging data suggest that paid home care workers assist with a wider scope of care. However, the extent and quality of the care they deliver remains poorly understood. Research Design and Methods Using the Empire State Poll, a telephone-based cross-sectional survey of 800 adults in New York State, we characterized the types of care that paid home care workers provided and the perceived value of that care. Results Of 800 participants surveyed, 274 reported that they or an immediate family member received care from a paid home care worker (34.3%). Of these, the majority (73.9%) reported that paid home care workers provided emotional and/or medical care, in addition to personal care. In adjusted models, providing emotional and medical care (compared to personal care alone) was associated with nearly a twofold greater perception of importance and experience by the care recipients. Discussion and Implications Our findings provide additional data on how paid home care workers contribute to patient care, from the perspective of the care recipient(s). The type of care provided is associated with varying magnitudes of perceived quality. Although limited to New York, these findings have implications for paid home care workers’ training and compensation. Future studies are warranted to investigate the specific factors that mediate the association between types of care provided and their perceived value. Older adults as well as those with chronic conditions and disabilities commonly rely on caregivers for help at home with their personal and medical care. While family caregivers have traditionally provided the backbone of this care, over the last decade, the demand for paid home care workers has increased substantially (1)(2)(3). There are currently 2.3 million paid home care workers, and the field is projected to grow by more than 1.5 million by 2030 (2). Paid home care workers, which include home health aides and attendants as well as personal care aides, provide personal assistance and health care support to older adults and people with chronic conditions and disabilities in their homes (2,4). Although typically thought to provide only personal care, emerging data suggest that paid home care workers assist with a wider scope of care activities beyond activities of daily living and instrumental activities of daily living. Specifically, qualitative studies have found that paid home care workers, especially home health aides, often assist with medical care, such as the maintenance and management of chronic conditions and acute complaints (5)(6)(7). Additionally, patients and family caregivers have reported that paid home care workers provide emotional support and companionship to them or their loved ones (8,9). Despite these contributions, few studies have quantitatively documented the types of care that paid home care workers provide on a day-to-day basis and the association between the types of care they provide and its quality. A better understanding of how they contribute to patient care and their perceived value is important because this workforce is increasingly caring for medically complex adults who are choosing to age in place, and because payment models in home care are shifting toward reimbursement based on quality metrics. If paid home care workers are indeed providing more medical or emotional care than previously thought, increased awareness of their contributions by the medical community and policymakers is needed, and formalized training will need to appropriately reflect these contributions. Furthermore, evidence supporting these expanded roles has clear implications for paid home care workers' professional development and payment. Studies have consistently found that paid home care workers feel marginalized, unappreciated, undervalued, and poorly integrated into the health care team (10)(11)(12). Without data that speak to the value of their contributions to patient care, these issues are unlikely to change. To that end, the objectives of this study were threefold: (a) to determine the utilization of paid caregiving among households in New York State, (b) to determine the types of caregiving activities in which paid home care workers contributed, from the perspective of the household, and (c) to examine the association between the types of care provided by paid home care workers and the value of that care, as measured by its importance and overall experience from the perspective of the care recipient(s). Study Design We conducted a cross-sectional study analyzing data collected as part of Cornell Survey Research Institute's 2020 Empire State Poll (13). Institutional review board approval was obtained in January 2020 by the Survey Research Institute (SRI) at Cornell University. Survey Development The Empire State Poll is an annual telephone survey of randomly selected adults aged 18 and older residing in New York State. Each year the survey includes a core set of questions about community, government, and economic issues as well as demographic data. In addition to the core questions, investigators may submit questions to be included in the annual survey. Submitted questions are reviewed by the SRI and pilot-tested with a sample of 25 individuals who do not participate in the final survey. Investigators are given feedback regarding any issues encountered during pilot testing and offered the opportunity to modify their questions for clarity and comprehensibility prior to launch of the final survey, which our team did. In 2020, the final survey contained 62 interview questions, including 13 submitted by this research team. Survey Administration The Empire State Poll was administered from January 24, 2020 to March 15, 2020. A total of 18 007 phone records were contacted to achieve the final sample size of 800 individuals (unique households; Supplementary Material I). The 2020 Empire State Poll had a cooperation rate (defined as the number of completed surveys divided by the number of potential interviews) of 25.6% and a response rate (defined as the number of completed surveys divided by the total eligible sample) of 14.6%. The average interview length was 18 min. The Cornell SRI used a dual-frame random digit dial sampling of landlines and cell phones in New York State. An additional sampling frame was added to recruit individuals with out-of-state cell phone numbers. Additional oversampling was utilized to ensure representation of Black and Hispanic individuals throughout New York State, as well as upstate and downstate geographic regions sampling was conducted in proportion to population totals. Data Collection Research assistants used a standard Computer-Assisted Telephoning Interviewing software package called CASES to call prospective participants. This software was used to collect and store sample information and measure call outcomes. All data were de-identified by removing confidential information and stored in a private and secure database on SRI's network, monitored by the Cornell Information Technology team. Survey Questions About Paid Caregiving Informed by The Convoy of Care Model, which theorizes the intersection of unpaid and paid caregiving and its impact on caregiver and care recipient outcomes, we developed a total of 13 questions that assessed public |
utilization of and attitudes toward unpaid and paid caregiving for personal and health-related needs; of those, 8 questions are relevant to this study (14). We asked all participants: (1) Have you or an immediate family member ever had a paid home care worker? A "yes" response triggered 7 additional questions that began with: Our next few questions are about your most recent experience with a paid home care worker: (2) Main Exposure The types of caregiving activities in which paid home care workers were engaged (personal care, medical care, emotional support, other), as assessed with the 1-item survey question above (#4). Main Outcome The main outcome of the study was the value that paid home care workers' care provided. We measured value with 2 separate constructs, one pertaining to (a) the importance of having a paid home care worker (as measured by the 1-item survey question above, #5) and a second pertaining to (b) the overall experience of having a paid home care worker (as measured by the 1-item survey question above, #6). Data Analysis We assessed the distribution of demographic and caregiving characteristics among participants who reported that they or their family members received care from a paid home care worker. We next assessed differences in demographic and caregiving characteristics across the types of care that the paid home care worker provided (personal, emotional, medical) using chi-square, Fisher's exact, and analysis of variance tests. We used robust Poisson regression to analyze the association between type of care provided and importance of the paid home care worker, adjusting for demographic/caregiving characteristics. We used Poisson regression because with frequent outcomes, prevalence ratios (PRs) are less likely to be overestimated and are easier to interpret than odds ratios (17). Characteristics were selected for the model based on previous literature (7) or a significant bivariate association (p < .05) with type of care. We repeated this process for the remaining outcome of experience with a paid home care worker. Covariates with missing data included method of payment for paid care (5%), rural residence (5%), hour of paid care (4%), and race (3%). Sample weights were applied to account for the sampling design and permit generalization to the entire state. The sample size of 800 produces a margin of error of ±3.5 percentage points. Data management and modeling were conducted using STATA version 16 (StataCorp LLC, College Station, TX). Results Of the 800 ESP participants, 274 reported that they or an immediate family member (household) had received care from a paid home care worker in the past (34.3%). For the main analyses, and given the study objectives, we excluded participants (n = 6) who did not provide information on the types of care that the paid home care worker provided, and 29 participants who reported that the paid caregiver did not assist with personal care. A total of 239 participants comprised the final analytic cohort (Supplementary Material II). Characteristics of the Sample Characteristics of the 239 participants who reported that they or their family member received care from an HHW are given in Table 1. They had a mean age of 48.6 (95% confidence interval [CI]: 45.8-51.3), 58.7% were female, 73.4% were White, 15.6% were Black, 33% had income less than $50 000, and 86.5% lived in urban areas of New York. With respect to paying for the paid home care worker, nearly one quarter (23.8%) reported using Medicare insurance, 18.8% Medicaid, and 40.6% reported paying privately without insurance. They reported receiving paid care for a mean of 34.3 (95% CI: 28.5-40.1) hours a week and 16% and 46.5% reported that the paid home care worker was "somewhat" and "very" connected to the person they cared for, respectively. With respect to the importance of the paid home care worker to overall care received, nearly three quarters of the participants (74.1%) reported that the HHW was "very important." When asked about the overall experience of having a paid home care worker, 31.1% and 54.5% reported that the experience was "somewhat" and "very" positive, respectively. Types of Care Performed by Paid Home Care Workers Of the 239 participants, 26.1% reported that the paid home care worker assisted with personal care only, 19.6% reported that the paid home care worker assisted with personal and emotional care only, 22.5% reported that the paid home care worker assisted with personal and medical care only, and 31.9% reported that the paid home care worker assisted with personal, emotional, and medical care ( Figure 1). How care was paid for differed significantly by the type of care rendered; for example, personal care only was more likely to be paid for privately (without insurance), whereas personal and medical care only were more likely to be paid for by Medicare (Supplementary Material III). Additionally, the number of hours of care differed significantly by the type of care rendered; participants who reported receiving personal, emotional, and medical care had more hours of paid help on average (46.9 hours/week), compared to those who received personal care only (25.4 hours/week), personal care and emotional care only (36.8 hours/week), and personal and medical care only (24.2 hours/week), p < .04 (Table 1). Note: CI = confidence interval. *Participant who responded to the survey may or may not be the one receiving care. Experience of Having a Paid Home Care Worker The association between the care performed by the paid home care worker and how positive the experience of having a paid home care worker was varied by the type of care the paid home care worker performed (Table 3). Having a paid home care worker who provided both personal and emotional care was associated with an 80% ; however, this association did not achieve statistical significance. Likewise, in adjusted models, the association between tasks performed by a paid home care worker and having a very positive experience remained positive and significant when emotional care was rendered in addition to either personal care or in addition to personal and medical care; adjusted models accounted for the demographics of the participant and household, as well as the number of hours in which care was provided and payment method for care. Discussion In this cross-sectional and representative survey of community-dwelling adults in New York State, a third of participants reported that their household received care from a paid home care worker. Although a quarter of participants reported that the worker provided personal care alone, the majority of participants reported that paid home care workers provided emotional and/or medical care, in addition to personal care, to them or their immediate family member. In general, satisfaction with the care provided by paid home care workers was high, as was the overall experience of having a paid home care worker. In multivariable models, we found that the type(s) of care provided by the paid home care worker was associated with different perceptions of value by the care recipient. That is, when more than personal care was provided (eg, emotional and/or medical), we found associations with greater perceptions of importance and a positive experience by the care recipient. Notably, providing emotional care (in addition to personal care) increased the magnitude of these associations more than the provision of medical care without emotional care. This study adds to the existing literature in 2 main ways. First, it improves our understanding of paid caregiving utilization and how specifically paid home care workers contribute to patient care, from the perspective of the care recipient(s). Previously, paid home care workers' contributions were inferred based on their job title and the formal scope of care and training requirements associated with their job title (18,19). This, however, is not always an accurate depiction of the care they provide, as demonstrated by a recent qualitative study by Reckrey et al. (5) of patient-paid caregiver dyads. This study found that in the course of their day-to-day activities, agencyemployed workers went well beyond assisting patients with personal care, often supporting patients with the management of their chronic conditions, promoting their general health, and fostering their mental health. Other qualitative studies have found the same, including a study of family caregivers by Shaw et al. (20) which found that paid home care workers often provided patients with functional and emotional support. Similarly, a recent survey by Sterling et al. (7) of over 300 paid home care workers found that nearly two thirds contribute to their patients' heart failure self-care activities, such as assisting with the preparation of low salt meals, reminding patients to take medication, and engage in physical activity. Our study expands these qualitative findings by quantitatively documenting the contributions of paid home care workers from the care recipients' perspective. Future studies are needed to determine if these findings persist nationally, because the scope of care of paid caregivers often varies by state. A second way in which our study adds to the literature is that it is one of the first to demonstrate that the type of care paid home care workers provide is associated with varying magnitudes of perceived value or quality. This is important because, despite anecdotal observations of this association, quantitative data have been lacking (21). Not only do our findings highlight the value of paid home care workers' contributions to patient care, but they are also important from a payment and policy perspective, both at the individual and organizational levels. Under the revised Fair Labor Standards Act, agency-employed workers are paid at least the federal or state minimum wage, and this pay is based on time, not on care contributions (22). Currently, 24% of paid home care workers live in households below the federal poverty line, compared to 9% of all US workers (2). Thus, to the extent that this workforce is providing care that is not being captured, acknowledged, or recognized, there is a need to reassess compensation models that will account for this contribution to patient care. Related, but separate to this, is that payment for Medicaid and Medicare-funded home care are increasingly shifting away from traditional fee-for-service models to value-based payment (VBP). VBP, which aims to contain health care costs while improving quality, is designed to reward home care agencies for the quality of services rendered to patients, rather than the volume of those services (23). Existing quality metrics for home care, however, are mostly process measures (eg, vaccination given) and utilization outcomes (eg, hospitalization; improvement in activities of daily living function) (24). Although the "overall quality" of home care is assessed, it is not specific to the contributions of the home care workers themselves (eg, the types of care provided or intensity of that care), suggesting that payment models may not sufficiently measure the quality of home care services (25). Our findings suggest that more comprehensive metrics may be needed to inform payment models for agency-employed home care workers. Notably, we found that providing emotional support was associated with greater perceptions of value. This adds to existing frameworks from the family caregiving literature, whereby the provision of emotional support is associated with better perceptions of care and higher-value care by care recipients (15,16). Research has demonstrated that family (unpaid) caregivers provide psychosocial support, including emotional and mental support (eg, touching, listening, attention, humor, assist with referral to mental health services), social and spiritual support (eg, respect spiritual needs, empathy, assist with attending religious or spiritual services), and coping with symptoms of disease (eg, anxiety, depression) (26,27). Parallel to this, studies have shown that paid home care workers have also participated in these activities for adults during coronavirus disease 2019 (COVID-19) (12), as well as for adults with high-need medical conditions, such as dementia, end of life, human immunodeficiency virus/acquired immunodeficiency syndrome, and heart failure (6,8,28). However, the specific components of the emotional-related care that paid home care workers provide and how that influences the care recipients' perceptions of care remain unknown from this study. Additionally, it is unclear what, if any, training, these workers receive regarding emotional care and the toll that providing this care takes on their own health and well-being. For example, if paid home care workers are helping patients combat depression or anxiety in the home, then they may require additional training on this topic and support for the consequences that this type of caregiving has on their own work experiences. This is especially relevant during the COVID-19 era, |
where the prevalence of mental health problems and social isolation is increasing among adults receiving unpaid and paid care (12). Implications Our findings have implications for the workforce, policies that govern payment and training, and future research. Despite providing day-to-day hands-on care, paid home care workers have historically felt invisible to the medical field and society at large (29). As demonstrated here, and from the perspective of their clients, these workers make substantive contributions beyond the personal care that they are often perceived to provide. Our findings raise awareness of the contributions of this workforce, and further, by demonstrating how these contributions are associated with value, they can inform policies to realign reimbursement for paid home care workers' wages and services (30). For example, state-level policies that govern home care curriculum and training may need to be adapted to reflect their current contributions, particularly with respect to emotional and certain medical aspects of care (31), which they may not receive standardized training on currently. Our findings also have implications for federal-level policymaking, for example, the Biden administration's recent American Jobs Plan, which includes an estimated 400 billion dollars of investments in home care workers over an 8-year period. This proposal seeks to expand home care in the United States while increasing the quality jobs for paid home care workers. Our findings provide a framework for investments in the expanded role that these caregivers play in a manner that would enhance quality of care and working conditions for this essential workforce, by highlighting 3 essential and interrelated categories of care (32). Finally, the mechanism underlying the associations between the types of assistance provided and the perceived quality of that care warrants further investigation. Future studies might investigate what factors explain or mediate this association, such as individual-level characteristics (eg, sociodemographic characteristics, experience level, attitude, and training of the paid home care worker and the clinical need and severity of the care recipient), interpersonal characteristics between the paid home care worker and the care recipient (eg, connectedness, trust, cultural fit), and factors pertaining to the employment of the worker (eg, method of payment for paid care, duration of services, and agency vs nonagency employment). Limitations Our study has limitations that warrant consideration. First, although we asked participants to report on their most recent experiences with a paid home care worker, we lack specific data on the temporality of the relationship, the duration in which the worker provided services and for whom (in the family unit), and specificity regarding the types of care within broad domains (eg, emotional support) that were provided. Second, the Empire State Poll did not include measures of participants' functional and clinical status, which are important determinants of the types of care provided by paid caregivers. Third, we lacked data on the personal characteristics of the paid home care workers, the relationship between them and their care recipients (eg, connectedness, cultural fit, trust, and objective payer data for the services they provided; these factors are likely to influence the degree to which households were satisfied with paid home care workers, but we were unable to account for these in our models. Future studies would benefit from including these additional data. Finally, although a representative random sample, participants were residents of New York State and the response rate was modest, thus their experiences may not be wholly generalizable. Conclusion In this cross-sectional survey of New York State households, we found that one third of households utilized a paid home care worker. Paid home care workers often provided care beyond personal care alone, including emotional and/ or medical care. Doing so was associated with greater perceptions of importance and positive experiences, by the household. Future studies are needed to understand whether these findings hold at the national level, and programs and policies that better match their contributions in the home to their training and payment are warranted. Distinct combinations of variant ionotropic glutamate receptors mediate thermosensation and hygrosensation in Drosophila Ionotropic Receptors (IRs) are a large subfamily of variant ionotropic glutamate receptors present across Protostomia. While these receptors are most extensively studied for their roles in chemosensory detection, recent work has implicated two family members, IR21a and IR25a, in thermosensation in Drosophila. Here we characterize one of the most evolutionarily deeply conserved receptors, IR93a, and show that it is co-expressed and functions with IR21a and IR25a to mediate physiological and behavioral responses to cool temperatures. IR93a is also co-expressed with IR25a and a distinct receptor, IR40a, in a discrete population of sensory neurons in the sacculus, a multi-chambered pocket within the antenna. We demonstrate that this combination of receptors is required for neuronal responses to dry air and behavioral discrimination of humidity differences. Our results identify IR93a as a common component of molecularly and cellularly distinct IR pathways important for thermosensation and hygrosensation in insects. DOI: http://dx.doi.org/10.7554/eLife.17879.001 Introduction Ionotropic Receptors (IRs) are a large subfamily of ionotropic glutamate receptors (iGluRs) that appear to have evolved in the last common protostome ancestor (Benton et al., 2009;Croset et al., 2010;Rytz et al., 2013). In contrast to the critical role of iGluRs in synaptic communication, IRs have diverse roles in chemosensory detection (Koh et al., 2014;Rytz et al., 2013). The best-defined functions of IRs are in olfaction, where they mediate sensory neuron responses to diverse chemicals, including many acids and amines (Rytz et al., 2013;Silbering et al., 2011). Most IRs are thought to form heteromeric ligand-gated ion channels, in which broadly expressed coreceptor subunits (e.g., IR8a, IR25a and IR76b) combine with more selectively expressed IR subunits that confer stimulus specificity Rytz et al., 2013). Many of these IRs are highly conserved in insects, indicating that they define sensory pathways common to a wide range of species (Croset et al., 2010;Rytz et al., 2013). Although most conserved IRs have been assigned chemosensory roles, we recently reported that one of these receptors, IR21a, mediates cool sensing (together with IR25a) in a population of neurons in the Drosophila melanogaster larva, the dorsal organ cool cells (DOCCs) (Ni et al., 2016). This finding raised the possibility that other IRs serve non-chemosensory functions. In this work we characterize one of these 'orphan' receptors, IR93a, which has orthologs across arthropods (Corey et al., 2013;Groh-Lunow et al., 2014;Rytz et al., 2013). RNA expression analysis in several insects and crustaceans indicates that this receptor gene is transcribed in peripheral sensory organs (Benton et al., 2009;Corey et al., 2013;Groh-Lunow et al., 2014;Rytz et al., 2013), but its role(s) are unknown. Using Drosophila as a model, we find that IR93a acts with different combinations of IRs in distinct populations of neurons to mediate physiological and behavioral responses to both thermosensory and hygrosensory cues. Results IR93a is expressed in larval thermosensory neurons and is essential for cool avoidance To investigate the expression and function of IR93a, we generated antibodies against a C-terminal peptide sequence of this receptor and obtained two Ir93a mutant alleles: Ir93a MI05555 , which contains a transposon insertion in the fifth coding exon, and Ir93a 122 , which we generated using CRISPR/Cas9 to delete 22 bases within the sequence encoding the first transmembrane domain ( Figure 1a). In larvae, IR93a protein is expressed in several neurons in the dorsal organ ganglion, one of the main sensory organs in the larval head (Stocker, 1994) (Figure 1b-c). These neurons encompass the DOCCs (labeled by an Ir21a promoter-Gal4-driven GFP reporter), and the protein localizes prominently to the dendritic bulb at the tip of the sensory processes of these cells ( Figure 1c). All expression was absent in Ir93a mutants, confirming antiserum specificity ( Figure 1c). These observations indicated that IR93a might function in cool sensing. Indeed, when larval thermotaxis was assessed on a thermal gradient (Klein et al., 2015), we found both Ir93a mutant alleles exhibited strong defects in cool avoidance ( Figure 1d). Cell-specific expression of an Ir93a cDNA in the DOCCs under Ir21a-Gal4 control fully rescued this mutant phenotype ( Figure 1d). These data demonstrate an essential role for IR93a in DOCCs in larval thermotaxis. IR93a is required, together with IR21a and IR25a, for cool-dependent physiological responses of DOCCs We next assessed whether IR93a is required for the physiological responses of DOCCs to cooling by optical imaging of these neurons using the genetically encoded calcium indicator, GCaMP6m (Chen et al., 2013). As previously reported (Klein et al., 2015;Ni et al., 2016), wild-type DOCCs exhibit robust increases in intracellular calcium in response to cooling ( Figure 2a). These responses were dramatically reduced in Ir93a mutants, and could be rescued by cell-specific expression of an Ir93a cDNA (using the R11F02-Gal4 DOCC driver [Klein et al., 2015]) (Figure 2a,b). This dramatic loss of DOCC temperature sensitivity resembles that observed in both Ir21a and Ir25a mutants (Ni et al., 2016), and is consistent with IR21a, IR25a and IR93a functioning together to mediate cool activation of the DOCCs. To provide a more direct readout of thermotransduction in these neurons than soma calcium measurements, we tested the requirement for IR93a, IR21a and IR25a in cool-evoked membrane voltage changes using the genetically encoded voltage sensor, Arclight (Jin et al., 2012). In wildtype animals, cool-dependent voltage changes were observed in the DOCC sensory dendritic bulbs (Figure 2c-d), where IRs are localized (Figure 1c). This response was completely eliminated in Ir21a, Ir25a and Ir93a mutants (Figure 2c-e), indicating that each of these IRs is required for temperaturedependent voltage changes in this sensory compartment. IR93a is co-expressed with IR25a and IR40a in the antennal sacculus In adults, Ir93a transcripts were previously weakly detected in a set of neurons in the third antennal segment surrounding the sacculus, a three-chambered pouch whose opening lies on the posterior surface of the antenna (Benton et al., 2009) (Figure 3a). With our IR93a antibody, we detected IR93a expression in neurons innervating sacculus chamber I (11.0 ± 0.5 neurons, n = 48 animals; mean ± SEM) and chamber II (13.9 ± 0.7 neurons, n = 23), with signal detected both in the soma and in the sensory cilia that project into cuticular sensory hairs (sensilla) ( Figure 3b). As in larval DOCCs, IR25a was expressed in IR93a-expressing cells in the sacculus (Figure 3c). By contrast, no expression was detected in these cells when using our Ir21a promoter driver (data not shown). We found instead that the IR93a/IR25a sacculus neurons express a distinct receptor, IR40a (Figure 3d-e) (Benton et al., 2009;Silbering et al., 2016). Larval anterior Morphological studies have suggested that neurons in sacculus chambers I and II are hygroreceptive (Shanbhag et al., 1995), raising the possibility that IR93a, IR25a, and IR40a are required for hygrosensory behavior. To test this hypothesis, we adapted an experimental paradigm (Perttunen and Salmi, 1956) in which flies choose between regions of differing humidity generated by two underlying chambers: one containing deionized water and the other containing water saturated with a nonvolatile solute (ammonium nitrate) to lower its vapor pressure (Figure 4a). This assay design created a humidity gradient of~96% relative humidity (RH) to~67% RH, with negligible variation in temperature ( Figure 4b). Consistent with previous observations (Perttunen and Salmi, 1956), wild-type flies exhibited a strong preference for lower humidity (Figure 4c). This preference was completely eliminated in Ir93a and Ir25a mutant flies, and significantly reduced, but not abolished, in Ir40a mutants (Figure 4c, Figure 4-figure supplement 1a). All of these behavioral defects were robustly rescued by the corresponding cDNAs, confirming the specificity of the mutant defects ( Figure 4c). Importantly, the loss of Ir21a (or other antennal-expressed IR co-receptors, Ir8a and Ir76b) did not disrupt dry preference. To exclude any potential contribution of the non-volatile solute to the behavior observed, we also tested flies in a humidity gradient (~89% to~96% RH) generated using underlying chambers of deionized water alone and air ( Figure 4a). Even in this very shallow gradient, wild type flies displayed a strong preference for the side with lower humidity (Figure 4d), and this preference was dependent on IR93a, IR25a and IR40a, but independent of IR21a ( Figure 4d). The distinction between the functions of IR21a and IR40a extended to thermotaxis, as Ir40a mutants exhibited no defects in this IR21a-dependent behavior (Figure 4-figure supplement 1a-b), consistent with lack of expression of IR40a in the larval DOCCs (data not shown). IRs mediate dry detection by sacculus neurons To test whether the IR40a/IR93a/IR25a-expressing sacculus neurons are physiological hygrosensors, we monitored their |
calcium responses to changes in the RH of an airstream (of constant temperature) directed towards the antenna. We used Ir40a-Gal4 to express UAS-GCaMP6m selectively in these neurons, and measured GCaMP6m fluorescence in their axon termini, which innervate two regions of the antennal lobe, the 'arm' and the 'column' (Silbering et al., 2016;Silbering et al., 2011) (Figure 5a-b). We observed that these sacculus neurons behave as dry-activated hygrosensors: decreasing the RH from~90% to~7% RH elicited an increase in GCaMP6m fluorescence, while increasing RH from 7% to~90% elicited a decrease (Figure 5c-g). Calcium changes were most apparent in the 'arm' (Figure 5c). Importantly, these physiological responses were IR-dependent: mutations in either Ir93a or Ir40a eliminated the dry response (Ir25a mutants were not tested), and these defects were restored with corresponding cDNA rescue transgenes (Figure 5d-g). These data corroborate the requirement for IRs in behavioral preference for lower humidity. The TRP channels Nanchung and Water witch do not mediate IRdependent dry sensation Previous work has implicated two Transient Receptor Potential (TRP) channels, Nanchung and Water witch, in hygrosensation (Liu et al., 2007), but it is unclear whether they have an essential function in this modality (Enjin et al., 2016;Ji and Zhu, 2015) and the cells in which these proteins act are unknown (Jourjine et al., 2016;Liu et al., 2007). In our gradient assay, we found that animals mutant for nanchung or water witch displayed partially diminished dry preference behavior ( Figure 6a). However, neither nan nor wtrw was required for the dry responsiveness of IR40aexpressing sacculus neurons (Figure 6b-e). Thus, these TRP channels are not essential for IR-dependent dry sensing, suggesting that they contribute to hygrotaxis through other mechanisms. In addition to the roles of IR93a in cool and dry sensing, it is very likely that this receptor defines additional sensory pathways. Our expression analysis has identified IR93a-positive cells that do not express IR21a or IR40a, such as non-DOCCs in the larval dorsal organ (Figure 1c). Moreover, the milder hygrosensory behavior phenotype of our protein null Ir40a mutants compared to Ir93a (or Ir25a) mutants hints that IR93a may have broader roles in this sensory modality than acting exclusively with IR40a. The populations of IR93a-expressing neurons characterized in this study are themselves heterogeneous. For example, IR40a/IR93a/IR25a-expressing sacculus neurons belong to two morphologically and physiologically distinct subpopulations. Arm neurons have contralateral projections (Silbering et al., 2011) and respond robustly to low humidity, while column neurons are exclusively ipsilateral (Silbering et al., 2011) and respond more weakly to humidity, as well as displaying mild thermosensitivity (Enjin et al., 2016). IR40a-expressing neurons also respond to ammonia (Silbering et al., 2016). Given that these neurons are housed in apparently poreless sensilla (Shanbhag et al., 1995), we speculate that this chemical activates these cells indirectly, for example, through modification of the humidity of the air, or the temperature of the cuticular surface, within the sacculus. A key future challenge will be to determine the mechanisms by which IRs contribute to the sensation of thermal and humidity cues. We previously showed that ectopically-expressed IR21a can confer cool sensitivity to other IR-expressing neurons, consistent with IR21a acting as a sensory specificity determinant (Ni et al., 2016). It will be important to determine if IR40a serves a more permissive role or functions in a similar capacity in dry sensing. The contribution (if any) of the Venus flytrap-like ligand-binding domains of these receptors is of particular interest. Although this domain recognizes glutamate in iGluRs, and diverse organic molecules in chemosensory IRs, it is conceivable that these domains mediate thermo-and hygrosensory detection in these receptors in a ligand-independent manner. For example, IR21a could transduce information via temperature-dependent conformational changes. The requirement for IR93a (and IR25a) in both thermosensation and hygrosensation also indicates that these modalities could share common mechanisms of sensory detection. For example, hygrosensation could involve a thermosensory component, based on evaporative cooling. Alternatively, both temperature and moisture detection could involve mechanosensation, based on swelling or shrinkage of sensory structures, as suggested in mammals and C. elegans (Filingeri, 2015;Russell et al., 2014). Further characterization of how IRs mediate temperature and moisture detection is currently limited by our inability to reconstitute thermosensory or hygrosensory responses in heterologous systems by expressing the known combinations of IRs (G.B., L.N., A.F.S., R.B. and P.G., unpublished data). IRs, like iGluRs, are thought to form heterotetrameric complexes , raising the possibility that additional IR subunits are required. It is also conceivable that other types of accessory signaling molecules act with IRs, and/or that the cellular and cuticular specializations of the thermosensory and hygrosensory structures are critical to allow monitoring of these ubiquitous and ever-changing environmental stimuli. Behavior Thermotaxis of early second instar larvae was assessed over a 15 min period on a temperature gradient extending from 13.5 to 21.5˚C over 22 cm (~0.36˚C/cm) as described (Klein et al., 2015). As thermotaxis data were normally distributed (as assessed by Shapiro-Wilk test), statistical comparisons were performed by Tukey HSD test, which corrects for multiple comparisons. To assay hygrosensory behavior, 8 well rectangular dishes (12.8 Â 8.55 Â 1.5 cm; ThermoFisher #267060) were modified to serve as humidity preference chambers. The lids of two 8 well plates were used. A heated razor blade was used to cut out the middle of one lid, and a nylon mesh was glued into place around the edges, providing a surface for the animals to walk on which separated them from contact with any liquid. A soldering iron was used to melt a small hole in a second culture plate lid, which could then be placed over the screen, creating a chamber~0.7 cm in height in which the flies could move freely. To monitor the gradients formed, an additional chamber was constructed with four holes equally spaced along its length to allow the insertion of humidity sensors (Sensirion EK-H4 evaluation kit) for monitoring the humidity and temperature. Prior to the start of each experiment, 4 wells on one side of the culture dish were filled with purified water, while the opposite 4 were filled with~4 ml water and sufficient ammonium nitrate to obtain a saturated solution (~3 g). The gradient was assembled with the screen and lid piece, and the whole apparatus wrapped in food service film to avoid any transfer of air between the inside and outside of the device. Gradients were transferred to an environmental room that maintained at constant external temperature and humidity (25˚C and 70%RH). Ammonium nitrate gradients were permitted to equilibrate for approximately 1 hr and were stable over many hours. For the water and air only gradients, the air only side humidified over time. These gradients were incubated for 25 min prior to use to allow the temperature to equilibrate; the humidity of the dry side typically rose bỹ 2% RH during the 30 min assay (values shown are at the 30 min time point). A small hole was poked through the food service film covering the device to allow animals to be transferred to the gradient. This hole was sealed using transparent scotch tape once the animals were inside. Experiments used 1-4 day old adult flies that had been sorted under light CO2 anesthesia into groups of 30 (15 male and 15 female) animals 24 hr before testing, and transferred to fresh tubes. Flies were allowed 30 min to settle on the gradient, at which point a photograph was taken of their position, and the number of animals on each side counted, allowing calculation of a dry preference index as follows: Dry Preference ¼ # animals on dry side À # animals on moist side total # of animals As moisture preference data did not conform to normal distributions (as assessed by Shapiro-Wilk test, p<0.01), statistical comparisons to wild-type control were performed by Steel test, a nonparametric test that corrects for multiple comparisons, using JMP11 (SAS). Calcium and Arclight imaging Calcium and Arclight imaging of larval thermosensors was performed as previously described (Klein et al., 2015). Pseudocolor images were created using the 16-colors lookup table in ImageJ 1.43r. Adult antennal lobe calcium imaging was performed as described for olfactory imaging (Silbering et al., 2012), with slight modifications to sample preparation and stimulation. Briefly, 3-7 day old flies were fixed to a Plexiglas stage using UV-glue (A1 Tetric Evoflow, Ivoclar Vivadent), the antennae were pulled forward and a small opening was made in the head capsule to allow visual access to the antennal lobes. For the stimulation compressed air from a tank was passed through activated charcoal and then either through an empty gas washing bottle or a gas washing bottle filled with distilled water producing either a dry airstream of~7% RH or a humid airstream of ~90% RH. A computer controlled solenoid valve (The Lee Company, Westbrook, CT) was used to switch the airflow between the two gas washing bottles. The flow was kept constant at 1 l/min with a parallel arrangement of two 500 ml/min mass flow controllers (PKM SA, www.pkmsa.ch) placed before the gas washing bottles. Activating the solenoid valve resulted in a complete reversal of RH from low to high or high to low within less than 10 s. For each animal tested, both high to low and low to high RH transitions were applied in random order. Following humidity stimulation, a final pulse of 10% ammonia was applied as a control to confirm cellular activity (Silbering et al., 2016) (animals showing no response to this positive control were excluded from the analysis). Data were processed using Stackreg (ImageJ) (Thevenaz et al., 1998) to correct for movement artifacts (animals with movement artifacts that could not be corrected with Stackreg were excluded from the analysis) and custom scripts in Matlab and R as previously described (Silbering et al., 2011). As quantified imaging data did not conform to normal distributions (as assessed by Shapiro-Wilk test, p<0.01), statistical comparisons were performed by Steel-Dwass test, a non-parametric test that corrects for multiple comparisons, using JMP11 (SAS). The perpetuation and epidemic recurrence of communicable diseases in human populations Recurrence of communicable diseases is a looming threat for human populations. Factors explaining the recurrences are partially known, involving demographics, biology, and complex relationships with the environment, but no comprehensive theory exists today. Here, we review some recent results obtained in modelling studies with a view to understanding better the mechanisms of perpetuation. Factors intrinsic to the interaction of pathogen and host have regained interest in this respect, especially with multiple pathogen and multiple population interactions. Extrinsic factors, including pure demography and environmental forcing are also strong predictors. With increasingly detailed data available, large-scale integrated models will help sorting out the multiple influences on recurrence. To cite this article: P.-Y. Boëlle, C. R. Biologies 330 (2007). Conclusion Il n'existe pas encore de théorie unique pour expliquer la totalité des manifestations cycliques et récurrentes des épidémies. La prise en compte des multiples acteurs dans des systèmes réalistes sera source de travaux dans les prochaines années. Introduction When it was first used by Hippocrates in a medical context, the Greek word 'epidemic' had already changed meanings from an initial 'back in his country' for the more dynamical 'propagating in the country', adequately qualifying the spread of a disease [1]. Still, the very etymology of the word made it clear that 'epidemics' were 'at home' in human populations that they visited from times to times [2]. Indeed, while outbreaks of communicable diseases may be limited in space and time, their propensity to recur place a looming threat on human populations. In May 2003, an editorial in the New England Journal of Medicine about SARS concluded: "if we are extremely lucky, the epidemic will be curtailed, develop a seasonal pattern that will improve prospects for regional containment" [3], showing that recurrence of the disease was taken seriously by experts in the field. This prediction did not realize, illustrating that our current understanding of epidemics allows betting rather than logic. Indeed, the factors making recurrence more likely remain unclear, as multiple relationships, acting at various levels, may be required [4]. Many infectious diseases present a seasonal pattern, especially those caused by respiratory pathogens [5]. Interpandemic influenza epidemics occur during fall and winter in temperate countries of the northern hemisphere [6]. Respiratory syncitial virus activity peaks in the same period, although seemingly independently [7]. Rotavirus infection displays also a very seasonal pattern, especially among the young [8], |
as do other childhood diseases [9]. In the latter, seasonal patterns vary with location, with yearly epidemics (for example, in France [10]) or epidemics of varying size every other year, or do not exist at all in the absence of imported cases [11]. Other diseases present a pseudo periodic behaviour that is more loosely related with seasonal variation. Chikungunya fever is reportedly recurring every four years in Senegal [12]. Smallpox outbreaks occurred with a frequency of five to seven years [13,14], while meningococcal meningitis presented cycles of 10 years [15]. Syphilis also shows recrudescence periods separated by approximately 11 years [16], while gonorrhea or HIV infection, although transmitted mainly by the same route, do not show evidence of periodic behaviour. These various situations illustrate diseases that perpetuate in human populations: cases may be found without interruption, even if incidence varies [17]. Seasonal outbreaks may rhythm the perpetuation, or seemingly haphazard outbreaks interspersed with large silent periods. In all cases, perpetuation requires that a portion of the population is susceptible to the disease, through either demographic replacement, or loss of immunity. Recurrence may require a reservoir population for the pathogen, which may be the human population itself or some interacting species. In the following, we will not discuss the case of zoonotic diseases unable to perpetuate in humans, but introduced from time to time by direct exposure: plague, for example, but also the current form of avian influenza. The relative importance of intrinsic and extrinsic factors in the perpetuation and recurrence of infectious diseases is debated [18,19]. Intrinsic factors are those coming directly into play in the interaction between a pathogen population and its host population. Extrinsic factors are those that may modulate this interaction, but not be affected by it, like climate for instance. In this article, we review some theoretical results regarding perpetuation and recurrence of communicable diseases. In a first part, intrinsic factors are explored, and extrinsic factors are then reviewed. Intrinsic factors associated with perpetuation and recurrence A number of determinants for the perpetuation of transmissible diseases in human populations have been listed [17], highlighting the importance of both demographical and biological parameters. Demographical parameters include the size of the population, the rate of population turnover, the rate and pattern of contacts; biological parameters include the duration of immunity, the susceptibility and transmissibility, the duration of infectiousness and the generation period (the duration between an index case and a secondary case). The importance of these factors is illustrated in the following paragraphs. One population, one pathogen: the SIR model In the famous SIR model, the human population is split in three subgroups: (S)usceptibles to infection, (I)nfectious and (R)emoved by cure or death, with the following differential equations describing contamination, cure and removal [20]: This basic model aims at describing the circulation of a pathogen in a single host population, leaving those infected immune or dead. These equations do not allow for perpetuation or recurrence unless a demographic birth and death process is also implemented or a loss of immunity with time taken into account in those already recovered. In either case, an equilibrium state where infectious perpetuate may be obtained, with quasi-periodic oscillations around the equilibrium. The approximate period of the oscillations is T = 2π √ DA, where D = 1/γ is the average duration of the infectious period and A = μ(1 − 1 R 0 ) is the average age at infection, with μ the birth (or loss of immunity) rate and R 0 = β/(μ + γ ) the basic reproduction number of the disease (i.e. the number of secondary cases borne from one initial index case in a completely susceptible population) [21]. The period of oscillations is therefore increased by a low birth rate (or low loss of immunity) and a short duration of the disease. The occurrence of oscillations is a direct consequence of the form of the interaction between susceptible and infected individuals (i.e. the law of mass action) in a finite population, leading to a critical threshold in the number of susceptible individuals below which the incidence of new cases fails to compensate those removed; the dynamics of births, which replenishes the susceptible compartment with time and allow for renewed increase in prevalence. This intrinsic periodic behaviour was first thought to account fully for recurrent epidemics [22], all the more because almost yearly periods may be obtained with reasonable parameter values. However, the oscillations damp with time and fail to explain sustained seasonal cycles [23]. Inclusion of realistic distributions for the duration of the infectious period (in the SIR model, an exponential distribution is assumed) shows that, while damping may take considerable time, it is nevertheless present [24]. Introducing stochastic components changes the behaviour of models with respect to perpetuation in an essential way, since there is a possibility of disease extinction [25]. A stochastic treatment to the SIR model leads to undamped oscillations in incidence, but the disease goes extinct with probability 1 [26]. Furthermore, taking into account realistic distributions for the epidemic period even hasten this process, as the troughs are more pronounced [24]. Therefore, although the simple SIR model provides for both perpetuation and recurrence, it fails short to provide a satisfying explanation to observed infectious disease dynamics. Several populations, one pathogen Having multiple coupled populations somewhat alleviates the problem of extinction. Measles in the UK has provided ample evidence of source dynamics explaining recurrences in isolated populations. In London, for example, measles never goes extinct, since the size of the children population is large enough to perpetuate the disease, with a simple seasonal pattern (see also extrinsic factors below). On the contrary, small places show scattered epidemics, interspersed by irregular time intervals [27]. These 'meta-population' dynamics may have vastly diverging effects depending on the disease. For example, measles dynamics tends to be synchronized through these interactions, while whooping cough, with an infectious period approximately thrice longer, is desynchronized [28]: this is explained by a subtle interplay between the duration of the infectious period, the coupling between populations and seasonal forcing. Perpetuation may be more likely with desynchronized epidemics in a meta-population approach. A very different meta-population dynamics is seen in vector-borne diseases. In dengue and malaria, the pathogen is transmitted by mosquitoes. The disease dynamics is therefore subject to the population dynamics of both species. Interestingly, this translated into an al-most three-year interepidemic period for both diseases, in very good agreement with the prediction of a SEIR model [29]. More generally, multi-host infectious diseases may be associated with original perpetuation and recurrence characteristics. This is so because multi-host pathogens have generally reservoir hosts in which they persist with little clinical manifestation, and cause outbreaks outside this reservoir [30]. Even if malaria was one of the first diseases submitted to a mathematical treatment, there is still considerable work to gain further insight into the emerging zoonotic threat [31]. One population, several pathogens In the preceding, a limited view of the interaction between pathogens and humans was adopted, taking no notice of diversity in the pathogen population. This may be true for diseases like measles, where little viral diversity exist. However, the influenza virus effectively escapes immunity by constant changes at key antigenic sites [32], causing an effective influx of susceptible individuals. The joint study of pathogen diversity and epidemic transmission is the subject of much current research [33]. When several strains of a pathogen co-circulate, we may expect independent behaviour and coexistence at levels determined by the relative fitness of each strain. However, this does not account for potential interference between strains: competing for susceptible hosts [34], reducing susceptibility to infection by genetically 'close' strains [35] or transmissibility after an initial infection [36]. With these refinements, unconstrained coexistence may be the rule at low levels of cross immunity, but only non-interfering strains may coexist at very high levels of cross immunity. Furthermore, in a range of intermediate levels of cross immunity, periodic or chaotic cycling between strains is possible [35][36][37][38]. In this case, the secular variation seen in the prevalence of the strains of some pathogens may result from complex interplay due to cross immunity. Examples conforming to these predictions may include lineages of N. meningitidis and serotypes of group-A streptococcus [39]. If the co-circulating strains have different pathogenicity, periodic or recurrent outbreaks will occur [40]. It is noteworthy that such rich dynamics may also be observed in a very particular case of interference known as 'antibody-dependent enhancement' [41,42]. In this case, previous infection does not act to reduce susceptibility or transmission to a new infection, but on the contrary may increase transmissibility during a second infection and infectiousness. This has been proposed in the case of dengue (although it is debated [40]), and for other viruses as well (among others coronaviruses of the SARS type [43]). Modelling shows that antibody enhancement may lead to undamped cyclical epidemics with different strains, much like more classical interactions. Diversity in pathogens is, like transmission, a dynamic process, and should be considered in a 'phylodynamics' approach [33]. Models were proposed in this respect where strain diversity changes with time given a constant mutation rate, accompanied by a decrease in cross immunity with increasing genetic distance [44,45]. In a detailed analysis of the behaviour of this model, it appears that combining short infectious duration with long-lived immunity induced by infection (total immunity against the same strain, and partial immunity to other strains) lead to the occurrence of renewed outbreaks with time, where at each time the dominant strain is different from that in earlier epidemics [45]. Calibrating the model to reflect influenza, it was found that the simulated phylodynamics were reminiscent of that observed in real influenza [44]: several closely related genetic strains co-circulating at any given time, with a continuous drift [46]. Networks and heterogeneous mixing An assumption of the basic SIR model is that of 'homogeneous mixing', whereby all infectious may contact all the susceptible individuals. This also means that little variation is expected in the number of secondary cases borne from two different infectious cases, since the effective contacts take the form of a completely random graph. Studying the impact of heterogeneity in transmission was first addressed by splitting compartments according to individual characteristics. The impact of such structuring was manifest in the study of sexually transmitted infections (STI), and led to the influential concept of 'core groups' having more numerous partners than the rest of the population [47]. These groups were essential to explain the maintenance of STIs in a population where it would have otherwise disappeared. The recent epidemic of SARS has put forward that in other diseases as well, the number of secondary cases could be highly variable. In fact, most cases were without descent, while few cases had a disproportionately high number of descendants [48]. The so-called 'superspreaders' have since then been recognized as fairly common in many communicable diseases, appearing as the norm rather than the exception [49]. While superspreaders may have a definite role in introducing a disease in a population, and be targeted first to prevent its spread, it is not clear as of today how their role would extend in perpetuating a disease. Superspreaders echo findings from complex network theoretical epidemiology. In those, the topology of contact networks has been changed from completely random to include small world or scale-free properties. A major finding of these studies has been the strong dependence of epidemic spread on network structure [50]. For example, contrary to random networks, infinite power-law networks do not display threshold behaviour for the persistence of a disease, even if a thresholdlike phenomenon may be recovered in finite-size networks [51]. Provided that susceptible individuals are introduced or recovered ones lose their immunity, the perpetuation of any disease is theoretically granted in such networks. Indeed, any initial number of infected individuals may lead to large-scale epidemics. Oscillatory dynamics in prevalence may appear in growing networks [52], while in other cases perpetuation exists without periodicity. Extrinsic properties leading to recurrence A large number of extrinsic factors have been suggested as promoting recurrence or perpetuation of a communicable disease, especially those based on seasonal forcing. From an evolutionary point of view, the existence of a regular forcing imposed by seasons must have led to selecting the type of interactions that allows perpetuation of pathogens in the host populations. Other |
explanations involve seasonal demographic processes leading to changes in host behaviour, to environmental changes or to changes in pathogen presence. Demographics The mere size of the population has a profound effect on the perpetuation of diseases. Large populations enable perpetuation, whereas the disease generally disappears in small populations. The critical community size corresponds to the threshold below which perpetuation is not possible [53]. Changing birth rates may profoundly affect the pattern of recurrence of diseases. This is best exemplified for measles epidemics, where sustained annual epidemics require high birth rates. In the UK, decreasing birth rates have been associated with the occurrence of large epidemics every other year [54]. Importantly, external stimulation of the model through the changing birth rates was sufficient for the trajectory to hop between attractors having cycles of various lengths [54]. The same profound effect of birth rates was seen for smallpox. Indeed, the intrinsic period of this disease, in a population of constant size, is 3 to 4 years; however, inspection of past time series show preferentially 5-year cycles [13], or even longer. The explanation of this period lengthening has been linked to extrinsic processes, involving for example the occurrence of famine and high wheat price [13]. Seasonal changes in host behaviour Many human activities exhibit a seasonal component, and this may have implications in disease transmission. For example, childhood diseases exhibit marked seasonality, although patterns may change between places. Incidence increases during the whole school year, and drops in the summer. A very good candidate to explain this seasonality pattern is the forcing imposed by school holidays, and it has been repeatedly found that this leads to improved fit to observed data [55]. For diseases with long infectious periods, school-term forcing leads to annual cycles. When the duration of the infectious period decreases, as well as the effective number of secondary cases, longer cycles may occur [56]. Large differences in contact rates also lead to longer cycles. On the contrary, most respiratory diseases exhibit a seasonal pattern with incidence increase during the cold season. Importantly, this is not a characteristic of the disease only, as it is well documented that the seasonal component increases with latitude: tropical countries experience similar attack rates as temperate countries, only the epidemic period is diffuse, instead of concentrated during a few weeks [5]. Up to now, no single theory has been sufficient to explain all seasonal patterns. First, there may be changes in contact rates as a consequence of crowding in colder seasons. It was recently proposed that even subtle changes in contact rate could lead to large amplitude changes due to resonance between the intrinsic cycle of the disease and that in contact rates [57]. In any case, seasonal forcing in the contact rate in SIRtype models introduces extremely rich dynamics [54]. It has also be suggested, and found in mice, that seasonal changes in immunity of the host were at stake [5]. Given that there is so much evidence for seasonal changes, one may wonder on the advantage of seasonal variation in incidence for perpetuation of a disease. We must recall first that the dynamics of the SIR model intrinsically leads to oscillations in the number of susceptible individuals. If there is a good correlation in the intrinsic cycle of the disease and seasonality in contact rates (i.e. if the pathogen has been selected so), the pathogen becomes more firmly established as the magnitude of seasonal changes increases. However, the same argument leads to favour an increase in duration between recurrences of the disease, and ultimately an escape from the seasonal behaviour [58]. Further theoretical results are required to resolve these apparent paradoxes. Environment: weather, satellites and El Niño It is well known that the environment plays a key role in the distribution of human diseases [59], through climatic factors. Could global processes be associated with the occurrence of disease in particular places? While not causal determinants, global processes could favour environmental conditions leading to recurrence or perpetuation. El Niño occurs irregularly (cycles between two and seven years). The El Niño cycle is associated with increased risks of some of the diseases transmitted by mosquitoes, such as malaria, dengue, and Rift Valley fever [60], but also of cholera [61]. Candidate mechanisms linked to the effect of environment on the vector, or on water movement have been proposed. More unexpectedly, the cold event of the El Niño oscillation (or La Niña) is also associated with higher influenza mortality [62]. Other environmental factors, like vegetation indices, have been linked to the occurrence of vector-borne diseases like the Rift Valley fever disease [63]. These indicators would be extremely helpful for predicting and organizing containment measures: this is a way forward for the coming years [64]. Conclusion The natural movement of human populations is a key determinant to the perpetuation of communicable diseases. While it was laid down very early that "the recurrence of epidemics depends solely on two factors, the time of importation of the morbid poison and the number of persons susceptible to it" [65], mathematical models have been instrumental in suggesting which mechanisms ruled importation and susceptibility. For some diseases, where the virus is not to genetically diverse and does not exist in external reservoir, it seems that the determinants of the oscillatory-like dynamics of communicable diseases are well characterized: this is now to a large extent the case for measles. As the 20 th century saw the solution to the one population/one pathogen problem, the 21 st century must be the time when complex systems, involving multiple populations, multiple pathogens and the environment will be fully understood [66]. Current interest in influenza shows the complexity of this task, and how many domains will have to communicate: molecular techniques may help to precise how strains evolve and perpetuate in real epidemics [67], and large-scale computer simulations how epidemic spreads in realistic populations [68]. What will make all these developments possible is a renewed effort into observation, to correlate better observations now taken disparately: human, veterinarian, and the environment. The recent studies into cholera-recurrence dynamics, implicating first El Niño events [61] as likely to promote outbreaks, then more recently linked to the vibriophages' population [69], illustrate perfectly the way to follow in the future: reasoning at several scales, using models to assemble detailed information gathered in the field in a comprehensive framework for analysis. Targeting Efficacy of Simvastatin for Hormone-Dependent Carcinomas through Solid Lipid Nanoparticles Folate conjugated Simvastatin loaded Solid lipid nanoparticles (SIM-SLNs) were constructed after the preformulation studies such as Differential calorie metry (DSC), Fourier transform infrared spectroscopy (FTIR) and partition coefficient studies. SIM-SLNs were evaluated in terms of particles size, zeta potential, drug encapsulation, and drug release. The cytotoxicity of the SLNs was evaluated on MCF-7 Human Breast Carcinoma cells (MCF-7 cells) by MTT assay. In vivo pharmacokinetic parameters and Bio-distribution were observed in Wistar rats. The particle size of SIM-SLNs (F6) was 177.4 nm with a zeta potential of -5.16 mV. The entrapment efficiency and drug loading was found to be higher in SIM-SLNs. SLN formulation released the drug in sustained manner over a period of 24 hours. Various kinetic models were fitted and it was observed that Higuchi’s model was the best fit model. SIM-SLNs showed highest cytotoxicity in tumor cells in vitro. In vivo studies showed that there is significant increase in the pharmacokinetic parameters and Bio-distribution of SIM-SLNs when compared to the pure drug suspension. Higher relative bioavailability would be due to avoidance of first-pass hepatic metabolism by intestinal lymphatic transport, which circumvents the liver. According to these results, the novel SIM-SLNs provided sustained release of the drug, and these systems are the preferred drug carriers for lipophilic drugs to overcome the oral bioavailability problem of drug and to increase targeting efficacy to the specific organs as breast, prostate and ovary. Introduction Hormone-dependent carcinomas are classically those of reproductive tract-Prostate and Testis cancer in men and Breast, Ovarian and Endometrial cancer in women [1]. Worldwide, breast cancer is the leading type of cancer in women, accounting for 25% of all cases. In 2012 it resulted in 1.68 million cases and 522,000 deaths. It is more common in developed countries and is more than 100 times more common in women than in men [2]. Breast cancer is a cancer that develops from breast tissue. Prostate cancer, also known as carcinoma of the prostate, is the development of cancer in the prostate, a gland in the male reproductive system.Most prostate cancers are slow growing; however, some grow relatively quickly. The cancer cells may spread from the prostate to other parts of the body, particularly the bones and lymph nodes [3]. Simvastatin, an anti-hyperlipidemic drug shows the anti-cancer activity on hormone dependent carcinomas. The fact that mevalonate plays a key role in cell proliferation and that many malignant cells present an increased HMG-CoA reductase activity, suggests that a selective inhibition of this enzyme could lead to a new chemotherapy for cancer disease. Results obtained in vitro have shown that statins can inhibit tumor cell growth. The obtained reduction of sterols synthesis by statins, suggests that inhibition of tumor cell growth can be related to the reduction of nonsteroidal isoprenoid compounds [4,5]. The main challenge in cancer chemotherapy is toxic side-effects induced by chemotherapeutic drugs. Single dose or short-time application (1-2 weeks) will probably causes serious health problems, but the use of biodegradable nano-sized particles for longterm or life-time therapy may produce better therapeutic-effects with less side effects [6,7]. Increasing the encapsulation efficiency of poorly water-soluble molecules will lead to the development of improved SLN formulations and drug-loaded SLN formulations to increase the efficacy and reduce the side-effects of chemotherapeutic drugs for anticancer treatment [8]. The purpose of the study is to construct Solid lipid Nanoparticles (SLNs) as nanomedicine for targeting hormonedependent carcinomas by sustained release of the drug and improved oral bioavalabity due to avoidance of first-pass hepatic metabolism through intestinal lymphatic transport, which circumvents the liver. Preformulation studies Differential scanning calorimetry (DSC): Differential scanning calorimetric analysis was used to characterize the thermal behavior of the drug, lipids and their physical mixtures. Sample was crimped in standard aluminium pans and heated from 20 to 400°C at a heating rate of 10°C/min under constant purging of dry nitrogen at 30 ml/ min. An empty pan, sealed in the same way as the sample, was used as a reference. DSC thermograms were obtained using an automatic thermal analyzer system. Temperature calibration was performed using Indium calibration reference standard. Fourier transform infrared spectroscopy (FTIR): FTIR spectroscopy can be used to investigate and predict any physiochemical interaction between different components in a formulation and therefore it can be applied to the selection of suitable chemical compatible excipients while selecting the ingredients, we would chose, those which are stable, compatible, cosmetically and therapeutically acceptable. Infrared spectra matching approach was used for detection of any possible chemical interaction between the drug, lipid and surfactants. A physical mixture of drug, lipid and surfactants was prepared and mixed with suitable quantity of potassium bromide. This mixture was compressed to form a transparent pellet using a hydraulic press at 15 tons pressure. It was scanned from 4000 to 400 cm -1 in a FTIR spectrophotometer (FTIR 8400 S, Shimadzu). The IR spectrum of the physical mixture was compared with those of pure drug, lipid and surfactants and peak matching was done to detect any appearance or disappearance of peaks. Partition coefficient studies: Partitioning behavior of Simvastatin(SM) was determined in different lipids viz. stearic acid,glyceryl monostearate and tripalmitin. 10 mg of Simvastatin was dispersed in a mixture of melted lipid (1 g) and 1 mL of hot phosphate buffer pH 6.8 (PB) and shaken for 30 min in an water bath shaker maintained at 10°C above the melting point of lipid. The aqueous phase of the above mixture was separated from the lipid by centrifugation at 10000 rpm for 20 min. The clear supernatant obtained was suitably diluted with phosphate buffer pH 6.8 (PB) and Simvastatin content was determined in UV-Visible spectrophotometer at 239 nm against solvent blank. The partition coefficient was calculated as: where, C SMI =the initial amount of SM added (5 mg and 10 mg) C SMA =the concentration of SM in pH 6.8 PB. Preparation of solid lipid nanoparticles (SLNs) Solid lipid nanoparticles were prepared by solvent evaporation technique. |
First, o/w microemulsions were prepared. The organic phase was stearic acid or Glyceryl monostearate or tripalmitin. Tween80 (hydrophilic surfactant) solution was used as continuous phase. The chosen lipid and drug and 10 mL of ethanol were heated at 60°C. 50 mL of aqueous surfactant solution containing Tween 80 heated at same temperature and the organic phase was added to the aqueous phase with mechanical stirring for 15 minutes. A clear microemulsion was obtained under stirring at a temperature close to the melting point of the lipid used. Folic acid solution was used as a ligand and added to microemulsion. Solid lipid nanoparticles were obtained by dispersing the warm o/w microemulsion dropwise into ice cold water in a beaker under continuous stirring. SLN dispersion was further stirred after complete addition of microemulsion for 4 hours at 2000 rpm. After completion of stirring, the SLN dispersion was subjected to ultrasonication for 15 minutes. Similarly drug loaded solid lipid nanoparticles were prepared by Solvent evaporation method using stearic acid (SA) and glyceryl monostearate (GMS) and Tripalmitin (TP) as lipids, Simvastatin as drug, Tween 80 was used as hydrophilic surfactant, Folic acid was used as ligand [9,10]. Evaluation of solid lipid nanoparticles Particle size and zeta potential: Particle size and zeta potential of the solid lipid nanoparticles were measured by photon correlation spectroscopy using a Malvern Zetasizer Nano ZS90 (Malvern Instruments, Worcestershire, UK), which works on the Mie theory. All size and zeta potential measurements were carried out at 25°C using disposable polystyrene cells and disposable plain folded capillary zeta cells, respectively, after appropriate dilution with original dispersion preparation medium. Where, D (0.9) corresponds to particle size immediately above 90% of the sample. D (0.5) corresponds to particle size immediately above 50% of the sample. D (0.1) corresponds to particle size immediately above 10% of the sample [10]. Entrapment efficiency and drug loading: Entrapment efficiency and drug loading of Solid lipid nanoparticles (SLNs) was determined by weighing empty riavials and add 3 mL of the formulation into riavials. Reweigh it again. Centrifuge the sample for 10 mins with 7000 rpm. After centrifugation take 1 mL of each supernatant layer of the sample and then make upto 10 mL with pH 6.8 buffer and do serial dilution and the supernatant liquid absorbance value was determined by UV visible spectroscopy at 239 nm using pH 6.8 PB as blank. External Morphological Study: External morphology of nanoparticles was determined using Scanning Electron Microscopy (SEM). Samples were diluted with ultrapurified water to obtain a suitable concentration. Then the samples were spread on a sample holder and dried using vacuum. They were subsequently coated with gold (JFC 1200 fine coater, Japan) and examined by a Scanning Electron Microscopy (SEM). In vitro release studies The release of Simvastatin from the SLNs was studied under sink conditions. Glyceryl monostearate (GMS-SLN) and Tripalmitin (TP-SLN) which showed higher drug content and entrapment efficiency (F3 and F6) were evaluated for in vitro release. 10 mL of formulation(F3 and F6) were kept in dialysis bags (MWCO 12000, HiMedia). The dialysis bags were placed in 100 mL of dissolution medium (pH 6.8 PB) were kept in orbitory shaker for 24 hrs at 37°C. Aliquots of the dissolution medium were withdrawn at each time interval and the same volume of fresh dissolution medium was added to maintain a constant volume. Samples withdrawn from pH 6.8 phosphate buffer were analyzed for Simvastatin content spectrophotometrically at 239 nm against solvent blank. Haemocompatibility studies Blood samples of healthy human volunteers were obtained from blood bank of government hospital, Ooty in evacuated siliconized glass tube containing sodium citrate. Red blood cells were separated by centrifugation at 1500 rpm for 10 min and then washed 3 times with saline. Stock solution of erythrocytes in Saline water was prepared such that the cell count was 1×108 cells/ml. Equal volumes of RBC suspension and nanoparticles dispersion were suspended in a microcentrifuge tube such that the final concentrations of nanoparticle dispersion and nanoparticles were 150 -1000 μg/ml and incubated seperately at 37°C for 1 h. 1% Triton X and Saline water were used as positive and negative controls respectively. After 1 h the tubes were centrifuged at 1500 rpm for 10 min and the hemoglobin released in the supernatant was detected by UV absorbance at 239 nm. All measurements were performed in triplicate (n=3) and the SD was calculated. The percent haemolysis was calculated by the formula; Where Abssample is the absorbance of supernatant of erythrocyte and nanoparticles suspension. Abs0% is the absorbance of supernatant of erythrocyte and PBS suspension. Abs100% is the absorbance of supernatant of erythrocyte and Triton X. In vitro cytotoxicity studies The monolayer cell culture was trypsinized and the cell count was adjusted to 1.0 × 10 5 cells/ml using DMEM medium containing 10% FBS. To each well of a 96 well microtitre plate, 100 µl of the diluted cell suspension (approximately 10,000 cells/well) was added. After 24 hours, when a partial monolayer was formed, the supernatant was flicked off, the monolayer was washed once with medium and 100 µl of different sample concentrations prepared in maintenance media were added per well to the partial monolayer in microtitre plates. The plates were then incubated at 37°C for 24 hrs in 5% CO 2 atmosphere, and microscopic examination was carried out and observations recorded. After 24 hours, the sample solutions in the wells were discarded and 20 µl of MTT (2 mg/ml) in MEM-PR (MEM without phenol red)/PB was added to each well. The plates were gently shaken and incubated for 3 hours at 37°C in 5% CO 2 atmosphere. The supernatant was removed and 100 µl of iso-propanol was added and the plates were gently shaken to solubilize the formed formazan. The absorbance was measured using a microplate reader at a wavelength of 540 nm. The percentage growth inhibition was calculated using the following formula and concentration of drug or test samples needed to inhibit cell growth by 50% values were generated from the dose-response curves for each cell line. Mean OD of individual test % Growth inhibition = 100 -X100 Mean OD of control group Drug suspension (0.3% w/v CMC) and solid lipid nanoparticles were administered orally by oral feeding tube at dose of 1.6 mg and 2.86 mg of drug for F3 and F6 formulations respectively. Blood (0.5 mL) was collected by retro-orbital puncture at 0, 0.30,1,4,8,12 and 24 hours after administration separately. Blood samples were placed into eppendorf tubes containing 0.3 mL of anticoagulant (sodium citrate) solution and centrifuged immediately. After centrifugation, the plasma obtained was stored at -20°C until further analysis. The organs like Heart, Brain, Kidneys, Liver, Lymph, Spleen, Ovaries and Uterus in female rats and Testis,Vas deferens in male rats were collected.The collected organs were washed with normal saline and homogenized and centrifuged for 10 mins at 10,000 rpm.Then the supernatant solution was analyzed by HPLC [11]. Bioanalytical method development and analysis Reverse phase HPLC method is the most popular mode for analytical and preparative separations of the compounds in chemical, biological, pharmaceutical and food samples. In reversed phase mode, the stationary phase is non polar and the mobile phase is polar. The polar compounds gets eluted first in this mode and non polar compounds are retained for longer time. In present study, methods for the estimation of Simvastatin present in the blood plasma samples were developed and validated. For the estimation of Simvastatin in blood plasma, the chromatographic variables, namely pH, solvent strength, solvent ratio, flow rate, addition of peak modifiers in mobile phase, nature of the stationary phase, detection wavelength and internal standard were studied and optimized for the separation and retention of the drug. The following are the optimized chromatographic conditions, preparation of standard and sample solutions and the methods used for the estimation of Simvastatin in plasma. Preparation of Simvastatin standard stock solution: 10 mg of Simvastatin was transferred into a 10 mL volumetric flask and the volume was made upto the mark with mobile phase to give 1 mg/mL (1000 µg/mL) solution. From this stock solution, 10 mL of 100 µg/mL solution was prepared and again from this solution 10 mL of 10 µg/mL was prepared. Preparation of analytical calibration curve solutions: From the standard stock solution 0.25-2 µg/ml standard solutions were prepared and stored below 8°C until further analysis. Preparation of blank plasma: Blank plasma (0.2 mL) was transferred into 2.0 ml centrifuge tube and 0.2 mL of precipitating agent (10% perchloric acid) was added. Finally made upto 2 mL with the mobile phase. The resulting solution was vortexed for 5 minutes and centrifuged at 4000 rpm for 10 minutes. The supernatant layer was separated and analyzed. Preparation of bioanalytical calibration curve samples: 0.2 mL of Rosvastatin solutions were transferred to 2.0 mL centrifuge tube respectively, to this 0.2 mL of plasma, 0.2 mL of precipitating agent were added. The resulting solution was vortexed for 5 minutes and centrifuged at 4000 rpm for 10 minutes. The supernatant layer was separated and analyzed. Preparation of plasma samples: Plasma samples (0.2 mL) obtained from study subjects was transferred into 2.0 mL centrifuge tube and 0.2 mL of precipitating agent was added. The resulting solution was vortexed for 5 minutes and centrifuged at 4000 rpm for 10 minutes. The supernatant layer was separated and analysed. Method of analysis: The bioanalytical calibration curve samples and plasma sample solutions were injected with above chromatographic conditions and the chromatograms were recorded. The quantification of the chromatogram was performed using peak area. Preformulation studies In DSC studies (Figure 1) it was concluded that the drug was entrapped into the polymer matrix without any chemical interaction. It was also further concluded that, the polymers were found to be compatible in entrapping the selected drug Simvastatin. The melting endotherm of the mixture proved that the crystalline form of the drug is converted to amorphous form and hence more amount drug solubilised in the lipid. As reported earlier, that rapid quenching of microemulsion does not allow the drug molecules dispersed in lipid phase to crystallize [12]. Furthermore, the presence of surfactants could not allow the drug to crystallize [13,14]. The spectra obtained from FTIR studies at wavelength from 4000 cm -1 to 400 cm -1 are shown in Figure 2. After interpretation of the above spectra it was confirmed that there was no major shifting, loss or appearance of functional peaks between the spectra of drug, physical mixture of drug and lipid (3548.09 cm -1 , 3451.46 cm -1 ). From the spectra it was concluded that the drug was entrapped into the polymer matrix without any chemical interaction. From the IR study it was concluded that, the selected lipid tripalmitin was found to be compatible in entrapping the selected drug Simvastatin. Evaluation of solid lipid nanoparticles The significance of zeta potential is that its value can be related to the stability of colloidal dispersions. The zeta potential (-5.16 mV) indicates the degree of repulsion between adjacent, similarly charged particles in dispersion and confirms the stability (Figure 3). The Polydispersity index (PI) is the measure of size distribution of the nanoparticle formulation. PI was measured using Malvern zetasizer. PI values range from 0.000 to 1.000 i.e. monodisperse to very broad particle size distribution. PI values of all the formulations indicate that particle size distribution was unimodel. The optimized TP-SLN (F6) batch having least particle size 177.4 nm with the PDI 0.650. The lipid core was found to affect the extent of entrapment efficiency and drug loading. As observed with GMS-SLN and TP-SLN, the maximum entrapment efficiency was 78% (Batch F3) and 91.28% (Batch F6) respectively. The entrapment efficiency was higher when drug to lipid ratio was 1:10 (Batch F6) compared to drug to lipid ratio of 1:5 (Batch F3) i.e. when amount of drug taken for preparation of SLN was more (10 mg), higher entrapment was obtained. TP-SLN showed higher entrapment efficiency compared to GMS-SLN because TP is more lipophilic than GMS and can accommodate more drugs in the lipid matrix. These results correlate well with partition coefficient studies. Similar results were obtained in case of drug loading, TP-SLN showed maximum drug loading of 15.15% (Batch F6) compared to GMS-SLN which showed maximum loading of 11.94%. These results correlate well with partition coefficient studies The External morphological studies (SEM) revealed that maximum nanoparticles were nearly spherical in |
shape (Figures 3 and 4). The nanoparticle size observed by SEM correlated well with the particle size measured by zeta sizer (Malvern instrument). In vitro dissolution studies were carried out in Phosphate buffer pH 6.8 and results are shown in Figure 5. The release profiles indicate that SLN formulations showed a retarded release of the drug from the lipid matrix when compared with plain Simvastatin solution (SM-SOL). The in vitro release data and graph of SLN formulations and SM-SOL in Phosphate buffer pH 6.8 is shown in Table 1 and Figure 5 . It was observed that SM-SOL showed 99.12% release in 1.5 h compared to 15.5% and 13.25% release for TP-SLN and GMS-SLN at the end of same time (p < 0.05). This is due to fact that there is no barrier for diffusion at dialysis membrane interface for Simvastatin molecules. Hence, higher release was observed in case of SM-SOL. But, TP-SLN released the drug in a sustained manner. In order to elucidate mode and mechanism of drug release, the invitro data was transformed and interpreted at graphical interface constructed using various kinetic models ( Table 1). The in vitro release data obtained for TP-SLN formulation, in phosphate buffer pH 6.8, was fitted into various kinetic models. The best linearity was obtained in Higuchi's plot for TP-SLN formulation indicating the release from matrix as a square root of time dependent process. The release exponent values 'n' for TP-SLN was found to be 0.619. Since, the release exponent 'n' values were between 0.5-1, it indicates that the SLN formulations undergo anomalous diffusion. Cytotoxicity studies Haemocompatability studies: Nanoparticles were subjected to rigorous blood biocompatibility tests. Erythrocyte-induced haemolysis in vitro can be considered to be a simple and reliable measure for estimating the membrane damage caused in vivo. Percent haemolysis was determined spectrophotometriclly, detecting plasma free haemoglobin derivatives after incubating the particles with blood and then separating the undamaged cells by centrifugation. Typically less than 5% haemolysis is considered acceptable for blood biocompatibility. The results of haemocompatability studies are shown in Table 2. The concentration of 125-1000 μg/ml was subjected to determined percentage haemolysis. In this 1% triton X 100 was used as positive control and Saline water was as a negative control. Positive control showed percentage haemolysis of 3.714 ± 0.65% where as negative control of 0.023 ± 0.001%. On increasing the concentration from 150 μg/ml to 1000 μg/ml, there was no significant increase in% haemolysis.. The results indicated that the nanoparticles were haemocompatible and did not produce any toxic effects. MTT assay: The percentage growth inhibition was calculated using the following formula and concentration of drug or test samples needed to inhibit cell growth by 50% values were generated from the dose-response curves for each cell line. Mean OD of individual test group % Growth inhibition = 100 -×100 Mean OD of control group The IC 50 value of the formulation is about 0.165 mg/ml and IC 50 value of the drug is about 0.23 mg/ml (Figures 6 and 7). Bioanalytical method development and analysis Bioanalytical calibration curve of Simvastatin was prepared in mobile phase and peak area was calculated. The chromatograms of Simvastatin showed stable baseline. The regression equation in the range of 0.5-2 µg/ml was as follows: y=3730x+332.2, R 2 =0.990. The concentration of Simvastatin was determined in plasma samples separated at different time intervals by HPLC analysis. The concentration of plasma samples was determined from the area of the chromatographic peak using the calibration graph. The results of in vivo bioavailability and bio-distribution has been shown in Figures 8 and 9. The enhanced bioavailability by the SLNs formulation might be attributed to direct uptake of nanoparticles through the GI tract, increased permeability by surfactants, and decreased degradation and clearance. Firstly, the uptake of Simvastatin in the SLNs-encapsulated form could be up taken through the GI tract, where the particle size played a dominant role in absorption rate [15]. The mechanisms of such uptake include diffusion of particles through mucus and accessibility to enterocyte surface, epithelial interaction and The size of SLNs in the range of 20 -500 nm allows the efficient uptake in intestine, particularly in the lymphoid sections of this tissue; therefore bypass the liver first-pass metabolism [16]. Secondly, the surfactants, such as Tween 80, have contributed to an increase in the permeability of the intestinal membrane or improved the affinity between lipid particles and the intestinal membrane, and also may exhibit bioadhesion to the GI tract wall [17]. Thirdly, by incorporation into nanoparticles, Simvastatin can be embedded into a solid lipid matrix thus not only reducing its exposure to bacterium as well as enzymatic degradation during absorption process, but also offering a long time contact with the wall of intestine in vivo due to the nice adhesiveness of SLNs to the mucosal surface of intestine. Also, positively charged particles are better taken up by intestinal lymphatics than neutral or negatively charged particles [16,17]. Apical potential of epithelial cells of gastrointestinal tract as well as other cells in the body possess a negative charge on their surface due to the presence of negatively charged proteins on the outer membrane of the cells there by better permeability and uptake would occur for positively charged colloidal particles due to electrostatic attraction between oppositely charged surfaces [17,18]. In addition, Simvastatin-SLNs could provide Simvastatin with a long circulation effect in vivo with sustained-release property, which prolonged the drug residence time in systematic circulation and resulted in better bioavailability [19,20]. In biodistribution study it is clear that SIM-SLNs showed a comparitive better amount of drug sistrubtuion to organs like lymph, prostate and ovary. The drug was bound to the tissues of various organs and the availability of the drug shows that the formulation is targeted to the specific sites. Conclusion In this study, we have described the preparation and characterization of Folate conjugated Simvastatin SLNs. In MCF-7 Human Breast carcinoma cells, FCSIM-SLNs showed greater cytotoxicity than free drug solutions. The SLNs showed a significant increase in oral bioavailability compared to pure drug suspension. Higher relative bioavailability would be due to avoidance of firstpass hepatic metabolism by intestinal lymphatic transport, which circumvents the liver. SLNs provided sustained release of the drugs, and these systems are the preferred drug carriers for lipophilic drugs to overcome the oral bioavailability problem of drug and to increase targeting efficacy to the specific organs as breast, prostate and ovary. Expression and immunogenicity of nsp10 protein of porcine epidemic diarrhea virus Porcine epidemic diarrhea virus (PEDV), a swine enteropathogenic coronavirus, causes lethal watery diarrhea to the piglets, which poses significant economic losses and public health concerns. The nsp10 protein of PEDV is essential regulatory subunits that are critical for virus replication. Since PEDV nsp10 is a crucial regulator of viral RNA synthesis, it is promising that nsp10 might become anti-virus drugs target or candidate for rapid diagnosis of PEDV infection. In this study, the PEDV nsp10 was inserted into pMAL-c2x-MBP / pET-28a vector, efficiently and stably expressed in E.coli system. Then the purified nsp10 protein was found to mediate potent antibody responses in immunized mice. The antibodies of immunized mice and PEDV infection swine strongly recognized purified nsp10 protein from cell lysates. Furthermore, cytokines test revealed that the expression of IL-2, IL-4, IL-10, TNF-α, IFN-γ were significantly higher than those in control group, indicated that purified nsp10 protein induce the cellular immune response mechanism in mice. Using modified seroneutralization test, we also demonstrated that sera from nsp10-immunized mice inhibited PEDV replication to some extent. These findings suggest that nsp10 has a high immunogenicity. This study may have implications for future development of PEDV detection or anti-virus drugs for swine. Introduction Porcine epidemic diarrhea (PED) is a highly contagious viral disease in pigs caused by porcine epidemic diarrhea virus (PEDV) that characterized by severe diarrhea, vomiting, and dehydration with a high mortality in piglets and brought substantial economic losses (Pensaert and Martelli, 2016;Sun et al., 2012;Wang et al., 2016). In China, PEDV was first identified in the 1980s, and in 2010, a large-scale outbreak of PED occurred in China, causing tremendous economic losses (Li et al., 2012;Sun et al., 2012;Tian et al., 2014). PEDV belongs to the order Nidovirales, family Coronaviridae and genus Alphacoronavirus and is an envelop virus with a single-stranded, positive-sense RNA genome. The genome of PEDV is approximately 28 kb and consist of the 5 ′ untranslated region (5' UTR), 3 ′ poly (A) tail, seven open reading frames (ORFs) which include ORF1a, ORF1b, S, ORF3, E, M and N genes (Duarte et al., 1993;Song and Park, 2012). The ORF1a and ORF1b encode two large replicase polyprotein (pp1a and pp1ab), which are subsequently processed into 16 nonstructural proteins (nsp1 to 16) (Subissi et al., 2014). Nsp10 protein is existent exclusively in CoVs so far which is a zincfinger protein through detecting the crystal structure of SARS CoV nsp10 protein (Joseph et al., 2006). Crystallographic or nuclear magnetic resonance structures have shown that nsp10 have the ADP-ribose 1-phosphatase (ADRP) activity and RNA binding activity (Anand et al., 2002). Nsp10 is a crucial regulator involved in viral RNA synthesis and is necessary for viral replication via regulating the nsp14 ExoN and nsp16 2'-O-MTase activities (Bouvet et al., 2010;Bouvet et al., 2012;Donaldson et al., 2007b). The nsp16 S-methyltransferase activity can only be activated when combined with nsp10 as dimer structure (Decroly et al., 2011). In addition, nsp10 also forms a complex with nsp14 to promote the cleavage effect of mismatched nucleotides by nsp14 (Bouvet et al., 2012). Donaldson et al reported that mutant variant of nsp10 inhibits the main protease, 3CLpro, blocking its function completely at the nonpermissive temperature which implicate nsp10 as being a critical factor in the activation of 3CLpro function (Donaldson et al., 2007a). Furthermore, PEDV nsp10 enhances the inhibitory effect of nsp16 on IFN-β production, negatively regulates innate immunity to promote viral proliferation (Shi et al., 2019). Considering these biological features of the nsp10 protein, it would be an appropriate target for developing effective detection tool or drugs against PEDV. In the study, nsp10 protein were constitutively expressed in Escherichia coli (BL21) and explored for its ability to induce immune responses. We found that nsp10 was capable of inducing an efficient antibody response in immunized mice and high expression of cytokines in lymphocytes of mouse spleen. Main reagents RNA extraction kit and SYBR Color qPCR Master Mix was purchased from VazymeBiotech Co.,Ltd. 2 K DNA Marker; 15 K DNA Marker; BamHI, XhoI, EcoRI, HindIII endonuclease; T4 DNA ligase; Taq DNA polymerase and PrimeScript 1st strand cDNA Synthesis Kit were purchased from Takara Bio company. Plasmid extraction kit was purchased from TIANGEN Biotech Co.,LTD. Urea was purchased from Xilong Chemical.14-100 kDa Protein Marker; 10-180 kDa Protein Marker; 5× SDS-PAGE Loading Buffer were purchased from TransGen Biotech. PVDF membrane was purchased from Millipore of Sigma-Aldrich Company. 6× His-Tag Monoclonal antibodies, MBP Rabbit polyclonal antibodies, HRP-Goat anti Mouse IgG and HRP-Goat anti Pig IgG were purchased from Proteintech Group. The ELISA kit was purchased from Huamei Biological Engineering Co. Ltd. NTA Agarose was acquired from QIAGEN of Sigma-Aldrich Company. Freund's complete adjuvant and Freund's incomplete adjuvant was purchased from ThermoFisher Scientific. PEDV nsp10 Bioinformatics Analysis The prediction of nsp10 immune epitopes were performed using the predictor of Bepipred on the IEDB website (http://tools.immuneepitope. org/main/). The putative three-dimension of nsp10 protein was generated using the software Swiss-Model(https://swissmodel.expasy.org/). Plasmid construction Total RNA Extraction Reagent (Vazyme Biotech, Nanjing) was used to isolate the total RNA from PEDV-infected vero cells according to the manufacturer's instructions. The nsp10 was amplified from total RNA of infected cells according to the nsp10 specific primers designed by PEDV strain CV777 (GenBank accession number KT323979.1). The nsp10 genes were separately inserted into the pET-28a vector and pMAL-c2x-MBP vector. The inserted gene fragments of constructed plasmid were confirmed by sequencing. Expression and purification of recombinant protein The recombinant plasmids (pMAL-c2x-MBP-nsp10, pET-28a-nsp10) were transformed into E.coli BL21 (DE3) strain separately. The expression of these proteins was induced by adding isopropyl β-D-1-thiogalactopyranoside (IPTG). After induction, the expression levels in the whole cell lysate, the supernatant and the sediment were examined by SDS-PAGE gel. Purification of recombinant proteins was performed as described previously (Wang et al., 2009). Briefly, cells were pelleted and suspended in 8 M urea buffer, followed by centrifugation at 12,000g, 4 • C for 30 min. |
The supernatant was applied to Ni-NTA affinity column. Purified proteins were eluted with buffers containing different concentrations of imidazole. Samples were stored at − 80 • C for further use. Immunoblotting Immunoblotting was performed as described previously. Briefly, protein samples were separated by SDS-PAGE gel and transferred to polyvinylidene difluoride (PVDF) membranes (Millipore). After protein transfer, the membranes were blocked for 1 h with 10% nonfat dry milk. The blots were then incubated with a primary antibody at 4 • C overnight. The primary antibodies could be the antisera from mice immunized with the nsp10 protein, antisera from healthy or PEDV-infected swine; commercial antibodies against His-tag, or MBP-tag. The membranes were then incubated with HRP-conjugated secondary antibodies: goat anti-mouse or -swine. Finally, the proteins were visualized with Clarity ECL Immunoblotting substrate (Bio-Rad). Mouse immunization and ethics Kunming mice were randomly assigned to the following groups and immunized with 50 μg of the recombinant proteins in the presence of complete Freud's adjuvant (subcutaneous). The mice were subsequently boosted by the same method twice at one-week intervals with incomplete Freud's adjuvant in the subcutaneous regimen. Blood was harvested at different time-points. Mice immunized with PBS served as negative controls. Mouse physiological Status were monitored daily for the following three weeks. Mouse experiments were approved by Animal Care and User Committee and Laboratory Animal Ethics Committee at Jiangxi Agricultural University and performed under the approved guidelines. Recombinant protein-coated ELISA ELISA plates were coated overnight at 4 • C with 50 ng/well of each purified recombinant protein or the cell lysate dissolved in coating buffer (0.016 M Na 2 CO 3 , 0.034 M NaHCO 3 , pH 9.6) followed by the blocked with 5% non-fat milk for one hour at room temperature. After extensive wash with PBST, mouse serum samples were added to wells and incubated at 37 • C for 1 h. After extensive wash, HRP-conjugated goat anti-mouse antibody was added to the wells for one hourincubation at room temperature. Finally, substrates were added to the plate. The absorbance of each well was measured with a Bio-Rad microplate reader at a wavelength of 450 nm. Cytokine profiling in serum and splenocytes Sera and splenocytes were collected from protein-immunized mice. Cytokines including IL-4, TNF-α in serum samples and supernatant from cells treated with purified nsp10 protein were measured by ELISA as described previously. Briefly, the harvested sera were added to the precoated ELISA plate. After incubation for one hour at room temperature, 50 μl of biotinylated antibody was added to each well and incubated for 90 min at 37 • C. After 4 times of wash, 100 μl of streptavidin-HRP was added to each well. The plates were sealed and incubated at 37 • C for 30 min. After adding substrate for development, the absorbance was recorded at 450 nm. The mRNA expression of Cytokines including IL-2, IL-4, IL-10, TNF-α, IFN-γ in splenocytes were detected by Real-time PCR as described previously (Matsuda, 2017). In brief, Gene expressions were quantified using TB Green®Premix Ex Taq™II (Takara Biotechnology, China) and the qRT-PCR was performed in an Applied Biosystems® 7500 Fast Real-time System. The primers were specific designed using Primer Premier 5.0 (Table 1). Seroneutralization test and Real-time quantitative RT-PCR The seroneutralization experiment was proceeded as previously described with minor modifications (Gauger and Vincent, 2014;Li et al., 2020). Vero cells were seeded with 5 × 10 4 cells per well into 96-well plates at least 6 h prior to infection. The test samples were serially diluted 2-fold from 1:2 to 1:256 with DMEM, eight wells were used for each dilution. Then the serial dilution of sera samples were applied to cells and incubated for 1 h at 37 • C. After that, the sera samples were removed and cells incubated with 200 TCID 50 of PEDV at 37 • C for 1 h. The cytopathic effect was examined for 2 days post-infection. The wells infection with 200 TCID 50 , 20 TCID 50 , 2 TCID 50 of PEDV were considered as virus-infection control. For standardization of the absolute quantification assay, the cDNA of PEDV N gene was inserted into pCAGGA-HA vector and the plasmid DNA isolated from positive clone DH5α. qRT-PCR was carried out in a 20-μl reaction mixture with 2 μl cDNA or standard plasmid, 10 μl 2 × SYBR qPCR Master mix, 0.2 μM forward and reverse primers. Amplification and detection were performed in Applied Biosystems® 7500 Fast Real-time System. The specific primers of PEDV N are listed in Table 1. Statistical analysis Data analysis was performed using GraphPad Prism 8.0 software to determine statistical significance. Statistical differences between two groups were tested with Student's t-test. For all the tests, P < 0.05 was regarded as the statistical significance. Prediction of potential antigenic epitopes and three-dimensional (3D) structure of PEDV nsp10 protein Schematic diagram of the genome of PEDV is shown in Fig. 1A. PEDV nsp10 protein sequence was subjected to Bepipred and Swiss-Model to forecast its linear B cell epitopes and 3D structure, respectively. Using a threshold value 0.5 for Bepipred, 5 potential B cell epitopes were found (Fig. 1B). It is speculated that main B cell antigen epitope of nsp10 may appear in the 25 to 31, 40 to 63, 77 to 82, 94 to 97, 120 to 124 amino acid residues (Fig. 1C). B cell antigen epitopes are shown in the Swissmodeled nsp10 protein structure (Fig. 1D). Altogether, these results indicate that PEDV nsp10 protein may harbor potential for inducing host immune responses. Construction of recombinant prokaryotic expression plasmid By using PEDV cDNA as template, specific primers were designed for PCR amplification. The fragment of 408 bp was identified by 1.5% agarose gel electrophoresis, which was consistent with the expected size ( Fig. 2A). After PEDV nsp10 gene was ligated with prokaryotic expression vector pET28a, pMAL-c2x-MBP by double digestion of BamH I, EcoR I, Xho I, Hind III, the ligation product was transformed into E. coli competent cell DH5α. A number of positive clones were randomly selected for PCR identification, and the amplified fragment was 408 bp (Fig. 2B). The correct clone strains identified by PCR were selected for plasmid extraction (Fig. 2C and E). The recombinant plasmid pMAL-c2x-MBP-nsp10 was identified by double enzyme digestion (Fig. 2D). Then the recombinant plasmid pET28a-nsp10 was identified by PCR using nsp10 specific primer and T7 primer respectively (Fig. 2F). The positive clone was sent to Sangon for sequencing. The sequence of PEDV nsp10 was identified based on a multiple alignment of PEDV CV777 strain. The sequence of positive clone was confirmed through DNAMAN sequence comparison. The recombinant expression vector pMAL-c2x-MBP-nsp10, pET28a-nsp10 was successfully constructed. Expression and purification of recombinant PEDV nsp10 protein To obtain nsp10 protein, the recombinant plasmid was transformed into E. coli BL21 (DE3) competent cells, positive clones were randomly selected and expressed. The bacteria were collected and examined by SDS-PAGE gel, several positive clones were successfully expressed. The MBP protein in vector pMAL-c2x-MBP was successfully expressed, and the relative molecular weight of the MBP-nsp10 and nsp10 protein was 56kD and 16kD respectively. To obtain the optimal condition of protein expression, we tested different induction temperature (Fig. 3A (caption on next page) concentration. In addition, the best expression temperature of pET28a-nsp10 was at 40 • C (Fig. 3C left panel), and with the increase of temperature, the protein expression level showed a trend of first increasing and then decreasing. Then nsp10 protein was expressed in different concentration IPTG to confirm optimal concentration of IPTG. In summary, the best condition of pET28a-nsp10 expression was: 40 • C, 1.4 mM IPTG concentration (Fig. 3C right panel) induced by 8 h (Fig. 3C middle panel). The MBP, MBP-nsp10 and nsp10 protein was purified by Ni-column and eluted by 250 mM imidazole (Fig. 3D). To confirm the identities of E.coli-expressed recombinant proteins, we performed immunoblotting assays with the commercial anti-His tag antibody or anti-MBP antibody. Expected bands were detected for all recombinant proteins in IPTG-induced bacterial lysate samples, but not un-induced ones, demonstrating the identities of E.coli-expressed proteins (Fig. 3E). Immunogenicity analysis of PEDV nsp10 protein in mice After administration of nsp10 protein, antibody titers of mice sera were determined by ELISA. Sera of mice immunized with PBS solution was used as negative control. The results showed that antibody titers of nsp10 protein immunized mice were the highest at the fourth week, 1:43000 of MBP-nsp10 (Fig. 4A right panel) and 1:13000 of nsp10 ( Fig. 4B left panel) respectively. In addition, we tested the antibody titers of MBP protein and found that the highest titer was at the third week, 1:31000. At the same time, we harvested sera from proteinimmunized mice to examine the immunogenicity of these proteins. Immunoblotting results showed that mice sera specifically detected the corresponding recombinant proteins expressed in IPTG-induced transformed E.coli, but not in the un-induced samples (Fig. 4C). As a negative control, sera from mice treated with PBS did not recognize any recombinant proteins. We further evaluated the nsp10-specific immunogenicity with the sera from PEDV-infected pigs. The data from immunoblotting revealed that nsp10 recognized PEDV-positive pig sera but not PEDV-free pig sera (Fig. 4C right). Moreover, 30 sera samples from healthy or PEDV-infected pigs were collected, then investigated blindly by using commercial ELISA kit or nsp10 protein. The results showed that there is a high degree of consistency between PEDV ELISA kit and nsp10-ELISA (Fig. 4D). In summary, these data showed that the recombinant nsp10 protein expressed in E.coli retains the antigenicity that sera from PEDV-infected pigs can recognize. Cellular responses analysis of PEDV nsp10 protein in mice Cellular immune responses play an essential role in the defense against viruses (Benova et al., 2020). we examined levels of Th2 (IL-4) cytokines in mouse sera upon immunization with nsp10 protein. The results showed that the treatment of nsp10 protein, but not PBS, triggered mice to secrete cytokines substantially (Fig. 5A). After splenocytes were harvest from mice immunized with nsp10 protein and stimulated by purified nsp10 protein of different concentration in vitro, levels of Th1 (TNF-α) from cell supernatant were examined by ELISA after different incubation time. As shown in Fig. 5B, nsp10 protein triggered more secretions of cytokine TNF-α, compared with PBS treatment group. To further evaluate the cellular immunity at mRNA expression level, expression of cytokine IL-2, IL-4, IL-10, TNF-α, IFN-γ in splenocytes from immunized mice were detected by Real-time PCR assay. The relative expression levels of all genes were calculated by the 2 -△△Ct algorithm. The internal reference was GAPDH, which is a housekeeping gene. Compared with PBS immunized group, the expression level of cytokine in the nsp10-immunized group was significantly increased (Fig. 5C). All the results above collectively demonstrate that nsp10 is capable of initiating cellular immunity. Mice serological protection assay and absolute PEDV quantitation A seroneutralization test was performed to determine whether the sera antibody from mice immunized with nsp10 had protection activity. We incubated cells with serum dilutions (final dilution from 1:4 to 1:256) of immunized mice sera, then the cells were infected with 200 TCID 50 of PEDV. We observed the cytopathogenic effect for 2 days and the partial absence of CPE in >50% of the wells was defined as protection. As shown in Fig. 6A, 2 3 dilution of serum can block CPE in 50% of eight repetitions wells. The findings suggested that nsp10 antibodies may protect cells from PEDV infection. To further detect the absolute quantity of the viral load in serum-treated cells, the N gene expression of PEDV was detected by qRT-PCR as the plasmid pCAGGS-HA-N was used for the generation of a standard curve (Fig. 6B). The result showed that the copy numbers of PEDV RNA from nsp10 antibodies serum-treated cells were decreased compared with the PEDV-infection cells (Fig. 6C). The results presented here suggested that nsp10 antibodies from mice sera may inhibit PEDV replication to some extent. Discussion The coronavirus nsp10 contains two zinc-finger motifs (Matthes et al., 2006) and may function as part of a larger RNA-binding complex (Su et al., 2006). Previous researches have shown that coronavirus nsp10 is a critical multifunctional factor for activation of multiple replicative enzymes. Nsp10 molecules may thus serve as a platform for the recruitment of nsp14 or nsp16 to the replication-transcription complex in order to either stimulate nsp14 ExoN activity or activate nsp16 2-O-MTase activity (Snijder et al., 2016). Both |
nsp10-nsp14 and nsp10-nsp16 complexes may coexist which play essential role in different steps of viral RNA synthesis as part of larger nsp assembling factory (Bouvet et al., 2014). Considering its importance for virus replication, nsp10 represents an attractive target for anti-coronavirus (A-B) KunMing mice (6-week old) were administrated subcutaneously with 50 μg of MBP, MBP-nsp10, nsp10. Antisera were collected and subjected to ELISA for detecting nsp10 antibody. (C) The lysates from un-induced or IPTG-induced cells and purified protein were used to react with corresponding protein-immunized mouse sera, PEDV-positive pig sera and control sera in the immunoblotting assays. (D) Comparison of nsp10-coated ELISA and commercial PEDV ELISA kit. Sera samples from PEDV-infected or mock-infected pig were test by nsp10-coated ELISA (left) and commercial ELISA kit (right). Each point represents a serum sample, and the line represents the cutoff value (n = 30). drug development (Lin et al., 2020;Wang et al., 2015). Specific molecules or antibodies targeting the nsp10 and inhibiting the interaction with nsp14 and nsp16 might be applied to reduce or prevent virus proliferation (Ke et al., 2012). Because the nsp10 subunit is among the more conserved coronavirus proteins, such compounds or antibodies could be used in a prophylactic approach to prevent coronavirus infection just like the role of vaccine. The current focus of PEDV nsp10 research is on the interaction of nsp10 with nsp14 and nsp16 to regulate innate immunity Shi et al., 2019). In the present study, the first aim was to stably express the full-length nsp10 protein of PEDV in E.coli system. Subsequently, we were able to successfully generate a stable E.coli system producing large amounts of nsp10 protein through optimization of expression conditions (Fig. 3A-C). Following the purification and concentration processes, nsp10 protein could be consistently harvested. Then mice were immunized with nsp10 protein and we determined whether or not they developed humoral immunity. Nsp10-specific antibodies were strongly detectable in mice sera collected from 4 weeks post-inoculation (Fig. 4A-B). PEDV-infected pig serum can recognize the E.coli-expressed nsp10 protein, indicating that it has good immunogenicity (Fig. 4C). Compared with commercial PEDV ELISA kit, nsp10-ELISA had high degree of consistency to detect sera samples of PEDV-infected or uninfected pig (Fig. 4D). It is suggested that PEDV nsp10 may serve as the candidate for developing PEDV detection kit. PEDV is the main causative pathogen of viral diarrhea in piglets which could cause rapid destruction and detachment of the intestinal physical structure (Wang et al., 2018). Over the years, a number of works has been done to investigate PEDV pathogenesis and prevention. Vaccination against PEDV is an important measure to prevent and control the rate of PEDV infection (Chen et al., 2017). However, due to the continued mutation of PEDV genome (Lin et al., 2016), the current available vaccines cannot provide complete protection for pigs. Therefore, it is important that PEDV infection can be efficiently detected when outbreaks of disease occur or effective vaccine will be developed. Whole virus vaccines are mainly traditional vaccines of PEDV such as live attenuated vaccines and inactivated vaccines. For live attenuated vaccines, the risk of reversion to high virulence strain limits its applicability, although live attenuated vaccines possess highly immunogenic (Lauring et al., 2010;Yong et al., 2019). As to inactivated vaccines, there are several advantages including good safety profile and ease production. However, incomplete inactivation and multiple immunizations are some disadvantages that cannot be neglected (DeZure et al., 2016). Compared with whole virus vaccines, subunit vaccines represent promising approaches because of high safety, no contagious viral nucleic acids, uniform antigen (Du et al., 2016;Yap and Smith, 2010). The S protein of PEDV is main target used for developing subunit vaccines since it can induce neutralizing antibodies and interfere virus entry into host cells (Makadiya et al., 2016). Non-structural protein of coronavirus does not provide protection as neutralization antibodies against virus infection. However, antibodies against non-structural protein also can provide protective immunity by Fcγ receptor-mediated viral clearance, complement-mediated cytotoxicity and complement-independent phagocytosis (Rastogi et al., 2016;Reyes-Sandoval and Ludert, 2019;Wan et al., 2021). In addition, non-structural protein can trigger cellular immune response that protects against virus infection (Grubor-Bauk et al., 2019;Mishra et al., 2020). Subsequently, accumulating data have emerged as to develop non-structural protein-based vaccines (Bailey et al., 2018;Lin et al., 1998). In addition to non-structural protein-based vaccines, non-structural protein-specific antibodies would also be expected to be used for developing antiviral drugs that inhibit virus replication. Research reported that PRRSV nsp9-specific nanobodies that delivered into MARC-145 cells and PAMs can efficiently inhibited the replication of several PRRSV strains (Liu et al., 2015;Wang et al., 2019). In this study, the results showed that expression level of cytokine IL-2, IL-4, IL-10, TNF-α, IFN-γ in splenocytes or sera from immunized mice increased, indicating that nsp10 can promote cellular immunity ( Fig. 5B-C). Modified seroneutralization test indicated that nsp10 antibodies from immunized mouse sera may protect cells from PEDV infection (Fig. 6). In addition, we aware of several limitations of this study, which all experiments involving the immunogenicity of PEDV nsp10 were carried out in mice. As pigs are the natural host of PEDV, in vivo studies on the protective immunity of PEDV nsp10 are certainly required in the future. In conclusion, full-length nsp10 amplified from CV777 strain was expressed in E.coli, which protein yields were considerable high. Immunization challenge studies in mice, demonstrate that PEDV nsp10 induced robustly humoral and cellular immune responses. Furthermore, nsp10 antibodies from immunized mouse sera may have protective effect on cells from PEDV infection. Despite these potential limitations, we believe that the nsp10 protein has potential for use in developing effective and safe anti-PEDV drugs for PED prevention. protein respectively. The culture medium was collected after incubation for 24 h, 48 h, 72 h, then detected cytokines (TNF-α). (C) The RNA of splenocytes from immunized mice was extracted and subjected to Real-time PCR for detecting the expression level of cytokines IL-2, IL-4, IL-10, TNF-α and IFN-γ. Declaration of Competing Interest There were no potential conflicts (financial, professional or personal) that are relevant to the manuscript. A Practical Approach to Quantitate Hepatic Excretory Function 1 A statistical comparison of different BSP tests was carried out in normal subjects and in patients with various degrees of chronic liver damage. Only the logarithm of the BSP retention correlated linearly with the physiologically more meaningful determination of the maximal excretory capacity of the liver (BSP Tm, Wheele's method). A double logarithmic transformation was required to correlate the second exponential component of the BSP plasma disappearance curve (k2) with the BSP Tm. When the limitations of these methods are kept in mind, the observed statistical relationships can be used to express hepatic functional deterioration in more physiological terms. In patients with pulmonary or renal disease it is an established procedure to use quantitative function tests to follow the progression of the disease and/or to assess the effects of various forms of treatment. By contrast, analogous approaches to studying patients with liver diseases have been used only rarely. The need to develop quantitative tests of hepatic function has not been pressing because specific and effective therapy has not been available. Recently, however, clinical trials have accumulated evidence to suggest that chronic active liver disease may be successfully treated with corticosteroid hormones (1)(2)(3). Consequently quantiative clinical tests to measure partial hepatic functions have become more desirable. With the presently available sulfobromophthalein-or BSP-tests-as in many other areas of medicine-there is an inverse relationship between simplicity of the procedures and specificity of the information they can give (Table 1). Indeed, only the measurement of the hepatic transport maximum for BSP can truly be considered to represent a quantitative expression of hepatic excretory function (4). In practice, the test based on Wheeler's method is cumbersome and time-consuming and, therefore, seems ill-suited for the clinical routine. By contrast, the coventional BSP retention test is widely used and generally accepted. It is technically simple and empirically quite sensitive for the detection of liver disease (5) BSPi.;. -BSP(xer. VD The equation shows that C45, is directly proportional to the difference of the amounts of BSP injected (BSPi,,j.) and excreted (BSPQ\Cr.) and inversely proportional to the volume of distribution of BSP (VD). The injected quantity of BSP is standardized to 5 mg/kg body wt. Only the proportion excreted by the liver is related to hepatic function. Since, however, the concentration of BSP offered to the liver varies continuously throughout the test, it cannot be expected that the amount of BSP excreted during 45 min bears any simple relationship to the excretory capacity of the liver. The well-established complexities of hepatic handling of BSP further complicate the problem: Apart from BSP binding to plasma and intrahepatic acceptor proteins, at least conjugation and reflux into the plasma have to be considered as well. The volume of distribution appearing in the denominator is larger than the plasma volume, and its relation to disease processes is not adequately defined. These aspects, therefore, limit interpretation of the results to an empirical and statistical basis. The third test used to assess liver function is an analysis of the BSP plasma disappearance curve. If frequent samples are taken, the curve may often be described as a double exponential function (6,7). Despite the formulation of kinetic models (8), the physiological basis for these curves has remained insufficiently understood. Consequently, studies in our department were directed to further evaluate the BSP retention and the BSP plasma disappearance curve (5,9). In a group of patients with various liver diseases and in normal control subjects, the three tests have been carried out in one session. A comparison of the results revealed that the simple tests may be interpreted in terms of specific partial functions. The BSP disappearance curves were measured after the intravelnous injection of 5 mg/kg of BSP within 30 sec. Blood samples were obtained at 3, 5, 7, 10, 15, 20, 25, 30, 35, 40, and 45 min after the beginning of the injection. The last sample was used to calculate BSP retention assuming an initial concentration of 10 mg/100 ml. The BSP Tim was determined indirectly with the two-infusion method as described by Preisig et al. (10). The principles of this method are based on the idea, that with the chosen experimental conditions, the intravenously infused BSP is either distributed within its volume of distribution-plasma volume (PV) and hepatic storage compartment (S)-or excreted at the rate (Tm) which is maximal for the liver to be studied (4). The distribution of BSP can be calculated as product of the rate of change in plasma concentration (AC/At) and the volume of distribution (PV + S). Consequently, the equation may be written as: Infusion = Tm + Ac/At (PV + S). As only the rate of infusion and the plasma concentrations can be measured there remain two unknowns (Tm and S). Two different infusions, therefore, have to be administered in order to produce two equations with two unknowns, which then can be solved. The practical application of the test requires consideration of many details which have been discussed elsewhere (4,11 ). The first finding of our comparative analysis consisted of the empirical fact that only the logarithms of the BSP-retention were linearly correlated with the excretory capacity of the liver as revealed by the BSP Tm (Fig. 1). This type of relationship implies that the degree of abnormality revealed by the BSP retention has no direct proportionality to the functional deterioration of a diseased liver. The higher the measured BSP retention, the more sensitive it is in revealing changes of excretory function. This observation led to a definition of new BSP retention units. They were constructed to express linearly functional deterioration of the excretory system and can be read directly from the nomogram of Fig. 2. The normal mean value was regarded as zero retention. The range up to 1.0 retention unit was taken as normal, whereas the higher values correspond to directly proportional decreases in the excretory capacity of the liver for BSP. Application of such retention units to clinical problems should always take into account, however, that they were de- fined on a statistical and not a pathophysiologically recognized relationship. Their justification, therefore, remains empirical. A similar analysis was carried out with the BSP plasma disappearance curve. Its most important finding consisted in the close correlation between the logarithms of the second component (k2) of |
the disappearance curve (Fig. 3) and the BSP transport maximum, suggesting that k2 is mainly determined by the excretory capacity of the liver (Fig. 4). In support of this assumption there was no significant relationship between k2 and hepatic blood flow (r = -0.08) as measured with the ICG-infusion and extraction technique (12). Physiologically, ko could be explained on the basis of a two-compartment model, where, in addition, the actual BSP plasma concentrations had to be taken into account. Even though the details, especially the mathematical derivation, are described elsewhere (9), the model may be regarded as an adequate basis to jutify the use of k2 for estimations of BSP Tm. In practice, this test requires only four determinations of BSP plasma concentrations, 30, 35, 40, and 45 min after the injection of BSP. The value of k2 is calculated from the graphically read half-life (T1I2) provided three of the four points can be accurately fitted to a straight line (k2 = 0.693/T,1/2). When this condition is not fulfilled, the resulting k2 is inaccurate and should not be used. In our experience, this difficulty occurred in 39% of normal subjects but only in 10% of patients with impaired function. Consequently this procedure is more suitable to assess degrees of abnormality than to define a normal population. The values for k2 obtained with this method, can be converted to "estimated" BSP Tm values with the aid of another nomogram, depicted in Fig. 5. The specific conditions suggested for the determination of k2 were found by trial and error. The failures to get a straight line between 30 and 45 min after injection of BSP in a high percentage of control subjects were due to the relatively rapid disappearance of BSP from the blood leading to low plasma concentrations which were difficult to measure accurately enough. In patients with impaired liver function, a hump in the curve occasionally rendered k, inaccurate. Severe cholestasis or the Dubin-Johnson syndrome usually were associated with a ko approaching zero. The secondary rise in plasma BSP concentration often seen in these cases generally occurred after 45 min and consequently did not interfere with the assessment of k2. If further studies on a larger scale confirm the results obtained in our laboratory, the use of BSP k2 and of BSP-retention units should be valuable in the follow-up of patients treated for chronic liver disease. Both tests are simple enough for the clinical routine and give information about quantitative aspects of the excretory function of the liver. Because of its better established physiological basis, the BSP k2 is directly applicable to the individual patient and, therefore, may be preferred. Detecting Overlapping Outbreaks of Influenza ntroduction Influenza is a contagious disease that causes epidemics in many parts of the world. The World Health Organization estimates that influenza causes three to five million severe illnesses each year and 250,000-500,000 deaths [1]. Predicting and characterizing outbreaks of influenza is an important public health problem and significant progress has been made in predicting single outbreaks. However, multiple temporally overlapping outbreaks are also common. These may be caused by different subtypes or outbreaks in multiple demographic groups. We describe our Multiple Outbreak Detection System (MODS) and its performance on two actual outbreaks. This work extends previous work by our group [2,3,4] by using model- averaging and a new method to estimate non-influenza influenza-like illness (NI-ILI). We also apply MODS to a real dataset with a double outbreak. Methods MODS is part of a framework for disease surveillance developed by our group. In this framework, a natural language processing system extracts symptoms from emergency department patient-care reports. These features are combined with laboratory results and passed to a case detection system that infers a probability distribution over the diseases each patient may have. These diseases include influenza, NI-ILI, and other (appendicitis, trauma, etc.). This distribution is expressed in terms of the likelihoods of the patients’ data. These are given to MODS which searches a space of multiple outbreak models, computes the likelihood of each model, and calculates the expected number of influenza cases day-by-day. This work differs from past work in three important ways. First, we address the problem of detecting and characterizing multiple, overlapping outbreaks. Second, we do not rely on simple counts, but use likelihoods given evidence in the free-text portion of patient-care reports as well as laboratory findings. Third, we explicitly account for non-influenza influenza- like illnesses. This is important because some forms of influenza-like illness (such as respiratory syncytial virus) are contagious and exhibit outbreak activity. This research was approved by the University of Pittsburgh and Intermountain Healthcare IRBs. Results We conducted a set of experiments with simulated outbreaks. MODS is able to detect a single outbreak six to eight weeks before the peak. It is also able to recognize a second outbreak approximately halfway between peaks for simulated double outbreaks. We conducted experiments using real outbreaks and compared our results to thermometer sales [5]. Using data from Allegheny County Pennsylvania for the 2009-2010 influenza season, on September 1 MODS predicted an outbreak with a peak on October 5. The thermometer peak was October 21. The figure “Prediction on October 1 for Allegheny County” compares MODS’ prediction on October 1 to thermometer sales. Using data from Salt Lake City Utah for the 2010-2011 influenza season, on November 1 MODS predicted an outbreak with peak on December 7. The first thermometer peak was December 29. On January 20 MODS predicted a second outbreak with peak on February 9. The second thermometer peak was March 5. The figure “Prediction on January 20 for Salt Lake City” compares MODS’ prediction on January 20 to thermometer sales. Conclusions We have built a Multiple Outbreak Detection System that can detect and characterize overlapping outbreaks of influenza. Although the system currently predicts outbreaks of influenza, it is built on a general Bayesian framework that can be extended to other diseases. Future work includes incorporating multiple forms of evidence, modeling other known contagious diseases, and detecting outbreaks of new previously unknown diseases. Prediction on October 1 for Allegheny County 2009-2010 Prediction on January 20 for Salt Lake City 2010-2011 Introduction Influenza is a contagious disease that causes epidemics in many parts of the world. The World Health Organization estimates that influenza causes three to five million severe illnesses each year and 250,000-500,000 deaths [1]. Predicting and characterizing outbreaks of influenza is an important public health problem and significant progress has been made in predicting single outbreaks. However, multiple temporally overlapping outbreaks are also common. These may be caused by different subtypes or outbreaks in multiple demographic groups. We describe our Multiple Outbreak Detection System (MODS) and its performance on two actual outbreaks. This work extends previous work by our group [2,3,4] by using modelaveraging and a new method to estimate non-influenza influenza-like illness (NI-ILI). We also apply MODS to a real dataset with a double outbreak. Methods MODS is part of a framework for disease surveillance developed by our group. In this framework, a natural language processing system extracts symptoms from emergency department patient-care reports. These features are combined with laboratory results and passed to a case detection system that infers a probability distribution over the diseases each patient may have. These diseases include influenza, NI-ILI, and other (appendicitis, trauma, etc.). This distribution is expressed in terms of the likelihoods of the patients' data. These are given to MODS which searches a space of multiple outbreak models, computes the likelihood of each model, and calculates the expected number of influenza cases day-by-day. This work differs from past work in three important ways. First, we address the problem of detecting and characterizing multiple, overlapping outbreaks. Second, we do not rely on simple counts, but use likelihoods given evidence in the free-text portion of patient-care reports as well as laboratory findings. Third, we explicitly account for non-influenza influenzalike illnesses. This is important because some forms of influenza-like illness (such as respiratory syncytial virus) are contagious and exhibit outbreak activity. This research was approved by the University of Pittsburgh and Intermountain Healthcare IRBs. Results We conducted a set of experiments with simulated outbreaks. MODS is able to detect a single outbreak six to eight weeks before the peak. It is also able to recognize a second outbreak approximately halfway between peaks for simulated double outbreaks. We conducted experiments using real outbreaks and compared our results to thermometer sales [5]. Using data from Allegheny County Pennsylvania for the 2009-2010 influenza season, on September 1 MODS predicted an outbreak with a peak on October 5. The thermometer peak was October 21. The figure "Prediction on October 1 for Allegheny County" compares MODS' prediction on October 1 to thermometer sales. Using data from Salt Lake City Utah for the 2010-2011 influenza season, on November 1 MODS predicted an outbreak with peak on December 7. The first thermometer peak was December 29. On January 20 MODS predicted a second outbreak with peak on February 9. The second thermometer peak was March 5. The figure "Prediction on January 20 for Salt Lake City" compares MODS' prediction on January 20 to thermometer sales. Conclusions We have built a Multiple Outbreak Detection System that can detect and characterize overlapping outbreaks of influenza. Although the system currently predicts outbreaks of influenza, it is built on a general Bayesian framework that can be extended to other diseases. Future work includes incorporating multiple forms of evidence, modeling other known contagious diseases, and detecting outbreaks of new previously unknown diseases. ISDS Annual Conference Proceedings 2017. This is an Open Access article distributed under the terms of the Creative Commons Attribution-Noncommercial 3.0 Unported License (http://creativecommons.org/licenses/by-nc/3.0/), permitting all non-commercial use, distribution, and reproduction in any medium, provided the original work is properly cited. Cognitively and socially induced stress affects postural control Postural control is an adaptive process that can be affected by many aspects of human behavior, including emotional contexts. The main emotional contexts that affect postural control are postural threat and passive viewing of aversive or threatening images, both of which produce a reduction in postural sway. The aim of the present study was to assess whether similar stress-related changes in postural sway can be observed using stress induced by social evaluative threat (SET) while performing arithmetic tasks. Twelve young adults performed an arithmetic and a postural control task separately, concurrently, and concurrently with added time pressure in the arithmetic task. In the final condition, participants were given negative feedback about their performance in the arithmetic task and performed it again while being observed (SET condition). Results showed that stress increased linearly with task demand. Postural sway and reaction times were not affected by the first two conditions; however, when time pressure was introduced, reaction times became faster and sway amplitude increased. Finally, introduction of SET caused the predicted reduction in postural sway and an increase in reaction times relative to the time pressure condition. Our results suggest that stress induced using a combination of arithmetic tasks and social evaluative threat leads to systematic changes in postural control. The paradigm developed in the present study would be very useful in assessing interactions between cognition, stress, and postural control in the context of postural instability and falls in older adults. Introduction Maintaining upright stance, or postural control, is an adaptive process that relies on sensory, motor and cognitive processes (Balasubramaniam and Wing 2002) and can also be affected by emotional contexts (for a review, see Hagenaars et al. 2014). An emotional context that has been extensively studied in postural control is postural threat, which is a situation imposing increased challenge on balance. Postural threat has been primarily assessed in a series of studies by Carpenter and colleagues (Carpenter et al. 1999(Carpenter et al. , 2001Adkin et al. 2000) who showed that when participants stood on the edge of an elevated surface (0.8-3.2 m high), real, or virtual (Cleworth et al. 2012); they exhibited a reduction in postural sway and posterior body movement away from the platform's edge compared with standing at ground level. This sway reduction was accompanied by an increase in sway frequency, muscle co-contraction, and ankle stiffness. Standing on an elevated surface also induced increases in fear of falling, sympathetic arousal, and stress response (Cleworth et al. 2012;Horslen et al. 2014). More recent studies focused on the underlying mechanisms of |
this adaptive response and showed that postural threat is likely to induce changes in two of the sensory channels involved in postural control, proprioceptive (Davis et al. 2011;Horslen et al. 2013), and vestibular (Horslen et al. 2014;Lim et al. 2017). A reduction in body movement as a response to emotionally engaging, fear-inducing stimulation, however, is not unique to postural-threat contexts. Fear responses and their underlying mechanisms have been studied in animals and have been classified into two categories: defensive action, characterised by fight or flight behavior in response to impending attack, and defensive immobility, characterised by freezing, bradycardia, and hyper-attentiveness (Lang et al. 2000). For example, fear-freezing, measured as the amount of time rats in a cage were not moving has been observed in the context of Pavlovian-conditioned aversive responses using an auditory stimulus (LeDoux et al. 1988). Using fear-freezing responses in animals as a starting point, freezing has also been assessed in humans using a paradigm comprising passive viewing of aversive or threatening images, for example, images of mutilation (Azevedo et al. 2005;Facchinetti et al. 2006). Similar to posturalthreat research, these studies showed a reduction in sway, an increase in mean power frequency of sway but also bradycardia, the latter also being associated with freezing responses in animals. Similarly, Roelofs et al. (2010) showed a freezing response in reaction to angry faces, reflected in reduced heart rate and body sway, and Hillman et al. (2004) showed a backward body movement away from unpleasant stimuli in women which was not observed in the case of pleasant or neutral stimuli. This response was also affected by the previous experience of an aversive life event, with individuals who had experienced such an event showing a greater reduction in sway when exposed to aversive images compared with a control group (Hagenaars et al. 2012). However, another study manipulating arousal and valence using this paradigm (Horslen and Carpenter 2011) showed that only arousal affected postural sway similar to postural threat, and identified methodological limitations in some of the passive-viewing studies. The primary limitation involved the short duration of the postural trials (< 10 s) in some of these studies (Hillman et al. 2004;Stins and Beek 2007;Roelofs et al. 2010), which was not long enough for the full range of time scales present in postural sway time series to be identified (Van der Kooij et al. 2011). Despite the methodological limitations identified in studies of freezing in humans, the consensus in the literature is that fear-related postural responses can be induced by at least two different types of emotion-specific paradigms, postural threat, and passive viewing of aversive or threatening images. Given that these two paradigms induce similar sway reduction responses, it would be reasonable to assume that this reduction is a general effect that goes beyond the two paradigms and could also be caused by other emotion-specific manipulations. In the psychology literature, fear-related emotional responses are primarily triggered using socially induced stress and anxiety (Dickerson and Kemeny 2004). A well-established, effective method of inducing high levels of stress in humans is the presence of social evaluative threat (SET), in tasks primarily including mental arithmetic, public speaking, and singing (Kirschbaum et al. 1993;Frisch et al. 2015). SET is characterised by emotional responses observed during tasks performed in circumstances, where an evaluative audience or a negative social comparison is present. For example, SET has been successfully used to induce stress in combination with a mental arithmetic task, the Montreal imaging stress task (Dedovic et al. 2005). This study asked participants to perform arithmetic calculations and used a mock performance indicator combined with negative feedback by both the task software (after each trial) and the experimenter (between blocks, which was the SET element). When negative feedback was provided, an increase in cortisol was observed relative to the control and rest conditions, suggesting an increase in stress. Furthermore, a metaanalysis of over 200 studies showed that SET, together with uncontrollability, induced the largest increases in cortisol levels and the longest times to recovery compared to all other stressors (Dickerson and Kemeny 2004), thus making SET a very effective way of inducing stress in humans. The high effectiveness of this method makes it an excellent candidate to use as a novel manipulation to induce stress in the context of postural control, to see whether posturalthreat-and passive-viewing-induced sway reduction can also be observed using SET-induced stress. Furthermore, performance of a mental arithmetic task while standing has been shown to affect both postural control and physiological arousal as measured by skin conductance (Maki and McIlroy 1996). Together, evidence suggests that SET in combination with mental arithmetic causes an increase in stress (Dedovic et al. 2013) and that mental arithmetic tasks affect physiological arousal and postural control (Maki and McIlroy 1996). However, a combination of these three tasks has not been previously used to induce stress in a postural control context. The aim of this study was to assess whether a reduction in sway is observed when stress increases using a combination of mental arithmetic and SET manipulations. To this end, first, we assessed postural control and arithmetic separately and then concurrently. Subsequently, to increase stress and task demands, we added an element of time pressure by introducing a progress bar based on participants' own performance in the arithmetic task. It was expected that under time pressure, participants would allocate more resources to the arithmetic task in an attempt to perform it more efficiently. Finally, to increase stress and task demand further, we added an SET manipulation, providing negative performance feedback. This approach was used to incrementally increase task demands and stress in each condition. This incremental increase was used to ensure that we could assess the contribution of each component of our design (adding the arithmetic task, time pressure, and SET) on stress, postural control, and performance on the arithmetic task. Theories of anxiety and cognition, such as the processing efficiency theory (Eysenck and Calvo 1992) and its successor, attentional control theory (Eysenck et al. 2007), predict that during relatively simple tasks, performers can compensate for anxiety-related inefficiencies in processing information through increasing cognitive effort. However, as task demands increase, such inefficiencies can no longer be compensated for and performance starts to show deficits. We predicted that self-rated stress would increase incrementally with task demand. More importantly, we predicted that this increase in stress following the SET manipulation would be accompanied by a reduction in postural sway reflecting a stiffening or freezing response in line with the previous research, (Adkin et al. 2000;Azevedo et al. 2005;Facchinetti et al. 2006;Roelofs et al. 2010). These findings would suggest that the reduction in postural sway observed when using postural threat and aversive images can also be induced using cognitively and socially induced stress. Performance on the arithmetic task was expected to improve following the inclusion of a time pressure manipulation, due to recruitment of additional resources to, or prioritization of the arithmetic task. However, we predicted that performance in this task would eventually decline under high levels of stress and task demands (e.g., Brooks 2014) as a consequence of anxiety-related reductions in processing efficiency. More specifically, we predicted that only in trials containing increased task demands and associated higher stress (i.e., trials including SET), would individuals demonstrate significant reductions in postural sway in conjunction with reduced performance in the cognitive task. Participants Twelve adults, eight females, and four males (mean age 20.7, SD 1.9 years) volunteered to participate in this study. Participants were undergraduate psychology students and received course credit for their participation. They reported no major neurological or musculoskeletal disorders and no intake of medication that affects postural control, for example, sleep medication or antidepressants (de Groot et al. 2013). All participants provided written informed consent and the study was approved by the School of Psychology, Queen's University Belfast ethics committee. Apparatus and tasks The study comprised a postural control task and an arithmetic task performed separately and concurrently. In the postural control task, participants were asked to maintain upright standing, barefoot with eyes open on the dual force plates of a Neurocom Smart Balance Master system (Natus Medical inc.) in 2-min blocks. Stance width was determined by the system's manufacturer and was 26 cm for height < 165 and 30 cm for height > 165 cm. During the task, participants were asked to wear a safety harness which ensured safety in case of loss of stability. No losses of stability were observed in this study. In all posture blocks, the standing surface was tilted in the anterior-posterior (toes down, toes up) direction in proportion to body sway, using support-surface sway reference (Black et al. 1983;McCollum et al. 1996). Sway reference is a well-established method of increasing postural sway by means of reducing proprioceptive information about body sway (Peterka and Black 1990;Peterka and Loughlin 2004;Doumas and Krampe 2010). For example, when participants sway 1 degree forward, the surface was tilted 1 degree down thereby maintaining a relatively constant ankle angle and inducing inaccurate proprioceptive information about body sway. In the present study, we used sway referencing in the AP direction with a gain factor of 2, in which body sway of 1° induces surface tilt of 2°. We used a higher gain to induce larger AP sway magnitude, because in young adults with eyes, open sway referencing at gain 1 is not as effective in inducing an increase in postural sway compared with eyes closed (Peterka and Black 1990;Clark and Riley 2007). We opted for using this method instead of quiet standing on a fixed surface, because we aimed to induce a certain amount of sway first, to increase the possibility of detecting a reduction in sway when SET was introduced. Vertical forces applied by the body on the force plates were recorded at a sampling frequency of 100 Hz and were used to derive the center of pressure (COP) time series in the anterior-posterior (AP) direction. Participants performed an arithmetic verification task (Fig. 1a) both seated and standing. When seated, stimuli were displayed on a 19″ computer screen (resolution: 1024 × 768 pixels), located 60 cm in front of the participant and 100 cm from the ground and when standing, they were displayed on a 17″ monitor positioned at the participants' eye level, with the same resolution, embedded in the Neurocom system's surround (Fig. 1b). The task was presented with and without added time pressure. In the case of presentation without time pressure, stimuli were presented as white numbers (font size 48) on a black background and comprised a simple addition of a single-digit to two-digit number, or of two three digit numbers, followed by the sum (e.g., 14 + 3 = 17 or 342 + 539 = 881). Two types of problems were used to make the task less repetitive. A new set of problems was used in each block. Participants were asked to respond by pressing the left button of a wireless mouse if the sum was correct and the right if it was incorrect, as quickly and as accurately as possible. Arithmetic task stimuli were presented in blocks comprising 32 trials each, equally divided into correct and incorrect sums, and into smaller (1 + 2 digits) and larger (3 + 3 digits) pairs of numbers. If no response was provided after 10 s, the trial was interrupted and the next trial was presented. In the case of presentation with time pressure, stimulus presentation was the same, but a progress bar was introduced at the bottom of the screen (Fig. 1a), starting with a green part, followed by a dividing white line and a longer red part, with a constant total duration of 10 s. The bar was only present in Blocks 4 and 5 (Fig. 1b, c). To induce time pressure, the duration of the green part of the bar was determined as 70% of the RT in the last block without time pressure (Block 3, see Fig. 1c). After each block, except the first and last posture blocks, participants were asked to provide a 0-10 rating of how stressed they felt (0-not at all, 10extremely stressed). Procedure Participants responded to an advert for an experiment on 'Effects of maths ability on balance' and were not aware of the SET element of the study. The experimental session was carried out by two experimenters. After obtaining written informed consent |
and demographic information from the participant, one experimenter left the room. The experiment's timeline is depicted in Fig. 1c. Participants were first asked to perform a 2-min posture trial without the arithmetic task (Block 1). Then, they sat at a table and performed a practice block of 6 trials of the arithmetic task without time pressure followed by a full block of 32 trials of the same task (Block 2) to establish a baseline performance in this task. This was followed by performing the posture task and the arithmetic task concurrently without the progress bar (Block 3) and then by performing the posture and arithmetic tasks concurrently with the progress bar (Block 4). The green part of the bar was determined as 70% of the participant's mean Reaction Time of correct responses in Block 3; however, participants were told that the green part was determined by the average RT of the previous participants. At the end of Block 4, the Social Evaluative Threat manipulation was implemented. The experimenter inside the room left the room for a few seconds and returned with the other experimenter who pretended to look at the data on a screen that the participant could not see. One of the experimenters then turned to the participant, informed them that they were underperforming in the arithmetic task and asked them to perform better in the next block (Block 5), which comprised the same tasks as the previous block (Block 4) but with a new set of arithmetic problems. Both experimenters closely observed the participant performing the task in this block. Following the SET manipulation, participants performed the three initial blocks again in reverse order to account for practice and possible fatigue effects (Fig. 1c). All participants performed this series of blocks in a fixed order (Blocks 1-8). In the end of the experiment, participants were fully debriefed both in writing and verbally about the purpose of the experiment. Data analysis In the arithmetic task, we calculated accuracy as the percentage of correct responses and reaction times (RT). In the postural control task, COP data from the Neurocom system in the AP direction were low pass filtered at 4 Hz using a fifth-order Butterworth dual-pass filter. The first 5 s of each 2-min block were considered as a stabilisation phase and were discarded. In blocks including simultaneous arithmetic and posture performance, only posture data for the time during which both tasks were performed were analyzed. This was because it was not possible to synchronise the two tasks to end at the same time without changing the number of stimuli in the arithmetic task or the length of the balance trial, given that in participants exhibiting faster RTs, the 32 trials ended earlier. Postural sway was assessed by calculating COP AP and ML amplitude (the maximum-minimum position of the COP), standard deviation (SD), and mean power frequency (MPF). However, we did not expect changes in sway in the ML direction in this task, because sway reference was applied only in the AP direction. Indeed, no differences were observed for ML in neither of the three variables; thus, we only report results in the AP direction. Data in blocks that were repeated before and after the SET manipulation (Blocks 1 and 8, 2, and 7, 3, and 6) were averaged to account for practice or fatigue effects. In addition, to assess the presence of practice or fatigue effects, we contrasted performance in these pairs of blocks using pairedsamples t tests. Our main analyses used repeated measures analyses of variance (ANOVAs) with condition (arithmetic/ posture only; posture and arithmetic; posture, arithmetic and time pressure; and posture, arithmetic, time pressure, and SET) as the within-subjects variables. We were interested in the way that our variables were affected by the addition of each level of task demands, starting with the balance task and adding the arithmetic task, the progress bar and finally SET. Thus, we performed planned comparisons only between successive conditions using pairwise t tests corrected for the three comparisons (α = 0.017). Results In this section, we report statistical analyses for self-rated stress, postural control (AP amplitude, SD and MPF), and performance in the arithmetic task (RTs and accuracy). Self-rated stress Results for self-rated stress are depicted in Fig. 2. A repeated measures ANOVA for condition (arithmetic only; posture and arithmetic; posture, arithmetic, and time pressure; and posture, arithmetic time pressure, and SET) showed that stress increased with condition as shown by a main effect of condition F(3,33) = 47.15, p < 0.001, η 2 = 0.81. Planned comparisons contrasting successive conditions showed that all three differences were significant [ts(11) = 4.24, 5.15, 3.22, all ps < 0.008]. Comparisons of blocks repeated before and after the SET manipulation showed that self-rated stress was not different between the two arithmetic only blocks (Block 2: mean 3.25, SD 1.66; Block 7: mean 3.5, SD 1.83; t(11) = 0.9, p = 0.4) and the two balance and arithmetic blocks (Block 3: mean 4.5, SD 1.83; Block 6: mean 4.7, SD 1.1.78; t(11) = 0.3, p = 0.77). Postural control Postural sway AP amplitude, SD, and MPF results are depicted in Fig. 3. The ANOVA for AP amplitude (Fig. 3A) showed a main effect of condition F(3,33) = 5.55, p = 0.003, η 2 = 0.335. Planned comparisons contrasting successive conditions (Fig. 3a) showed no differences between the posture only and the posture and arithmetic conditions, but showed an increase in amplitude when time pressure was introduced t(11) = 3.41, p = 0.006, followed by a decrease in amplitude from the time pressure to the SET condition t(11) = 3.54, p = 0.005. Contrasts between blocks repeated before and after the SET manipulation showed that AP amplitude was not different between the two posture only blocks [Block 1: mean 9.28 cm, SD 4.25 cm; Block 8: mean 7.6 cm, SD 4.78 cm; t(11) = 1.6, p = 0.14], but amplitude was greater after the SET manipulation compared with before [Block 3: mean 5.94 cm, SD 3.37 cm; Block 6: mean 7.31 cm, SD 4.14 cm; t(11) = 2.23, p = 0.047]. AP SD showed a similar pattern of results (Fig. 3b) and a main effect of condition F(3,36) = 6.42, p = 0.002, η 2 = 0.37. Pairwise comparisons, however, showed that the only reliable difference was the decrease in SD between Arithmetic The two measures of the arithmetic task, accuracy and RT (Fig. 4) were analyzed using separate repeated measures ANOVAs with condition as factor (arithmetic only; posture and arithmetic; posture, arithmetic and time pressure; and posture, arithmetic time pressure, and SET) followed by planned comparisons. Accuracy (Fig. 4a) showed a decrease with condition as shown by a main effect of condition F(3,33) = 5.83, p = 0.003, η 2 = 0.35; however, none of the pairwise comparisons between successive blocks were significant [ts(11) = 1.66, 0.7, 1.88, ps = 0.13, 0.5, 0.09]. Contrasts between blocks repeated before and after the SET manipulation showed that accuracy was not different between the two arithmetic only blocks [Block 2: mean 90.37%, SD 10.78%; Block 7: mean 89.84%, SD 11.24%; t(11) = 0.24, p = 0.8] but increased between the two posture and arithmetic blocks [Block 3: mean 81.25%, SD 18.46%; Block 6: mean 93.75%, SD 6.53%; t(11) = 2.55, p = 0.027]. Reaction time (Fig. 4b) showed similar values between sitting and standing without the progress bar. However, when the bar was introduced (posture, arithmetic, and time pressure condition) reaction times became faster, but became slower again after SET was introduced (bar2 condition). This pattern of results was confirmed by a main effect of Error bars represent ± SEM. *p < .017 compared to the previous condition condition F(3,33) = 25.47, p < .001, η 2 = 0.70, followed by planned comparisons which confirmed the faster RTs when time pressure was first introduced t(11) = 6.08, p < .001, and then the slower RTs when SET was introduced t(11) = 6.26, p < .001. Comparisons between blocks repeated before and after the SET manipulation showed that RTs were not different in the beginning and in the end of the experiment [Block 2: mean 3153.8, SD 560.11; Block 7: mean 3479.9, SD 1092.04; t(11) = 1.31, p = 0.2] but faster RTs were observed before and after the SET manipulation [Block 3: mean 4116.78 ms, SD 1350.38 ms; Block 6: mean 2471.98, SD 684.22 ms; t(11) = 4.47, p < .001]. Discussion The aim of this study was to assess whether a reduction in sway is observed when stress increases using a combination of mental arithmetic and SET manipulations. Results showed that when SET was introduced in the context of an arithmetic task, postural sway amplitude and SD decreased, confirming our main hypothesis. To achieve a sufficient increase in stress, we used a gradual introduction of task demand comprising postural, cognitive, and SET elements and we showed that stress increased with task demand in our four conditions. We also assessed whether performance in the arithmetic task was affected by the gradual increase in task demand. As predicted, RT became faster when the progress bar was introduced, suggesting that the time pressure manipulation was successful. However, when SET was introduced RT became slower, suggesting that the high stress induced by the SET impeded performance in the arithmetic task. This increase in RT was accompanied by a decrease in accuracy, although this effect did not reach significance. This study is the first to demonstrate that SET results in a reduction of postural sway. Our results suggest that the emotional contexts that affect postural sway may not only be postural threat and passive viewing of aversive images but also stress induced using cognitive and SET contexts. We show a reduction in sway with SET, which is in line with studies on postural threat (Carpenter et al. 1999(Carpenter et al. , 2001Adkin et al. 2000) and aversive or threatening images (Azevedo et al. 2005;Hagenaars et al. 2012;Roelofs et al. 2010), suggesting that the reduction in sway in the present study could be evidence for a common 'freezing' or 'stiffening' mechanism in response to increased threat, anxiety, and stress. However, it is important to acknowledge that to make a strong claim about a stiffening or freezing strategy, it is critical to also show an increase in sway frequency, but our results did not show this increase. Furthermore, we also observed an unexpected increase in sway amplitude when time pressure was introduced. However, both these findings can be interpreted in the context of cognitive resource allocation, and its well-established interactions with postural control (Boisgontier et al. 2013) and stress (Eysenck and Calvo 1992). There is a large body of evidence (for a review, see Boisgontier et al. 2013), suggesting that postural control relies on cognitive resources. This reliance is primarily used by older adults-who use cognitive resources to prioritize balance using a 'posture first strategy' to avoid falls-but also by young adults (Maylor et al. 2001;Doumas et al. 2008a;Smolders et al. 2010;Doumas and Krampe 2015). In the present study, we used an approach similar to the previous posture-cognitive dual-task studies. However, we introduced increasing cognitive task demands as a way to increase stress (Pruessner et al. 2008;Dedovic et al. 2013). Thus, due to the involvement of both cognitive resource allocation and stress in this study, our results can be interpreted in the context of theories of anxiety and cognition, namely, processing efficiency theory (Eysenck and Calvo 1992). According to this theoretical framework, during relatively simple tasks, performers can compensate for anxiety-related inefficiencies through increasing cognitive effort. However, as task demands increase, such inefficiencies can no longer be compensated for and performance starts to show deficits. We argue that this theoretical approach can be used to interpret our results for at least the first three conditions. In the first two conditions, we observed an increase in stress, but this increase did not affect performance in posture or arithmetic tasks, because the increase in stress-related attentional inefficiencies was compensated for by greater cognitive effort. However, in the third condition, when task demands increased by means of introducing time pressure, stress increased further, along with a reduction in RT and increased postural sway. This result suggests that when time pressure was introduced, additional cognitive resources were directed to the cognitive task to accommodate the increased |
task demands. Finally, when SET was introduced cognitive and postural task demands were the same but stress increased further, thereby exacerbating anxiety-related processing inefficiencies (Eysenck and Calvo 1992). This increase in stress could not be compensated for through the allocation of cognitive resources, and as a result, RTs became longer and the postural control system triggered a reduction in sway resembling a freezing or stiffening strategy similar to the previous studies. The question remains how to interpret changes in balance control during the most challenging condition. It is possible that reductions in sway could be purely due to arousal-based physiological stiffening adaptations. However, this explanation appears incomplete due to the absence of changes in sway frequency and the further observation that adaptations to sway did not change in line with increased self-reported stress. A more likely explanation could be that reduced sway represents a shift of resources back towards postural control in an attempt to restrict unwanted destabilizing movements. This suggestion is supported by increases in RT in the most demanding condition. Our study had a number of limitations. Stress was primarily measured using self-report. Support for our choice of self-report arises from recent evidence showing that when standing on an elevated surface, self-reported fear is positively correlated with physiological changes such as changes in vestibular reflex gains (Naranjo et al. 2016). However, the previous studies assessing emotional contexts in postural control have used both self-report and physiological measures of emotion and arousal. For example, studies assessing postural threat and fear of falling have primarily used electrodermal activity which measures skin conductance, showing that it is a reliable measure of postural threat (Cleworth et al. 2012;Osler et al. 2013). On the other hand, studies using passive viewing of threatening images used heart rate, based on the idea from the animal literature that freezing is accompanied by bradycardia (Azevedo et al. 2005;Hagenaars et al. 2012;Roelofs et al. 2010). Furthermore, studies assessing stress using social evaluative threat have used cortisol measures (e.g., Dedovic et al. 2013). The addition of one or more of these measures, especially heart rate, would have been useful in strengthening the results of the present study in terms of whether our findings are in line with a freezing response in response to SET. In addition, subjective task difficulty, and the subjective effort allocated to the different aspects of the task was not measured. A final limitation of our study was that, due to the gradual nature of our stress manipulation, it was not possible to randomise the order of conditions. Specifically, we could not introduce the progress bar before participants performed the arithmetic task, or the SET task before the progress bar. We believe that this incremental increase in task demands was a strength of our paradigm; however, the fixed order of conditions was likely to induce practice effects which could confound our results. This is the reason why we used an experimental design which assessed individual tasks first (posture or arithmetic), then a combination of tasks and then individual tasks again (ABBA design) similar to our previous studies assessing posture-cognitive dual-task performance (Doumas et al. 2008b(Doumas et al. , 2009(Doumas et al. , 2012Smolders et al. 2010;Doumas and Krampe 2015). However, this solution may not be ideal, because averaging may conceal differences between first and last blocks. Thus, to exclude the possibility of practice effects, we contrasted performance from similar blocks (1 and 8, 2 and 7, and 3 and 6) in all variables. Indeed, results showed that postural sway amplitude and variability did not show improvement over the course of the experiment, neither in the comparison between the two posture only blocks in the beginning and in the end of the experiment (Block 1 vs 8) nor between the two posture and arithmetic blocks (Block 3 vs 6). This lack of sway reduction through practice suggests that the reduction in sway we observed as a result of SET cannot be attributed to practice effects. Fear, anxiety, and stress are emotions that are highly prevalent in daily life. For example, math anxiety is more prevalent in women than men (Dowker et al. 2016;Suárez-Pellicioni et al. 2016) and it would be instructive to assess, in a larger sample than in the present study, whether the induced math anxiety using our paradigm and the subsequent changes in postural control are different in women and men. Furthermore, identifying the manner in which emotion and balance interact is important in addressing balance problems and falls in older adults. Older adults at high risk of falling often self-report higher rates of fear of falling as well as movement self-consciousness (Wong et al. 2009), which, in turn, is related to a postural stiffening strategy (Huffman et al. 2009;Zaback et al. 2015). Therefore, using the approach developed for the present study could be used to gradually increase stress in older adults to assess postural responses in a controlled and safe laboratory setting. It is possible that older adults' postural control will be more sensitive to stress manipulations compared with young adults', especially when SET is introduced. Using this paradigm in an age-comparative context would also be very useful in assessing age differences in the relationship between anxiety or fear of falling and associated stiffening. Future research could also assess whether the well-established anxiety-related reduction in sway under stress is adaptive or maladaptive in terms of whether it increases or decreases susceptibility to a fall in young but also in older adults. In general, stiffening is considered a conservative strategy used to prepare the body for a potential disturbance (Young and Williams 2015). It is thought that such adaptations are likely to be beneficial during relatively static balance tasks, but maladaptive in situations, where dynamic responses or preplanned movements are required. Therefore, future research needs to assess whether this stiffening or freezing strategy observed under stress protects a person from a fall after a real-life perturbation, for example, when a bus starts to move or when a person is 'bumped' in a crowded room. Further research in this direction is needed to identify ways in which emotional contexts affect postural control in real-life settings and fall-prone populations. distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. Designing an Integrated Care Initiative for Vulnerable Families: Operationalisation of Realist Causal and Programme Theory, Sydney Australia Introduction: In July 2015 Sydney Local Health District (SLHD) implemented an integrated care initiative for vulnerable families in the Inner West region of Sydney, Australia. The initiative was designed as a cross-agency care coordination network that would ensure that vulnerable families: had their complex health and social needs met; kept themselves and their children safe; and were connected to society. We will describe the development of the design that drew on earlier realist causal and program theoretical work. Methods: Realist causal and program theory were used to inform the collaborative design of an initiative for vulnerable families. The collaborative design process included: identification of desirable and undesirable outcomes and contextual factors, stakeholder consultation, interagency planning, and development of a service proposal. Results: The design elements included: identification of vulnerable family cohorts; care coordination; evidence-informed intervention(s); general practice engagement and support; family health improvement; placed-based neighbourhood initiatives; interagency system change and collaborative planning; monitoring of individual and family outcomes; and evaluation. Conclusions: The design study described advances toward the implementation of a whole-of-government integrated health and social care initiative. The initiative was designed as a cross-agency care coordination network that would ensure that vulnerable families: had their complex health and social needs met; kept themselves and their children safe; and were connected to society. In so doing we aim to break intergenerational cycles of poverty, violence and crime, poor education and employment opportunities, psychopathology, and poor lifestyle and health behaviours, through strengthening family resilience, improving access to services, and addressing the social determinants of health and wellbeing. In July 2015 Sydney Local Health District (SLHD) implemented an integrated care initiative for vulnerable families in the inner west region of Sydney, Australia. The initiative was designed as a cross-agency care coordination network that would ensure that vulnerable families: had their complex health and social needs met; kept themselves and their children safe; and were connected to society. The initiative was part of a New South Wales (NSW) Government Integrated Care Strategy initiated in March 2014 and drew on the analysis of an earlier design study for vulnerable families undertaken from 2010 to 2014 [2]. The objective of the NSW Integrated Care Strategy was to transform the delivery of care to improve health [and wellbeing] outcomes for patients and reduce costs deriving from inappropriate and fragmented care, across hospital and primary care services. The NSW Government indicated that this objective was to be achieved by: a) focusing on organising care to meet the needs of targeted patients and their carers, rather than organising services around provider structures; b) designing better connected models of health [and social] care to leverage available service providers to meet the needs of our smaller rural communities; c) improving the flow of information between hospitals, specialists, community and primary care providers; d) developing new ways of working across State government agencies and with Commonwealth funded programs to deliver better outcomes for identified communities; and e) providing greater access to out-of-hospital community-based care, to ensure patients receive care in the right place for them. The NSW Government's stated policy intent was to respond to the challenges of an ageing population and increased numbers of people living with chronic and complex conditions by investing in new, innovative models of integrated care, transforming the health system to routinely deliver person-centered, seamless, efficient and effective care, particularly for people with complex, long term conditions. Funding was made available, through a competitive grant process, for health districts to develop innovative integrated care projects. The functional elements of the NSW Health Integrated Care Strategy are shown at Table 1. The Inner West Sydney District collaborated to design initiatives for vulnerable families in 2013. A critical realist design method was applied and included programme theory developed from an analysis of our realist causal propositions [3]. The resulting Theory of Change (ToC) Logic Model included three design components, namely: sustained health home visiting; family and community integrated service development; and infrastructure support [3]. The three design elements were further developed for inclusion in: 1) a Working with Vulnerable Families Business Case, and 2) Child and Family Health Planning Priorities ( Table 2). Importantly the design included population, system and individual-level health and social care elements. During 2014 the Inner West Sydney District collaborative planning for child and family health and wellbeing was extended to focus on improved outcomes for all children and their families. At the time of the launch of the NSW Integrated Care Strategy in March 2014, the planning process had identified four strategic themes, namely: improving system capacity; health and wellbeing promotion; early intervention and supporting place-based initiatives. In addition a detailed outcome framework had been developed based on earlier studies of child and family population-level indicators [4,5]. This paper will describe the design development for an integrated care initiative for vulnerable families including children. The design will draw on our earlier causal and programme theoretical work, the 2013 collaborative design for vulnerable families, and the 2014 NSW Government policy framework for integrated care [2,3,6]. The research is part of two ongoing programmes of research and programme development that seeks to 1) build and confirm a "Theory of Neighbourhood Context, Stress, Depression, and the Developmental Origins of Health and Disease (DOHaD)" [7]; and 2) strengthen the delivery of well child care through mixed method theory building and the application of interagency policy and program interventions. The work was undertaken during 2014 and 2015 with the integrated care initiative commencing July 2015. Theory and Methods As previously described [6,8,9], the overall research design is a longitudinal, multilevel, critical realist design and evaluation of applied programme interventions that seek to break intergeneration cycles of social disadvantage and poor child health and development outcomes and strengthen delivery of well child care. Intervention initiatives were designed and implemented by interagency and community collaborations. In doing so we aimed to move from "explaining underlying social mechanisms |
to generate social interventions in partnership with the affected populations" [10]. The main research programme will consist of four phases (Figure 1). The methodology used for the four phases is reported separately [9]. In summary the four phases are: 1) operationalisation of programme theory and intervention development and planning; 2) evaluation of the interventions; 3) theory testing studies; and 4) dissemination of the findings. In this paper we report on one of the collaborative design projects undertaken in Phase 1: Operationalisation. The operationalisation of causal and programme theory is briefly described in the Results section with Table 4. The full analysis will be reported separately. Critical realism and programme design As noted above, critical realism provides the philosophical and methodological underpinning to this programme of work. Critical realism seeks to discover the structures (C) and underlying mechanisms (M) that cause empirically observed patterned events or outcomes (O). These events are tendencies that result when certain conditions exist, or remain unrealised if the conditions are absent. Examining the pre-existing structural conditions of a context is therefore important. Critical realism also holds that mechanisms, in natural and social reality, are stratified. Depression (event), in the strata of the self, is governed by physical, biochemical and psychological mechanisms and laws. These mechanisms are not governed by laws • The implementation of processes and systems that ensure the integrated care plan meets the needs and preferences of patient/carers as defined by patients or carers themselves (shared decision making). Using patient reported measures in care delivery • The implementation of a system of patient reported measures for enrolled patients that measure both the patient's perceptions of both their care experience and their outcomes, due to the care that they receive. • This includes the timely provision of the information to clinicians/team delivering care to enable shared care planning/shared decision making. Supporting and promoting self-management • A set of defined care interventions specific to the targeted patient cohort to support self-management. • This also includes strategies to increase capacity for patients and carers to better self-manage their condition. Building patient/carer health literacy • The implementation of processes and systems (such as training and information) that improve the patient's understanding of their health condition(s), how to maximise their ability to manage it themselves, how/when to access health services and what role they play in managing their health condition(s). • This also includes care plan access, and active participation to the extent possible in care planning. Patient identification and selection Defining local health needs • The set of local health system parameters which broadly identify the types of patients that require the implementation of an integrated care pathway to improve the effectiveness of healthcare delivery (such as potentially avoidable hospital admission, ED presentations, delays in receiving specialist treatment). Identifying target cohorts • Patient level parameters (such as demographic, e.g. age; clinical, e.g. diagnosis; utilisation, e.g. number of medications; other, e.g. measure of social disadvantage) that define the group of patients that will be targeted/enrolled in the integrated care program. Developing systematic approaches to risk identification • The standardised approach to risk identification (such as signs of health deterioration) and methodology (such as automated processes in PAS/EMR/EHR) for identification of the targeted cohort of patients who would benefit from an integrated model of care. • The targeted risks and cohorts can vary locally, and can vary over time within locality as programs mature. Innovative ways of working Establishing new business models • The identification and implementation of business models across the continuum of care are being to promote care delivery which improves patient care and experience through improved coordination and integration. • The models sit alongside service models (which operationalize service delivery). • They potentially incorporate financial and/or non-financial elements. • The models may include the selection of alliance partners (such as GPs, NGOs or other government organisations) and investment in new roles, as well as the use of known business models (such as Person Centred Medical Homes or a Commissioning Framework). Ensuring appropriate and timely access to specialist care • Needs for the identified cohort. • The function may be achieved in a number of different ways (for example, quarantining appointments in hospital based clinics or purchasing services from a telehealth provider). Shared/joint care planning and management with the patient/carer • The development of shared or joint care planning and care management strategies between the initiator of the care plan, the patient, and other healthcare professionals who are to be involved in the care and service delivery to targeted patients. Establishing roles focused on organising patient-centred care • The establishment of roles (such as case managers, care navigators, care facilitators) to support the implementation of the integrated care model of care across care settings (such as hospital, primary care, specialist care, community care). Embedding agreed models of care • The uptake of models of care for patients with specified conditions that are based on evidence based medicine and adhered to by those clinicians seeing targeted patients. • This includes the process of designing and agreeing the models with stakeholders to optimise uptake. (Contd.) operating at the level of social activity, but are nevertheless affected by them. Critical realist theories therefore, may explain event mechanisms by antecedent causes, or explain mechanisms operating at one level by those operating at a more basic level. A higher-level mechanism, is said to be emergent from a more basic mechanism. Lay-der [11] illustrated this layering of reality in his Research Map (Figure 2). This study uses a modification of Layder's levels, namely, Self, Situated Activity, Intermediate Level and Macro Level. Mechanisms, emergence, a hierarchy of levels, and pre-existing historical conditions are all central to the critical realist design process described here. Functional Component Key Feature Primary and Community care as the hub Connecting people to their healthcare team • The assignment of targeted patients to a clinical provider (individual/practice) whose role is to be the lead clinical provider with responsibility for the shared care plan and initiating communication with other care providers (such as specialist, GP, aged care, community care). Systematic assessment, review of patients • The implementation of a system of standardised assessments, regular patient reviews, and uploading of relevant clinical metrics by clinical care providers based on developed integrated care pathway protocols. Building capacity/capability in primary and community care • The enhancement of resources (such as care navigators, training programs, care pathways, share care planning tools) in the primary and community care settings to support integrated care delivery to targeted patients. Information Sharing Establishing a trackable cohort list • The establishment of an electronic patient list/register that identifies all patients enrolled in the integrated care initiative and enables the monitoring of the patient journey, as reflected through the patient's use of healthcare services. Establishing shared access to patient information • The extent of electronic patient information on enrolled patients available to clinicians across care settings who are delivering the agreed integrated model of care (such as care plans, e-referral, discharge summaries, medication profiles, test results, service events). Realist causal propositions are expressed in terms of mechanisms (M), context (C), and outcomes (O). The MCO propositions in our previously reported theory [1] are in the MCO form proposed by Danermark and colleagues [12]. For evaluation studies, Pawson and Tilley [13] have proposed a CMO configuration. In realist programme evaluation terminology the mechanism (M) is an intervention mechanism (IM), and not a causal mechanism. Denyer and colleagues [14] draw attention to the importance of specifying the intervention separate from the mechanism and proposed the use of a CIMO-logic (Context, Intervention Mechanism, Outcome). Thus a CIMO is a hypothesis that the programme theory produces a change (O) because of the action of an intervention (I) on an underlying mechanism (M) operating in particular contexts (C). We will use the CIMO logic in this study and will apply it to the development of the Theory of Change (ToC) logic model (Figure 4). Realist programme evaluation usually starts with a programme that has been already designed. The approach assumes that whenever a programme is implemented it is testing an existing programme theory consisting of realist programme hypotheses (CMOs). The process of designing a programme intervention using realist causal and programme theory is not well explicated. For the purposes of this study we have drawn on the work of Keller and colleagues [15] who present a realist design-evaluation framework that combines design theory and realist evaluation. Collaborative Design The collaborative design of the integrated care initiative involved: 1) consultation and planning forums; 2) shared outcome planning; 3) collaborative interagency planning; and 4) preparation of a fully funded business plan, ToC and logic model. The development of a ToC using collective and collaborative processes can be difficult [16]. We used the set of steps proposed by Mackenzie and Blamey: 1. Identification of the long-term outcomes that the initiative seeks to achieve 2. Identification of the interim outcomes and contextual features that will be required to meet these longer-term outcomes 3. Specification of the activities that will be put into place and the contextual requirements to realise these interim outcomes 4. An explicit recognition of the resources that will be required to turn these goals into reality. The design analysis integrated our earlier causal and programme theoretical work [6], analysis of local child and family population-level indicators [5]; the 2013 collaborative design for vulnerable families [2]; and the 2014 NSW Government policy framework for integrated care. Consequently the critical realist theoretical framework guiding the collaborative effort incorporated 5 elements: 1. Historical analysis of the context to theorise the preexisting social structures and mechanisms [17] 2. Proposed design elements of an intervention, stemming from inputs from forums, realist syntheses, interviews and collaborations during 2013 and 2014 [2, 6] 3. Proposed design elements arising from an analysis of the NSW Government Integrated Care Strategy 4. The development of a programme theory hypothesising the pre-existing situational conditions and causal mechanisms, and specifying how the proposed intervention would trigger desired psychological, motivational and behavioural responses to bring about change [18] 5. The construction of Theory of Change (ToC) logic model explicating a proposed implementation theory [18]. Ethics The planning undertaken here did not include human subjects. Ethical approval was not sought. The indicator reports used secondary data and did not require ethics approval. The earlier cited mixed method multilevel studies had ethics approval from the University of New South Wales. Results and Design The proposed design elements for the integrated care initiative drew on: 1) collaborative design processes undertaken for vulnerable families in 2013 and the 2014 planning for a five-year child health and wellbeing plan; and 2) the NSW Ministry of Health Integrated Care Strategy ( Table 3). Shared identification and intake: A populationbased approach was proposed for identifying the most vulnerable families, developing cross-agency assessment and referral pathways, and improving hospital maternity services' recognition of the needs of families. The Pathways to Care component included: strengthening of perinatal screening and coordination systems; establishment of a centralised intake system; development of primary care referral pathways using the New Zealand Canterbury HealthPathways Sydney tools; development of a shared family risk assessment tool; building of a electronic medical record tool to support clinicians identifying families; and the building of a database to support the monitoring of outcomes among identified family members. Care coordination: The design provided for the establishment of a nurse-led family care coordination service model that would support families over a long time period with the intention of bridging the episodic nature of existing family support services. The proposed role was to: a) provide leadership and support in the building of local service networks and referral pathways for vulnerable families; b) liaise with and support service providers to ensure referral to appropriate services in accordance with shared care plan; c) coordinate and track service provision for the identified vulnerable families, including ongoing information coordination for providers; d) provide information, support and referral services for members of the enrolled families. The intention was to make use of all available local government and non-government resources to support the complete needs of families. The component included the trialling of "Patchwork", a carecoordination digital web-based support tool developed in the UK by FutureGov [19]. Evidence informed practice: The overall design included evidence-informed elements: nurse-led sustained health home visiting, wrap-around care, family group conferencing, child |
and family service centres, and targeted and multimodal parenting programmes. Sustained nurse home visiting was implemented in parallel with the elements described here and provided an important new service for vulnerable families with infants less than 2 years of age. In addition to including evidence-informed elements in the design, the initiative included the establishment of Knowledge Translation Networks that would identify and promote the use of evidence-informed practice. General Practice engagement and support: the engagement of families with a general practice, and supporting those general practices, was identified as a priority. The intention was to encourage families to have a general practice "health home". An important objective of this component was to reduce emergency department and hospital admission for ambulatory and primary care sensitive conditions. The Australian Commonwealth funded Medical Benefits Scheme (MBS) is premised on the central coordinating role of family physicians in general practice. This component of the design sought to ensure that families had access to all available financial subsidies for their health and social care. Family Health Improvement: The purpose of this component was to strengthen the delivery of public health and preventive health measures to families through the service network and general practice. Sector capacity building projects were proposed that would build on concurrent local initiatives including: a parenting communication initiative ("Love Talk Sing Read Play") and a well childcare sector capacity building project. The intention was to also include preventive health measures into family care plans. Health protection measures for families were also identified as important. The design made provision for enhanced immunisation and healthy housing initiatives. Place-based initiatives: The design assumed that for service integration to be successful it needed to be locally derived within a well organised primary and community sector. The design proposed two placed-based initiatives within the cities of Sydney and Canterbury-Bankstown. The design called for the trialling of wrap-around care in the place-based projects. The full nature of the local initiatives was not detailed in the design with intention of ensuring that they would be locally developed through community and consumer consultation. System Change: The development of the integrated care design had occurred in the context of a strong cross agency collaborative partnership. The partnership was at that time developing a five-year Child Health and Wellbeing Plan and the integrated care initiative provided an opportunity to demonstrate many of the system change elements included in that plan. The integrated care initiative proposed that opportunities for shared planning, commissioning and evaluation be explored. A number of policies, tools and practices were identified including: informed consent to share information with partners, person-centred care, branding and promotion, web-based and social media tools, client and family self-assessment, robust and trusted privacy, and skilled and well-supported "health home" general practices. The proposal also sought to explore more significant system changes including: cross-agency "task group" models of care; funding and performance agreement changes to ensure shared outcomes in contracts; joint service delivery teams with shared accountability structures; and joined-up entities or new shared-purpose organisations. Child and family Outcomes: The design identified the importance of having an outcome focus and included the monitoring of separate child and family outcomes. Included in this were several new initiatives that were considered central to integrated care, including: a shared outcome framework, patient reported outcome measures, patient reported experience measures, electronic medical record data-linkage projects, and population outcome monitoring. Evaluation: The designed initiative was a "complex intervention" and consequently the evaluation framework drew on Medical Research Council guidance for evaluation of complex public health interventions [20][21][22]. As noted above, the overall research design is a longitudinal, multilevel, critical realist evaluation of applied programme interventions that seek to break intergenerational cycle of social disadvantage and poor child health and development outcomes. MRC guidance argues that only through close scrutiny of causal mechanisms is it possible for evaluation to contribute to developing more effective interventions, and provide insights into how findings might be transferred across settings and populations. Consequently a critical realist mixed method approach was chosen to examine the quantity and quality (or process) of what was actually implemented in practice, the context, the mechanisms and seek to answer the question why [13]. Programme Theory We have previously published the Programme Theory informing this design [3]. It is presented here to assist readers' understanding of the design process. Theory of Change (ToC) The overall design retained the four strategic themes of the local collaborative design process (i.e. improving system capacity; health and wellbeing promotion; early intervention and supporting place-based initiatives). The place-based initiatives, system change, and health and wellbeing strategies were proposed to be implemented using a co-design approach with local communities and interagency partners. Consequently detailed ToC logic models were not developed for components related to those interagency strategic themes. By contrast the interagency partners sought to develop a ToC Logic Model that would guide the implementation of the early intervention and clinical aspects of the initiative. That ToC was particularly relevant for the 1) identification, 2) care coordination and 3) evidence-informed practice components of the design. It also sought to inform the clinical elements of the place-based projects. The ToC is shown at Figure 3. Prior to implementation in July 2015 a full Logic Model of the overall design was constructed to inform NSW Ministry of Health monitoring of the implementation. An adaption of that logic model is shown at Figure 4 with the inclusion of identified interventions and programme mechanisms. Discussion In designing an integrated care initiative for vulnerable families we have drawn on our previously reported empirical studies, the translation of causal theory to programme theory, collaborative design processes, and the integrated care policy elements advanced by the NSW Ministry of Health. Critical realism has provided the methodological underpinning of this programme of work and has assisted to explicate both the contextual conditions and the underlying causal and programme mechanisms. Consequently we have been able to move from our earlier theoretical models toward the design of whole of health and social care system interventions. In so doing we have moved from causal and programme mechanisms at the individual level toward mechanistic propositions relating to service systems and providers. For example, situated activity -face to face activity, and intermediate level social and service organisational mechanisms, continue to highlight the important of trust and willingness to share power. Thus the development and implementation of system change initiatives, such as general practice focused "health homes" and interagency "task groups" will be very reliant on approaches to sharing power and building trust between actors. To address those challenges the design has moved beyond the usual integrated care components of cohort identification, risk stratification and care-coordination to address the underlying programme mechanisms of trust and power sharing. Strong partnerships with general practice, and social and education sector partners, will be critical to the success of initiatives for children and families. True power sharing is difficult for health sector actors and will remain a challenge for the integrated care initiative described here. The design seeks to address not only intergenerational cycles of violence, family dysfunction and psychopathology, but also the social determinants of health as described in our earlier empirical and theoretical work. Thus throughout the design attention has been given to: meeting all the material needs of families, improving access to services, reducing family social isolation and marginalisation, building local social networks and community cohesion, and improving community health and life-skill literacy. Central to the initiative reported here is the strong interagency collaborative and the development of a shared outcomes framework. The final funding proposal included significant contributions from the local primary care agency, Family and Community Services NSW, non-government partners and academic institutions. A limitation was the lack of input from the Education sector which was attributable to restructuring of that sector at the time of design preparation. The consequence of this limitation is that the design has not fully explored the causal mechanisms operating within the school setting or the programme possibilities that might be operationalised through those settings. For example, the familyschool dimension was not examined and consequently the potential of school initiated family support strategies are absent from the design. A further limitation was the strong health sector focus despite the collective approach to planning. This can be attributed to the NSW Health integrated care tendering process, which was focused on initiatives for chronic health conditions. Despite this constraint, the final design was able to incorporate strong social care elements that took the design beyond the boundaries of the health sector. The underlying programme theory for the integrated care design remains tentative and will require further explication as part of the evaluation design. The design propositions developed followed the context-intervention-mechanism-outcome (CIMO) logic advanced by Denyer [14]. That approach was extended beyond the theoretical to be included in the overall logic model. Thus the final logic model included both implementation theory and programme theory elements as proposed by Blamey and Mackenzie [18]. The value of considering and analysing the underlying causal, implementation, and programme mechanisms has been that consideration was given to questions of "why" and "how". Consequently the design sought to build "mutually supporting" activities that would maximise the chances of success. Thus the analysis of how mechanisms operate in a context can help researchers to look out for, and establish potential areas of impact, through theorising about, and then establishing its causes. This can then facilitate the further refinement and improvement of programme design and implementation. Conclusion The design study described advances our earlier empirical and programme design studies toward the implementation of a whole-of-government integrated health and social care initiative. That initiative was designed as a cross-agency care coordination network that would ensure that vulnerable families: had their complex health and social needs met; kept themselves and their children safe; and were connected to society. In so doing we aim to break intergenerational cycles of poverty, violence and crime, poor education and employment opportunities, psychopathology, and poor lifestyle and health behaviours, through strengthening family resiliency, improving access to services, and addressing the social determinants of health and wellbeing. Neurobehavioral and Menstrual Disorders among Adolescent Females Environmentally Exposed to Pesticides, Menoufia Governorate, Egypt Background— Adolescent females living in agricultural areas where crops are routinely sprayed by pesticides are expected to be environmentally exposed to pesticides ˊ health hazards partially as those occupationally exposed. Objective— to assess menstrual and neurobehavioral disorders among adolescent females environmentally exposed to pesticides. Methods— This was a cross-sectional study conducted on 100 pesticide exposed adolescent females who had one or more of family members are pesticides ˊ seasonal applicators and 50 non-exposed adolescent females matched for age and education, served as controls at Menoufia governorate, Egypt during the period of pesticide application season of cotton crop from the first days of May to the end of September 2017. A self-administered and a series of neurobehavioral tests were administered and serum Acetylcholinesterase (AChE) activity was assessed. Results— A significant lower AChE activity levels were found in the exposed group than controls (Mean±SD=238.49± 23.83 vs 303.35±78.54 IU/L; respectively). There were significant higher mean scores of trail making test (parts 1 and 2) and significant lower mean scores of (similarities test, Benton visual retention test, block design test, Santa Ana dexterity test (dominant and non-dominant hands) and Beery visuo-motor imitation test in the exposed group than the controls (P<0.05). Also, the exposed group reported more prevalent irregular menstrual cycle (26.8%) and intermenstrual bleeding (28.2%) compared to the control participants (8.1% and 8.1%; respectively). Conclusion— ˊ applicators have significantly lower neurobehavioral performance, report more prevalent menstrual irregularities and have lower levels of serum AChE compared to a control group. The neurobehavioral deficits demonstrated a dose–response relationship AChE levels in the exposed participants. This necessitates the need for implementation of health education programs to prevent or reduce health effects associated with pesticide exposure to adolescent females. done using student t-test. Pearson correlation coefficient test was used to study the correlation between two quantitative variables. P values less than 0.05 were considered statistically significant. Introduction Pesticides are toxic chemicals that are widely used throughout the world in agriculture on crops as well as for domestic purposes for mosquito and cockroach control operations (Rani et al., 2017). Organophosphate pesticides (OPs) represent a class of pesticides with high toxicity and are used in both farmlands and households (Zhu et |
al., 2015). The pesticide exposure of farmers' families, especially children, can be potentially significant as there is a combination of para-occupational, environmental and domestic exposures (Silvério et al., 2017). Residential exposure depends on proximity of the house to areas treated with pesticides, the persistence of pesticides used in or around the home and domestic uses at home on pets (flies and ticks) and also on humans (lice and scabies) (Bouvier et al., 2005;Mamane et al., 2015). Direct exposure to pesticides can occur during the application whereas indirect exposure of farm spouses and children to pesticides can occur through spray drift, transfer of contaminated dust and soil from treated fields to farm vehicles and buildings, handling of contaminated clothing and personal contact with exposed individuals as they may track pesticides indoors on their clothes, shoes, skin and hair (El-Sebae, 1993; Lee et al., 2015). Also indirect exposures can occur through non-target species as air, water and soil and they represent routes of long-term generally low-level exposures. Pesticide use raises a number of environmental concerns. Over 98% of sprayed insecticides and 95% of herbicides reach a destination other than their target species (Miller, 2004). According to the Egyptian culture no females are recruited in pesticidesʹ application, hence their exposure to pesticides could be exclusively environmental either by living nearby agriculture fields, domestic use of home pesticides, or to pesticidesˊ residues spoiling clothes or skin and/or hair of their family members (brothers or fathers) working as pesticidesˊ applicators. Pesticide exposure can cause a variety of adverse health effects ranging from simple irritation of the skin and eyes to more severe effects such as endocrine disruption, blood and neurobehavioral disorders as well as the possibility of an increased risk of cancer especially, childhood cancers such as Non-Hodgkin lymphoma and leukemia (Clayton et al., 2003;Mamane et al., 2015). Organophosphate pesticides exposure can result in some neurobehavioral disorders such as memory loss, loss of coordination, reduced speed of response to stimuli, reduced visual ability, altered or uncontrollable mood; general behavior and reduced motor skills (Farahat et al., 2003;Silvério et al., 2017). Menstrual disorders like longer, irregular and missed menstrual cycles in addition to intermenstrual bleeding were detected among females exposed to carbamates and ∕ or OPs compared to women who never used pesticides (Farr et al. 2004). Pesticides are thought to pose a considerably higher risk to children than to adults as they are more vulnerable due to the significant anatomical and maturational physiological changes occurring in the brain during developmental periods including adolescence (Abdel-Rasoul et al., 2008). Most evidence for health effects of pesticides in adults comes from studies of occupationally exposed males. Relatively less is known about pesticide-related health effects in females, and there may be sex-specific risk differences with respect to reproductive toxicity. In addition, comparatively little is known about non-occupational pesticide exposure pathways (Rohlman et al., 2007). The aim of this study was to assess menstrual and neurobehavioral health disorders in the studied adolescent females that might arise due to environmental exposure of adolescent females to pesticides. Study design This was a cross-sectional study. Place and duration of the study The study was conducted at three randomly chosen districts (Shebin El-Kom, El-Bagour and Menouf) out of the ten districts in Menoufia governorate. The pesticide exposed female adolescents' homes lie within less than 1000 meters from the agricultural fields. Study sample The exposed group was chosen randomly from adolescent females ( aged from 9-18 years) whose one or more of their family members were pesticidesˊ seasonal applicators at cotton fields and / or working privately with their own backpack sprayers for other crops all over the year. Their homes lie within less than 1000 meters from the agricultural fields. The control group included 50 adolescent females that were matched with the exposed group regarding age, residence, socioeconomic standard, educational level and their families were never involved in field pesticidesˊ applications. Their homes lied by more than 1000 meters away from the agricultural fields. All the chosen female adolescents in both groups were single. Exclusion criteria included history of chronic medical disorders (diabetes, hypertension, asthma, thyroid disease, liver or kidney disease, peripheral neuropathy, vitamin deficiency) and illiteracy. Study methods each participant was subjected to the following: Personal data as age, residence, level of education, distance of house from fields and past medical history of diseases. b. Menstrual history as age of menarche, regularity of cycles, intermenstrual bleeding. c. Pesticide exposure history as applying pes ticides at home, use of empty pesticide container, handling contaminated clothes of their relatives, existing in fields within 3 days after pesticide spraying. d. Occupational history of their pesticide applicators' relatives: working days and duration of work. e. Neurological symptoms as headache, dizziness, blurred vision, impaired memory and concentration. II: Neurobehavioral test batteries: included: Age appropriate versions of the Wechsler Adult Intelligence Scale (WAIS) that were validated in an Arabic speaking population were used to assess neurobehavioral function for the studied participants. Better performance is evaluated by higher scores obtained on tests of Information, Similarities, Arithmetic, Digit Symbol, Block Design, Digit Span, and the Benton Visual Retention Test. In contrast, lower latencies or time to complete Trail Making A and B indicate better performance. Examiners were blind to the status of the participant as exposed or control (Abdel-Rasoul et al., 2008). III: Serum acetylecholine esterase (AChE) assessment Five milliliters of blood were drawn from all participants and serum AChE was determined according to Weber (1966) using standard kits (Test-combination Boehringer Mannheim GmbH Diagnostica). Serum AChE was selected because it is a better short-term indicator of cholinesterase inhibition than red blood corpuscles (RBCs) AChE due to its more rapid response to exposure; it is used as an indicator of recent, acute exposure to cholinesterase inhibiting pesticides. Also, because the primary pesticide being applied is chlorpyrifos, which has a preferentially inhibiting effect on serum AChE rather than RBC AChE (Abdel-Rasoul et al., 2008). Consent Written informed consents were signed by all participants before being enrolled into the study. Ethical approval Medical Ethics Committee at the Menoufia Faculty of Medicine approved the study protocol before starting. Data management Data were analyzed using IBM SPSS version 22 (SPSS Inc., Chicago, Illinois, USA). Chisquared test (χ 2 ) was used to examine the relation between qualitative variables. Z test is a significance test for testing proportions. For quantitative data, comparison between two groups was done using student t-test. Pearson correlation coefficient test was used to study the correlation between two quantitative variables. P values less than 0.05 were considered statistically significant. Results There was non-significant difference between exposed and control groups regarding sociodemographic data (age, income and education level), BMI, past clinical history (P>0.05) (data weren't illustrated in tables). A significant lower activity levels of AChE was found in the exposed compared to the control group (Mean± SD =238.49± 23.83 versus 303.35 ±78.54 IU/L; P < 0.001) as shown in figure (1). Table 1 shows that there was a significantly higher prevalence of dizziness, blurred vision, difficulty in concentration, impaired memory, feelings of depression, fatigue and involuntary movements in limbs among exposed group than controls. Table 2 shows that there were significantly higher mean scores of trail making test (parts 1 and 2) among the exposed group than the controls and significant lower mean scores of ( DISCUSSION This study revealed significantly higher prevalence of health effects in adolescent females environmentally exposed to pesticides compared to control group. Exposed participants had lower. AChE activity levels, impaired neurobehavioral performance, more prevalent neurological manifestations and menstrual irregularities. The significant association between the exposure to pesticides and decrement in the different functions of neurobehavioral performance (P<0.05, Table 2) as exposure to OP pesticides has been associated with irreversible inhibition of the enzyme cholinesterase thus there is a massive accumulation of the neurotransmitter acetylcholine within the synaptic cleft, different nerves and receptors in the body including neuromuscular (skeletal, smooth and cardiac) and glandular tissues because acetylcholinesterase is blocked leading to overstimulation of nicotinic and muscarinic receptors in the central and peripheral nervous system and muscle overstimulation. This decrement was also revealed by the presence of the significantly negative correlation between the number of days worked by pesticide applicator family members at the current season and years worked as pesticide applicators with Similarities, BVRT, Santa Ana dexterity test dominant hand and Beery visuo-motor imitation subtests and by significant positive correlation with Trail Making B. Also, the significant correlation between AChE activity level and the same tests (P<0.05, Table 4). These findings were consistent with other research results (e.g. Roldan-Tapia et al., 2005), that also reported a significant relationship between the period of exposure and working in applying OP pesticides and performance deficits and increased prevalence of neurological manifestations. This decrement of neurobehavioral performance and reporting more prevalent neurological manifestations in the exposed participants (P<0.05, Table 1) are consistent with a previous study that examined adult male pesticide workers in Egypt (Farahat et al., 2003) which found significant deficits in complex visual-motor processing and executive function, attention and short-term memory and memory/ perception. The exposure of the adults males was similar to the current study and some of the measurementss were the same across studies. Other studies in Egyptian populations comparing adult male pesticide applicators to controls have also reported increased prevalence of neurological, neuromuscular, and psychological symptoms and psychiatric disorders (Amr, 1997;Abdel-Rasoul et al., 2008). The current results also showed that there was a significantly lower activity level of AChE in the exposed adolescent females compared to the control ones (P<0.05, Figure 1). AChE is used as a biomarker for exposure to OP pesticides (Abdel-Rasoul et al., 2008). In spite of this recommendation, there are several studies that reported cholinesterase level showing a consistent significant association between exposure to pesticides and cholinesterase inhibition in farm workers. All studies found that AChE was significantly lower in the exposed participants than the controls (Amr, 1997;Farahat et al., 2003; Abdel-Rasoul et al., 2008). Exposed adolescent females who reported neurological manifestations had their relatives worked significantly more days than exposed participants who do not report these manifestations. These results are similar to studies with adults which found that increased time spent working is associated with increased symptom reporting (Farahat et al., 2003;Rohlman et al., 2007). The association between the level of AChE activity and neurobehavioral performance, was confirmed by presence of significant correlations between the AChE activity level and the performance in 6 out of 8 neurobeahvioral tests (Trail Making, Similarities, Santa Ana dexterity test dominant and non-dominant hand and Beery visuo-motor imitation subtests). These findings address the concerns raised by Roldan-Tapia et al. (2005), about the relationship between neurobehavioral performance and inhibition of AChE. Exposed adolescent females showed a significantly higher prevalence of menstrual disturbances than controls from the same communities (P<0.05, Table 3). These findings confirm previous reports indicating that menstrual disorders like longer, irregular and missed menstrual cycles in addition to intermenstrual bleeding were detected among adult females exposed to carbamates and ∕ or OPs compared to adult females who never used pesticides (Farr et al., 2004). Exposure to pesticides has been associated with menstrual cycle disturbances most probably due to potential effects of endocrine disrupting pesticides on the female reproductive system as modulation of hormone concentrations, ovarian cycle irregularities and impaired fertility. Limitations of the study The main limitation of our study is the scanty number of studies and cohorts.on the adolescent females environmentally exposed to pesticides Furthermore, not all neurobehavioral tests were used in all studies and cohorts. Greater precision would be achieved by the inclusion of additional studies. The second important limitation is the inconsistencies in exposure classification among the cohorts, ranging from purely categorical classification (agricultural workers/applicators vs controls) to more quantitative indices of exposure level including measurement of exposure across time. Conclusions and Recommendations Pesticides have been linked to numerous adverse health effects that are different in females than in males. Adolescent females living in agricultural areas and from families whose one or more members are pesticidesˊ sprayers have significantly lower neurobehavioral performance, report more prevalent manifestations of menstrual disturbances and have lower activity levels of serum AChE compared to a control group. The neurobehavioral deficits demonstrated a dose-response relationship between days and years of exposure and neurobehavioral performance and prevalence of manifestations reported. Since adolescents around the world are exposed to OP pesticides, |
these studies suggest an urgent need to evaluate this potential problem. Future research should focus on the type of health educational training considering the perceived benefits and disadvantages while developing plans to decrease health disorders among adolescent females. Mean levels of serum AChE among exposed and control groups Prevalence of neurological manifestations in exposed and control groups Mean ± SD of neurobehavioral performance for the exposed and control groups Menstrual history of the exposed and control groups Table 3 shows that the exposed group had a significantly higher prevalence of irregular menstruation and intermenstrual bleeding compared to the control group. Correlations between neurobehavioral tests' performance with AChE activity levels, duration of work of pesticide applicator family member among the exposed group Short Fiber Reinforced Composite: a New Alternative for Direct Onlay Restorations Objectives: To determine the static load-bearing capacity of direct composite onlay restorations made of novel filling composite resin system which combines short fiber-reinforced composite resin (FC) and conventional particulate filler composite resin (PFC). Methods: Three groups of onlay restorations were fabricated (n = 8/group); Group A: made from conventional particulate filler composite resin (Z250, 3M-ESPE, USA, control), Group B: made from short fiber-reinforced composite resin (EverX posterior, StickTeck Ltd, member of GC group, Turku, Finland) as substructure with 1 mm surface layer of PFC, Group C: made from FC composite resin. The specimens were incrementally polymerized with a hand-light curing unit for 80 s before they were statically loaded with two different sizes (3 & 6 mm) of steel ball until fracture. Failure modes were visually examined. Data were analyzed using ANOVA (p = 0.05). Results: ANOVA revealed that onlay restorations made from FC composite resin had statistically significantly higher load-bearing capacity (1733 N) ( p < 0.05) than the control PFC composite resin (1081 N). Onlays made of FC composite resin with a surface layer of PFC gave force values of 1405 N which was statistically higher than control group ( p < 0.05). No statistically significant difference was found in the load-bearing capacity between groups loaded by different ball sizes Significance: Onlay restorations combining base of short fiber reinforced composite resin as substructure and surface layer of conventional composite resin displayed promising performance in high load bearing areas. INTRODUCTION A variety of techniques are currently available to restore teeth with moderate coronal defects in the posterior region and the selection of the appropriate modality is dependent on the evaluation of and compliance with numerous criteria. Routine use of metal ceramic crowns instead of gold alloy partial crowns and onlay restorations enforces the removal of healthy enamel and dentin. Adhesively cemented ceramic onlays have been used as an alternative in order to minimize the removal of tooth structure. The greatest success using ceramic onlays was previously limited to anterior teeth with porcelain veneers [1][2][3]. This was not surprising, as their fracture resistance and abrasiveness were clearly inferior to that of gold alloys. However, after extensive material improvements, ceramic restorations are accepted for restoring posterior teeth with extended lesion today [4]. Despite their *Address correspondence to this author at the Department of Restorative Dentistry & Periodontology, Institute of Dentistry, Libyan International Medical University, Benghazi, Libya; Tel: 00218 3967435; Fax: +358 2 333 8390; E-mail: [email protected] less than esthetic appearance, the physical properties of gold alloys have created a standard that has been difficult to match; for example, toughness and a high compressive loadbearing capacity [1,5]. Particulate filler composite resin (PFC), at one time considered only as a treatment option for anterior teeth, has steadily been found to have wider applications. With the improvements in the mechanical properties of PFCs, their use has been widened not only to the posterior intra-coronal area, but also to extra-coronal restorations, and even complete crowns and fixed partial dentures [6]. Many studies have been undertaken to investigate the filler phases, resin compositions, and curing conditions to improve the mechanical properties of PFC [1,7,8]. However, further significant improvements are needed in order to extend the use of PFC to high stress-bearing applications such as direct posterior restorations involving cusps and indirect restoration, inlays and onlays [1,8]. In terms of indirect restorations, inlays/onlays have been used for almost 25 years. They were introduced in the hope of overcoming problems associated with the lower degree of conversion related to direct posterior PFCs being placed using conventional incremental techniques. The most common problems that occurred were vari-various types of fractures in high stress-bearing areas [1,9]. It was hoped that the use of the indirect technique would improve the load-bearing capacity of the composite by raising the degree of conversion obtained by laboratory postcuring of the restoration. It is reported that extra-oral polymerizations of the composite followed by cementation appear to improve the marginal fit and minimize contraction stress [1,10]. The mechanical properties of the composites were also improved by post-cure heat treatment, although such improvements were modest and sometimes not statistically significant [1,11,12]. The relatively high brittleness and low load-bearing capacity of current PFCs still hinder their use in large stress-bearing restorations [1,[13][14][15]. It therefore follows that there is a considerable need for improved mechanical properties, especially load-bearing capacity and wear resistance, whilst still retaining esthetic properties. Recently, short fiber reinforced composite FC resin was introduced as a dental restorative composite resin [16][17][18][19]. The composite resin is intended to be used in high stress bearing areas especially in molars. The results of the laboratory mechanical tests revealed substantial improvements in the load bearing capacity, the flexural strength and fracture toughness of dental composite resin reinforced with short Eglass fiber fillers in comparison with conventional particulate filler composite resin [16][17][18][19][20]. The short fiber composite resin has also revealed control of the polymerization shrinkage stress by fiber orientation and, thus, marginal micro leakage was reduced compared with conventional particulate filler composite resins [18]. It can be hypothesized that by using a FC composite substructure under PFC, the static load-bearing capacity of the material combination could be improved. Load application over the restoration is one of the factors that could influence the load bearing capacity. Thus, the aim of this study was to determine the static load-bearing capacity with two different loading stresses of direct composite onlay restorations made of novel filling composite resin system which combines short fiber-reinforced composite resin and conventional particulate filler composite resin. MATERIALS AND METHODS Onlay preparation with three cusps coverage of upper 1 st molar was prepared on zirconium model (Fig. 1). Preparation was made with 2-3 mm of axial and occlusal reduction. A transparent template matrix of an ideally contoured upper 1 st molar crown was used to aid standardized onlay restorations construction. Onlay restorations were fabricated according to the groups they belonged (Fig. 1 The onlay restorations of each group (n = 8) were polymerized with hand-light curing unit (Optilux _501, Kerr, CT, USA) for 80 s from all directions (wavelength: 380 and 520 nm with maximal intensity at 470 nm, light irradiance 800 mW/cm2). Subsequently after polymerization, the zirconium model with the onlay restoration was fixed to the metal base of testing device before statically loaded (spherical Ø 3 & 6 mm) (Fig. 1). Luting cement was not used and the restorations were tightly fixed solely by resin. The static compressive fracture test was performed using a universal testing machine (model LRX, Lloyd Instruments Ltd., Fareham, UK) at a speed of 1 mm/min, and data were recorded using PC software (Nexygen Lloyd Instruments Ltd.). The loading event was registered until restoration fracture. Failure patterns of each loaded restorations were visually analyzed. Data of the fracture load values were statistically analyzed with SPSS version 13 (SPSS Inc., Chicago, IL, USA) using analysis of variance (ANOVA) followed by the A B C Tukey's post hoc test at a significance level of 0.05 to determine the differences between the groups. RESULTS The mean load-bearing capacities of the onlay restorations with standard deviations (SD) are given in Fig. (2). ANOVA revealed that onlay restorations made from short fiber FC composite resin had significantly higher loadbearing capacity (1733 N) (p < 0.05) than those made only from conventional particulate filler composite PFC (control) (1081 N). Onlays made of FC composite resin with a surface layer of PFC gave force values of 1405 N which was statistically higher than control group (p < 0.05). No statistically significant difference was found in the load-bearing capacity between groups loaded by different ball sizes. Except with Group A (made from only PFC) had tendency to fracture more easily with smaller loading ball. Visual inspection revealed two types of fracture patterns according to the material used: catastrophic splitting of loaded cusps in restorations made of PFC composite resin, and chipping restoration fracture was found in all restorations made from either plain or substructure FC composite resin. DISCUSSION In our laboratory study we examined the fracture resistance of composite specimens simulating posterior an onlay restorations. After years of follow-up of indirectly or directly made posterior composite restorations, the clinical studies show that fracture of the restoration is the most common reason for failure, with no significant differences between the two techniques [1,21,22]. On the other hand, promising laboratory results were documented with the use of short fiber FC composite resin in high stress-bearing applications [16][17][18][19][20]. It was hypothesized that a FC composite resin substructure could reinforce the composite in onlay restoration for use in high stress-bearing areas of the dental arch. The data showed substantial improvements in load-bearing capacity when materials combination were used (Fig. 2). The function of bulk short fiber composite substructure is based on supporting the surface particulate filler composite layer and working as crack stopper layer. Reinforcing effect of the fiber fillers is based on stress transfer from polymer matrix to fibers but also behavior of individual fiber as a crack stopper. Random fiber orientation had a significant role in mechanical properties. Clinical study reported by Van Dijken have shown that restorative composite with microfibers suffer extensive wear [23], which can be partly explained because of the used fiber length was well below of critical fiber length. In order a fiber to act as an effective reinforcement for polymers, stress transfer from the polymer matrix to the fibers is essential [24,25]. This is achieved, if the fibers have a length equal or greater than the critical fiber length [24]. It has been measured using fiber fragmentation test that the critical fiber lengths of E-glass with bis-GMA polymer matrix vary between 0.5 and 1.6 mm [26]. Deteriorated or initially poor adhesion between the fibers and polymer matrix increase the critical fiber length. In this case, the mechanical friction of fibers to polymer matrix at the interface can compensate the poor adhesion [27]. Based on this, the short fiber composite resins used in this study have fiber fillers equal or greater to critical fiber length. To receive support from the short fiber composite substructure for the surface particulate composite, the structural rigidity of the short fiber composite substructure should be higher than that of surface particulate composite resin. In this, the fiber orientation likely has a significant role. On the other hand, if the function of the short fiber composite substructure is based on the mechanism of a crack stopper, the distance from the surface of the stress initiation point to the fibers is of importance. Therefore, the volume fraction or thickness of short fiber composite could contribute to the crack propagation and load-bearing capacity. Previous study by authors, showed when short random fiber-reinforced composite (FRC) was used as substructure for particulate filler composite, the load-bearing capacity of the materials combination increased linearly as thickness layer of FRC increased [1,28]. From this point of view, in this experimental study, short fiber composite substructure was covered with only 1 mm layer thickness of hybrid particulate filler composite resin. The alteration of ball size changes the position of load applications, modifying the concentration of tension patterns [29]. Habekost et al. showed that force required to cause fracture with 10 mm diameter ball was greater than with the 3 mm diameter ball [29]. However, contrary to expectation, in this study no statistically significant difference was found in the load-bearing capacity between groups loaded by 6 mm or 3 mm metal ball. This might be because of the small diameter's difference between balls was not mechanically influence. Stress applied to the teeth |
and dental restorations is generally low and repetitive rather than being isolated and impactive in nature. However, because of a linear relationship between fatigue and static loading, the compressive static test also gives valuable information concerning load-bearing capacity [30]. The fracture resistance values determined by the various investigators were recorded under different measurement criteria. These criteria were either initial cracking that was interpreted as crack development or a reduction in the load by an absolute or relative amount [1,31,32]. For this study, the maximum force on the final fracture was determined. Fracture patterns were analyzed visually and two types of fracture patterns were found, where each fracture type occurred according to the type of material. Because of the brittleness of composite resin, the catastrophic splitting of cusps was found in all specimens made from PFC only. In contrast, short fiber composite allowed the crack to propagate through the surface PFC and FC composite resin to make a compound-like fracture with no delamination found. Methodologically, limitations like sample size and aging process, such as alternate thermal stress, mechanical stress, wear and water storage, should be taken into consideration. Despite the importance of laboratory studies to answer some questions in a short time, the real performance of restorations can only be determined by long-term clinical trials. CONCLUSIONS Within limitations of this experimental study, the results showed good performance of a novel materials combination of bulk short fiber composite substructure and surface layer of particulate filler composite in high stress bearing areas. Nanomiemgel - A Novel Drug Delivery System for Topical Application - In Vitro and In Vivo Evaluation Aim The objective of this study was to formulate and evaluate a unique matrix mixture (nanomiemgel) of nanomicelle and nanoemulsion containing aceclofenac and capsaicin using in vitro and in vivo analyses and to compare it to a marketed formulation (Aceproxyvon). Methods Nanomicelles were prepared using Vitamin E TPGS by solvent evaporation method and nanoemulsion was prepared by high-pressure homogenization method. In vitro drug release and human skin permeation studies were performed and analyzed using HPLC. The efficiency of nanomiemgel as a delivery system was investigated using an imiquimod-induced psoriatic like plaque model developed in C57BL/6 mice. Results Atomic Force Microscopy images of the samples exhibited a globular morphology with an average diameter of 200, 250 and 220 nm for NMI, NEM and NMG, respectively. Nanomiemgel demonstrated a controlled release drug pattern and induced 2.02 and 1.97-fold more permeation of aceclofenac and capsaicin, respectively than Aceproxyvon through dermatomed human skin. Nanomiemgel also showed 2.94 and 2.09-fold greater Cmax of aceclofenac and capsaicin, respectively than Aceproxyvon in skin microdialysis study in rats. The PASI score, ear thickness and spleen weight of the imiquimod-induced psoriatic-like plaque model were significantly (p<0.05) reduced in NMG treated mice compared to free drug, NEM, NMI & Aceproxyvon. Conclusion Using a new combination of two different drug delivery systems (NEM+NMI), the absorption of the combined system (NMG) was found to be better than either of the individual drug delivery systems due to the utilization of the maximum possible paths of absorption available for that particular drug. Introduction Current therapies for the treatment of skin inflammation are not wholly effective and could be attributable to the types of topical delivery systems used. There is therefore a need to develop a controlled-release drug delivery system that would effectively deliver anti-inflammatory agents to reduce pain, inflammation, disease progression and prevent adverse reactions [1,2]. Aceclofenac is a Non-Steroidal Anti-Inflammatory Drug used in the treatment of inflammation and degenerative disorders of the musculoskeletal system. It is widely prescribed for the treatment of osteoarthritis and rheumatoid arthritis [3,4,5]. Capsaicin on the other hand, is used alone or in combination with other anti-inflammatory drugs to effectively reduce itching associated with skin inflammatory conditions [6,7,8,9]. The skin is an exceptionally effective barrier and it prevents the permeation of most of the drugs applied for therapeutic purposes [10,11]. Very few drugs have the capability to permeate in significant amounts through the skin. Most of the topical dosage forms available on the current market have poor penetration, which leads to poor therapeutic benefit [12,13]. Hence a delivery system that makes the skin more permeable and penetrates the skin by multiple mechanisms to enhance topical drug delivery is of great formulation interest. Long-term oral administration of aceclofenac causes serious gastrointestinal side effects like GI bleeding and ulceration [14,15]. Therefore, an improved topical aceclofenac formulation with a high degree of percutaneous permeation could be a useful alternative for the treatment of locally inflamed skin. One of the most promising drug delivery systems for enhancing skin permeation of drugs is the microemulsion or nanoemulsion system [16,17]. Nanoemulsions are thermodynamically stable transparent (translucent) dispersions of oil in water stabilized by an interfacial film of surfactant and co-surfactant molecules with droplet size less than 1000 nm. Kakumanu et al. [18] have illustrated in the epidermoid skin carcinoma xenograft mouse model that the nanoemulsion formulation of dacarbazine dramatically increases its efficacy. In this study, we have developed a drug delivery system called nanomiemgel through a novel formulation strategy, which utilizes the ''Multi Absorption Mechanism'' (MAM) concept and has a broad applicability. Nanomiemgel consists of two types of matrices; A & B. Matrix A comprises the nanoemulsion whilst matrix B comprises the nanomicelles. The hypothesis of the present study is that every nano drug delivery system is unique and its rate, extent and mechanism of absorption depend on the size, charge and composition of the nano drug delivery system. So, when a combination of completely different drug delivery systems is utilized for the delivery of a drug, the absorption of the combined system would be better than either of the individual drug delivery systems due to the utilization of the maximum possible paths of absorption available for that particular drug (Fig. 1). The objective of the present research work was to investigate the beneficial effect of the combination of the nanomicelles (NMI) and nanoemulsion (NME) as a novel carrier system for the topical application of aceclofenac (ACE) and capsaicin (CAP). In this study, the in vitro permeation of the drug through dermatomed human skin and inflamed mice skin, bioequivalence determination of the nanomiemgel (NMG) with marketed formulation Aceproxyvon (MKT) using dermal microdialysis and the in vivo anti-inflammatory activity of the NMG (nanomicelles plus nanoemulsion plus gel) were evaluated. The plan of work of this research work was to prepare NEM and NMI individually, and incorporation of NEM and NMI at 1:1 ratio into the gel to get a uniform matrix of the combined drug delivery system (NMG) and then characterization of NEM, NMI and NMG by different means followed by a systematic investigation of these formulation for in vitro and in vivo efficacy in psoriatic like inflammation model and also to compare the effects with MKT using dermal microdialysis. Formulation of aceclofenac and capsaicin nanoemulsion The aceclofenac (ACE) and capsaicin (CAP) nanoemulsion (NEM) was prepared by first dissolving 750 mg of ACE and 5 mg of CAP in 8 mL of olive oil and miglyol (1:1), followed by the addition of 6 mL of ''Polysorbate 80 and Transcutol'' mixture (1:1). The oil and surfactant mixture-containing drug were sonicated for about 15 minutes to get clear oil & surfactant mixture. To the mixture, 11 mL of deionized water was added while homogenizing to get a primary emulsion. The obtained o/w emulsion was homogenized further for 5 minutes at 3000 rpm to obtain a microemulsion. This microemulsion when passed through a NanoDEBEE (Bee International, South Easton, MA, USA) at 24,000 psi for 5 cycles resulted in a NEM. For the measurement of the mean droplet size and polydispersity index (particle size distribution) and f-potential, a zeta sizer (Nicomp 380 ZLS (Particle Sizing Systems, Port Richey, FL) was used. Formulation of aceclofenac and capsaicin nanomicelle The ACE and CAP nanomicelle was prepared with Vitamin E TPGS using the solvent evaporation method where the organic solvent was removed through evaporation. The Vitamin E TPGS (7.55 gm), aceclofenac (750 mg) and capsaicin (5 mg) were added to 2 mL of acetone. When a clear solution was obtained, 25 mL of distilled water was added. Then the organic solvent was removed gradually through evaporation. Change of solvent quality and hence, selectivity, from organic to aqueous was gradual; the polymer and the drugs were able to aggregate into micelles rather than precipitating from the solution into the bulk. The solvent of choice was acetone, due to its high water miscibility and low vapor pressure, which simplified the solvent removal. continuous low rpm stirring to form a uniform gel that was free from lumps and bubbles. The pH of gel was then neutralized with triethanolamine (TEA). After cooling the gel phase to 40uC, the NEM and NMI formulations were incorporated into the carbopol gel and mixed uniformly to obtain the NMG. The NEM and NMI were dispersed into the carbopol gel to achieve the final concentration of ACE and CAP at 1.5% & 0.01% respectively, whereas the final concentration of carbopol was maintained at 2%. This NMG was used for the drug release, skin permeation and other in vivo studies. Texture analysis Different gel formulations were placed one after other on a cone shaped plate and subjected to texture analysis with the TA.XT Plus texture analyzer (Stable Micro Systems, Surrey, UK) interfaced with a computer. The principle of operation is that when a 10 g surface trigger is attained (i.e. the point at which the disc's lower surface is in full contact with the product), the disc proceeds to penetrate to a depth of 30 mm. At this point (most likely to be the maximum force), the probe returns to its original position. The 'peak' or maximum force is considered as a measurement of firmness i.e. the higher the value, the firmer the sample is. The area of the curve up to this point is taken as a measurement of consistency i.e. the more the value, the thicker the consistency of the sample. The negative region of the graph, produced on probe return, is as a result of the weight of the sample which is lifted primarily on the upper surface of the disc on return(due to back extrusion) and hence gives again an indication of consistency/ resistance to flow off the disc. The maximum negative force is taken as an indication of the cohesiveness of the sample i.e. the more negative the value, the more 'cohesive' the sample. The area of the negative region of the curve may be referred to as the 'work of cohesion' i.e. the higher the value, the more resistant the sample is to withdrawal which is an indication of the cohesiveness and also the consistency/viscosity (Fig. 2). All the measurements were made on gels equilibrated to ambient temperature. The gels were compressed twice at 0.5 mm/s speed until 10-15 percent of their original amount was left. The results were reported as the means of triplicate tests. Atomic Force Microscopy Topographical and phase images were obtained for all the nanoformulations prepared. In order to determine the size of the NEM, NMI and NMG and to obtain high resolution images, Atomic Force Microscopy (AFM) analyses were carried out using a Nanosurf Flexa microscope, Switzerland, in the non-contact dynamic mode. The NEM and NMI suspensions were first diluted into ultra-pure water (18 MV.cm 21 ) at 1% v/v, whereas the NMG suspension was diluted into ultrapure water at 0.1% v/v. The diluted suspensions were then mixed on a vortex for 1 minute. Subsequently, 5 mL of the diluted nanoformulations were poured on optical microscope slides (5 mm 65 mm) previously cleaned upon sonication in ethanol. The slides containing the drop casting samples were dried at room temperature (22uC) in a desiccator (silica gel) for 3 days. All measurements were performed at room temperature (22uC) under controlled humidity (20-30%). The Gwyddion software was employed for images treatment for quality enhancement. HPLC analysis HPLC system (Waters Corp, Milford, MA) with a symmetrical reverse phase C18 analytical column (5 mm, 4.66250 mm) was used for the analysis of ACE and CAP present in the NMG. The mobile phase used was 10 mM acetate buffer (pH adjusted to 3) and acetonitrile at 45:55 v/v with a flow rate of 1 mL/min. Retention time for ACE and CAP were 6 and 3.5 min respectively (Fig. 3). All |
the samples were analyzed at 275 nm. In-vitro drug release In vitro drug release study was performed using USP type II apparatus in 500 ml of phosphate buffer pH 6.8, with speed and temperature maintained at 100 rpm and 37¡0.5uC, respectively. 2 gm of each formulation (NMG, NMI, NEM and MKT) and free drug gel (FD) (equivalent to 30 mg ACE and 0.2 mg CAP) was placed in a dialysis bag made up of cellulose (molecular weight cut off of 12,000 g/ mole; Sigma, USA) and subjected to the dissolution study. 1 mL samples were withdrawn at regular time intervals (1,2,4,6,8,12,24,48 and 72 h) and same amount of dissolution medium was replaced after each withdrawal. The withdrawn samples were analyzed for the drug content by using developed RP-HPLC method at 275 nm. Stability studies Different formulations (NMG, NMI and NEM) including MKT were filled in the glass vials and subjected to the stability studies as per the ICH guidelines at 2-8uC (refrigeration condition), 25¡2uC/60¡5%RH (room temp) and 40¡2uC/ 75¡5%RH (accelerated condition) for 6 months. Appearance and clarity were analyzed by visual inspection. The particle size of NMG and drug content were evaluated using zeta sizer and HPLC respectively. Ex vivo permeation through dermatomed human skin Dermatomed human skin was obtained from Allosource (Centennial, CO) with a thickness of 0.5-0.1 mm. The skin was then stored at 280uC until used. The dermatomed human skin was thawed and washed with distilled water for 30 min to remove the excess of glycerol. Skin permeation studies were performed using established procedures. The human skin permeation studies were performed by mounting the dermatomed human skin on Franz diffusion cells (Permegear Inc., Riegelsville, PA). The surface area of the dermatomed human skin exposed to the formulation in the donor chamber was 0.64 cm 2 and the receiver fluid volume was 5 ml. Free drug (FD), Aceproxyvon (MKT), NEM, NMI and NMG were applied evenly on the surface of the human skin in the donor compartment. The skin permeation study was performed using six diffusion cells and represented as an average of six cells. The receiver compartment was filled with 1% w/v Tween 80 in PBS (pH 7.4) and stirred at 300 rpm. The temperature of the receiver compartment was maintained at 37¡0.5uC using a circulating water bath to simulate the skin temperature at physiological level. To replicate the clinical conditions, a non-occlusive condition was followed and the surface of the skin was exposed to the surrounding air. After 24 h of skin permeation, the receiver fluid was collected and centrifuged at 15000 rpm for 15 min and analyzed for drug content using the HPLC method. Skin extraction For the evaluation of drug retention in dermatomed human skin, the entire dosing area (0.64 cm2) was collected with a biopsy punch. Stratum corneum (sc), epidermis (ed) and dermis (de) were separated using cryotome. SC, epidermis and dermis were minced and boiled with 250 mL PBS (pH 7.4) separately for 10 min. To these samples 250 mL of acetonitrile was added to solubilize the drug. All the samples were then centrifuged at 15000 rpm for 15 min. The supernatant was collected and analyzed by HPLC for drug content. Visualization of skin penetration Microscopic analysis on the rat skin was carried out following skin permeation study with FITC (flourescein isothiocyanate) dye loaded NEM, NMI and NMG to demonstrate the superior skin permeation ability of NMG over NEM and NMI using confocal laser scanning microscopy (Leica Microsystems Inc., IL, USA). Cryosectioning was done to get both cross and lateral sections up to 400 mm depth using cryotome (Shandon, UK). The cryotomed skin sections of NEM, NMI and NMG were visualized and analyzed for skin associated fluorescence using a 106 objective piece throughout the study for all the samples. In Vitro Evaluation of Probe Recovery The transfer rate of the probes, possible binding effect and in vitro recovery of ACE and CAP were assessed as per previously described procedures [19]. The standard stock solution of ACE (10 mg/ml) and CAP (0.01 mg/ml) was prepared in 1% (w/v) Tween-80 containing PBS solution. A linear microdialysis (MD) probe with 10 mm dialysis membrane (LM-10, Bioanalytical Systems, West Lafeyetter, IN, USA) was used for these studies. In Vivo Evaluation of Probe Recovery In vivo recovery of ACE and CAP was done by retrodialysis methods. CD SD hrBi hairless rats were anaesthetized with an intraperitoneal injection of urethane (1.5 g/kg) and placed on a temperature controlled heating pad (37¡2uC). To do the dermal implantation of the LM-10 probe, the skin of the dorsal region of the rat was punctured with 19-gauge intravenous needle (BD Company, Franklin Lakes, NJ, USA) and MD probe was inserted through the guiding needle cannula. Then the needle was retracted leaving the dialysis membrane in the skin horizontally. ACE and CAP standard stock with 1% (w/v) Tween-80 in Krebs-Ringer solution was passed through the probe using an infusion pump at a flow rate of 2 ml/mL and the dialysate samples were collected every 60 min for 480 min. The recovery was determined from the ratio of the concentration loss to the initial concentration in the perfusate. In vivo Skin Microdialysis In vivo Skin microdialysis was done using the MD setup with slight modification of standard procedure [20]. In short, the linear MD probes were used and were continuously perfused with 1% (w/v) Tween-80 in Phosphate Buffer Saline (PBS) solution at a flow rate of 2 ml/min. Dialysate samples were collected every hour into micro-fraction collector throughout the study period. Two MD probes were implanted in parallel position on the dorsal surface of the rat skin. The needle was inserted into the skin very carefully so that the needle could be visible through the superficial skin layer. For the topical application of different formulation gels, the donor chambers were fixed on the rat skin at the point where MD membranes were implanted using glue. 200 ml of FD, MKT, NEM, NMI and NMG gels (n54) were applied on the skin surface using donor chambers. Subsequently, the dialysate samples were collected every hour for 24 h and were analyzed for the ACE & CAP levels using HPLC. The concentrations of ACE and CAP at different time points in the dialysate samples were subjected to pharmacokinetic analysis to estimate various pharmacokinetic parameters like maximum plasma concentration (C max ), time to reach maximum concentration (t max ), area under the plasma concentration-time curve (AUC 0Rt and AUC 0Rv ) and elimination half life (t 1/2 ) using WinNonlin software. Imiquimod (IMQ) induced psoriatic plaque like model The model selected was IMQ-induced psoriatic-like plaque in mouse skin because it closely resembles the clinical presentation of human psoriasis with respect to the erythema, skin thickening, scaling, epidermal alterations, neoangiogenesis and infiltration of inflammatory proteins such as T cells, neutrophils and dendritic cells [21,22]. C57BL/6 mice of age 8-11 weeks were kept under specific pathogen-free conditions. Topical application of IMQ suspension was carried out for 5 consecutive days on the shaved backs of the mice. The dose of IMQ applied (5 mg/day) was optimized based on the induction of skin inflammation. Subsequently, the inflamed skin area was treated with FD, MKT, NEM, NMI and NMG gels topically for 5 consecutive days. Groups subjected to no IMQ and IMQ only treatments were utilized as the negative (NC) and positive (PC) control groups, respectively. Scoring severity of skin inflammation, ear thickness and spleen enlargement An objective scoring system was developed based on the clinical Psoriasis Area and Severity Index (PASI) to score the severity of inflammation induced on the back of the mice. The extent of erythema, scaling and thickening was scored independently on a scale from 0 to 4: 0, none; 1, slight; 2, moderate; 3, marked; 4, very marked. The scoring was performed every 24 h for 5 days. The reduction in the thickness of the inflamed ears of the mice was measured using caliper (Marathon Caliper, Digital Electronic, VWR, USA) and this parameter was also used to assess the treatment success of the inflammation induced after the topical application of different gel formulations. After 5 days of treatment with the different formulations, the mice were sacrificed and the spleen weights were taken to determine the differences among the different groups and to evaluate the efficiency of different formulations. Histology The inflamed skin of mice was collected at the end of the experiment and stored in 10% neutral phosphate buffered formalin. Following fixation, the samples were dehydrated and embedded in paraffin. Microtome sections of about 5 mm thickness were taken from the inflamed skin and stained with hematoxylin as well as eosin. The Olympus BX40 light microscope equipped with computercontrolled digital camera (DP71, Olympus Center Valley, PA) was used to visualize the images on the slides. Immunohistochemistry (IHC) IL-23 is a cytokine produced by macrophages that takes an important role in the inflammatory response [23,24]. Hence, IHC study of IL-23 was performed as per the procedure described by Chougule et al. [25]. In brief, formalin-fixed, paraffinembedded skin sections were used for IHC studies according to the protocol specified in the ImmunoCruz mouse ABC staining kit (SantaCruz Biotechnology Inc, CA). The section slides were washed in xylene and hydrated with different concentrations of alcohol. The slides were incubated with the primary antibody against IL-23 overnight at 4uC. Horseradish peroxidase-conjugated secondary antibody was applied to locate the primary antibody. The specimens were stained with DAB chromogen and counterstained with hematoxylin. The presence of brown staining was considered as a positive identification for activated IL-23. The Olympus BX40 light microscope equipped with computer-controlled digital camera (DP71, Olympus Center Valley, PA) was used to visualize the images on the slides. Permeation through inflamed mice skin The psoriatic dorsal skin portions of the mice following treatment with the different formulations were subsequently collected and subjected to drug permeation studies using a similar experimental set up as for the permeation studies with dermatomed human skin. The different mounted skin portions were all then treated with NMG. The concentrations of ACE and CAP were maintained as same as used for the permeation studies through the dermatomed human skin. The permeation study was conducted to ascertain the effect of inflammation (plaque, scaling and epidermal thickness) on the skin permeation of drugs to determine the efficiency of the topical delivery of these formulations during psoriasis treatment. Combination Index The combination index (CI) value was calculated separately for different parameters like the cumulative percent of drug permeated into the receptor compartment, amount of drug retained in the different layers of the skin, the thickness of the inflamed ear of the mice treated with IMQ to induce psoriasis and the PASI score to evaluate the combined effect of NEM and NMI. The CI was calculated using following equation [26]: The CI values were interpreted as follows: CI.1.3: antagonism, CI51.1 to 1.3: moderate antagonism, CI50.9 to 1.1: additive effect, CI50.8 to 0.9: slight synergism, CI50.6 to 0.8 moderate synergism, CI50.4 to 0.6: synergism, CI50.2 to 0.4: strong synergism. Statistical analysis The ACE and CAP content of the skin tissue was expressed as mg per g of the tissue. The differences in the skin permeation among the formulations FD, MKT, NEM, NMI and NMG gels were examined using ANOVA. The means were compared between two groups using the student's t test. Mean differences with p,0.001 were considered to be significant. All data is expressed as the mean ¡ S.D., n53. Physical characterization of NMG The NEM and NMI, each containing ACE and CAP had mean particle size of 229¡16 nm and 185¡10 nm, respectively and polydispersity index (PI) of 0.18¡0.06 and 0.12¡0.08 respectively. The entrapment efficiency of ACE and CAP in NEM was 96.74¡4.24% and 95.72¡3.58%, respectively whereas for NMI, was 94.88¡3.76% and 92.69¡3.08%, respectively. Texture Analysis The firmness, consistency, cohesiveness and viscosity were determined to optimize the texture parameters of the different formulations. The different texture parameters of the nano formulations, FD and MKT gels were given in Table 1 In vitro drug release The release of ACE and CAP from NMG was determined to be an intermediate between NMI and NEM. The pattern of drug release of ACE and CAP was almost similar for NEM, NMI and NMG and followed first order release kinetics with a best fit R 2 value of less than 0.99. No sustained |
release of drug was determined for FD whilst there was sustained drug release for only 10 hr for MKT. However, for NEM, NMI and NMG, the drug was released in a sustained manner for more than 72 hr (Fig. 4). Atomic Force Microscopy Topographic, phase mode images and mean nanoparticle radius of NEM, NMI and NMG are shown in Figs. 5-7. The dilution/drop-casting samples preparation method allowed the visualization of individual nanoparticles without aggregation even for NMG samples. All samples exhibited a globular morphology with an average diameter of 200, 250 and 220 nm for NMI, NEM and NMG, respectively. Stability studies The particle size of MKT, NEM, NMI and NMG remained unchanged when stored at refrigeration temperature and room temperature. However, there was an increase in the particle size and a decrease in drug content for the different formulations when they were stored at accelerated conditions (40¡2uC/75¡5% RH) for 6 months. The average particle size for the nano formulations increased from the initial size from 0 to 6 months as follows; NEM: 192¡13 to 241¡24 nm, NMI: 65¡11 to 102¡23 nm and NMG: 139¡18 to 185¡26 nm. The changes in particle size at 25¡2uC/60¡5% RH were; for NEM: 192¡13 to 208¡11 nm, NMI: 65¡11 to 81¡14 nm and NMG: 139¡18 to 156¡16 nm and the changes at refrigeration temperature (2-8uC) were; for NEM: 192¡13 to 201¡10 nm, NMI: 65¡11 to 77¡12 nm and NMG: 139¡18 to 145¡14 nm. Similarly, there was decreased drug content observed for both ACE and CAP for all the formulations; however, the decrease was more prominent for MKT than the nano formulations (Fig. 8). Ex vivo skin permeation through dermatomed human skin The permeation of ACE and CAP into the receiver compartment is illustrated in (Fig. 9). The cumulative amounts of ACE and CAP permeating into the receiver compartment from NMG after 24 hr were 5.68 and 5.63-fold more than FD and 2.02 and 1.97 fold more than MKT, respectively ( Table 2). The SC, epidermal and dermal retention of ACE and CAP after 24 hrs for the different formulations are shown in (Fig. 10). The amounts of ACE and CAP retained in the SC for NMG after 24 hr was 4.96 and 5.01-fold more than FD and 1.78 and 2.28-fold more than MKT, respectively. The retention of ACE and CAP after 24 hr was also determined to be significantly more for NMG than both the FD and MKT. The retention of ACE and CAP in the epidermis for NMG after 24 hr was 8.98 and 4.39-fold more than FD and 2.28 and 1.83-fold more than MKT respectively, where as the retention of ACE and CAP in the dermis for NMG after 24 hr was 6.58 and 4.60 fold more than FD and 2.0 and 1.80-fold more than MKT, respectively ( Table 2). Visualization of skin penetration Results from the confocal laser scanning microscopy confirmed that FITC loaded NMG was able to penetrate more and deeper into different skin layers compared to FITC loaded NEM and NMI (Fig. 11). Microscopic studies also demonstrated that there was a decreased green fluorescence signal from FITC with increased Nanomiemgel -A Novel Drug Delivery System for Topical Application depth in cryotomed skin sections with NEM and NMI where as with NMG significantly intense green fluorescence signal was observed especially in skin sections correspond to 260-400 mm depth. In vitro/in vivo microdialysis recovery studies The in vitro recovery of ACE and CAP estimated from the bulk solution was 29¡6% (r2 less than 0.99) and 26¡4% (r2 less than 0.99). The plot of ''Dialysate'' versus ''bulk'' concentrations exhibited a linear relationship at the concentrations tested. ACE and CAP were both detectable in the dialysates of NMG for 24 hrs, whereas in the case of MKT, ACE was not detectable at the 24 hr time point and CAP was detectable at only three time points i.e. 10, 12 and 14 hr. The respective C max of ACE and CAP for NMG (26.24 mg/mL and 0.098 mg/mL) were 2.94 and 2.09-fold more than MKT (8.94 mg/mL and 0.047 mg/mL) (Fig. 12) which signifies the supra bioavailability of NMG. The t max of both ACE and CAP for NMG were 18 hr whereas for MKT, the t max for ACE and CAP were 14 hr and 12 hr respectively. Imiquimod (IMQ) induced psoriatic plaque like model The scaling of the dorsal skin of the mice, a phenomenon typical for psoriatic like skin lesions, was observed after topical application of IMQ (Fig. 13). The PASI scores for all the groups were 4 at the end of 5th day of IMQ topical application. However, after 5 days of treatment, there was significant reduction in the inflammation induced by IMQ for NMG (p,0.01) and MKT (p,0.05). The PASI scores for FD, MKT, NEM, NMI and NMG formulation treated groups were decreased to 3.55¡0.17, 2.70¡0.15, 2.05¡0.13, 1.35¡0.10 and 0.75¡0.07, respectively (Fig. 14). There was a significant decrease in drug content at 6 th month compared to 0 month in the formulations stored at 40uC but not in the samples stored at 25uC and 2-8uC and the decrease was more significant with marketed formulation. Data represent mean¡SD, n56, significant where *p,0.05, **p,0.001, NS: Not significant compared to % drug content in first month. Also, imiquimod caused a pronounced increase (26.00¡3.02 to 174.00¡16.27 mm) in the ear thickness of the mice by the third day of topical application. However, NMG significantly (p,0.01) reduced the ear swelling by 3.22 and 2.01-fold more than FD and MKT, respectively. After topical application of the different formulations for 5 consecutive days, the ear thickness was decreased to 89¡21 mm, 56¡15 mm, 48¡13 mm, 42¡10 mm and 27¡6 mm for FD, MKT, NEM, NMI and NMG respectively (Fig. 15). Further, topical IMQ increases spleen mass and alters its cellular composition. Increase in the percentage of macrophages and plasmacytoid dendritic cells are some unique features in IMQ treated mice. There was an increase in the spleen weight for IMQ only treated group which was significantly (p,0.01) reduced for NMG treated group at the end of the study, compared to FD and MKT. The spleen weights of FD, MKT, NEM, NMI and NMG were 1.18, 1.94, 2.69, 3.34 and 4.63-fold less than IMQ (positive control; PC). The average spleen weights for the (Fig. 16). Histology Analysis of H&E stained sections of IMQ-only treated skin (Positive control; PC) showed increased epidermal thickening with elongation of epidermal rete ridges, disturbed epidermal differentiation and infiltration of leukocytes into both the dermis and epidermis (Fig. 17). However, NMG showed minimum epidermal thickening and extension of the rete ridges with intact SC that was comparable to the negative control (no IMQ treatment, no drug treatment) but much less than Immunohistochemistry (IHC) There was pronouncedly less brown staining for the IL 23 protein for NMG and NMI treated groups compared to the MKT, NEM and FD, which showed extensive brown staining that was comparable to the positive control (PC; IMQ only treatment). The general pattern of brown staining for IL-23 observed was NC,NMG,NMI,,MKT,NEM,,FD,PC (Fig. 18). Permeation through psoriatic like inflamed mice skin The objective of the permeation study through the inflamed psoriatic like skin portions of the mice treated with the different formulations was to find out if there was any significant difference in the permeation resulting from the topical treatment with the different formulations. Permeation of ACE and CAP through the psoriatic skin portions of mice treated with different formulations is depicted in Fig. 19. The amounts of ACE and CAP that permeated into the receptor compartment from NMG after 24 hr through the PC and NC treated psoriatic skin portions were found to be the highest and lowest, respectively, among all the results observed. The amounts of ACE and CAP that permeated into the receptor compartment after 24 hr through the FD and NMG treated psoriatic skin portions were similar to the PC and NC, respectively (Fig. 20). NMG treated psoriatic skin portion showed significantly more permeation of ACE and CAP from NMG than FD (p,0.01) and MKT (P,0.05). The permeation was 2.0 and 2.08-fold more than FD and 1.39 and 1.52-fold more than MKT of ACE and CAP, respectively (Table 3). Combination Index The combination Index (CI) values for the different parameters like cumulative percent of drug permeated into the receptor compartment through the dermatomed human skin, amount of drug retained in the different layers of the skin and the inflamed ear thickness of mice after induction by IMQ was in the range of 0.71 to 0.76. Discussion It is desirable to deliver the therapeutic agent for treating inflammatory skin diseases into the deeper epidermal and dermal layers to achieve maximum therapeutic success [27]. Many researchers have utilized nanoparticulate systems to enhance the skin permeation and deposition of the payloads they carry. However, permeation studies have demonstrated that nanoparticles do not cross the SC but possibly lodge in the SC layers and release the encapsulated drug in a controlled manner into the upper epidermis; the released drug then passively diffuses into the skin layers below [28]. Although, recent investigations have demonstrated that nanoparticles can transport the loaded agents into the skin layers through hair follicles, the total amount reaching the dermal site is very limited [28]. In the present study, we prepared a novel drug delivery system NMG by combining two different drug delivery systems (NEM and NMI) and incorporating them into a carbopol gel. The performance of the combination was compared with the individual drug delivery systems both in in vitro and in vivo and we observed that there was significantly (P,0.05) enhanced permeation of both ACE and CAP into the deeper skin layers delivered by NMG than the FD, NMI, NEM and MKT respectively. Carbopol was selected as the gelling agent for NMG because it forms a homogenous dispersion due to its high water solubility and ability to form a transparent gel after the addition of an alkaline reagent [29]. The effect of the different amounts of carbopol on the viscosity profile and particle size of NMG was investigated because minimum particle size was desired for the viscous gel. The 2% w/v carbopol gel showed greater viscosity and particle size compared to the 1% carbopol gel possibly because there was increased cross-linking of the polymeric network; hence it was selected for the rheological investigations of NEM, NMI and NMG because the spreadability and contact time of a topical gel on the skin surface is directly related to its rheological behavior [30]. Also, the texture analysis showed that there was higher firmness (294.61 g) and consistency (1402.70) but lesser cohesiveness (241.12) of the MKT gel and this can be attributed to the nature of the ingredients used to prepare the gel. Further, the lack of sufficient water content possibly caused the gel to harden and this contributed to the high resistance observed during the entry of probe into the gel. This resulted in the force 1 & area of force vs time (2158.45) between line one and two of TA curve being higher. The texture parameters of the NMG contrarily were close to an average of the parameters for NEM and NMI because NMG is a combination of the two. The substantial firmness and consistency properties observed for NEM were possibly due to the higher oil content of the nano formulation. However, there was a higher cohesiveness observed generally in the nano formulations than the MKT due to the addition of 0.5% Pluronic F-125 to the 2% carbopol gel. This possibly resulted in the enhanced permeation of the drugs observed, through the formation of a thin film on the skin surface and the production of a good occlusive effect. Further, the nano formulations stored at 2-4uC and room temperatures showed excellent stability without significant increase in the particle size and drug content during the entire study period, which may be attributed to the hydration effect of carbopol. However, a gradual increase in both the particle size and distribution was observed at accelerated storage conditions, possibly because there was an increased loss of water at the higher storage temperatures, which caused the aggregation of the particles and the destabilization of the system. Importantly, the in vitro drug release from NEM, NMI and NMG followed Korsmeyer Peppas kinetics demonstrating that the drug was released in a controlled manner through |
a combination of both diffusion and erosion mechanisms. Also, the controlled drug release from NEM, NMI and NMG may be due to the longer diffusion pathway caused by the entrapment of the drug in nano formulation [31]. The skin permeation studies on the other hand showed a significant (p,0.05) increase in ACE and CAP penetration into the receiver compartment as well as drug retention in the deeper skin layers generally for the nano formulations (NEM, NMI and NMG) compared to the FD and MKT. The analysis also showed that NMG formed drug deposits in the skin layers, which resulted in the sustained release of the drug over 24 hr with a concentration that was about 3-fold higher than MKT, although MKT also induced sustained drug release, possibly due to the inherent strong protein binding nature of both ACE and CAP. The penetration of NMG through the skin can be due to the usage of both of the routes that are utilized by NEM and NMI; hence, NMG permeated the skin by both paracellular and passive diffusion across the skin cells to deliver its payloads into the deeper layers of the skin. As every drug delivery system is unique, its rate, extent and mechanism of absorption depend on the size, charge and composition of the drug delivery system. So, when a combination of completely different drug delivery systems (NEM+NMI) was utilized for the delivery of a drug, the absorption of the combined system (NMG) was found to be better than either of the individual drug delivery systems due to the utilization of the maximum possible paths of absorption available for that particular drug. In this work, micelles were prepared using vitamin E TPGS and it was established fact that surfactants disturb the arrangement of cells temporarily and thereby enhance the absorption of poorly soluble drug molecules which otherwise called para-cellular absorption whereas emulsion was prepared using olive oil and because of their lipophilicity and nano size they were assumed to be taking the advantage of trans-cellular route. This was also evident from the observation of more intense green fluorescence signal in deeper skin layers from FITC loaded NMG clearly demonstrated its superior skin penetration ability over NEM and NMI thereby strengthened the hypothesis of the multi absorption mechanism with combination of two different drug delivery systems. These results were also comparable and well correlated with the in vitro skin permeation studies conducted through dermatomed human skin. The microdermal dialysis study of NMG vs MKT demonstrated that ACE and CAP could be safely delivered into the skin of hairless rat in an enhanced approach with NMG more than the MKT. There were no signs of irritation observed even after 24 hrs at the site of application with both NMG and MKT. The higher C max values of NMG signify its better permeation effect possibly because of the presence of permeation enhancers, the nano size and the unique combination of the two drug delivery systems (NMI and NEM). The increased t max and elimination half life of NMG signifies the sustained release of ACE and CAP, which could be because of the efficient entrapment of both drugs in the nanocarrier. It should be noted that both gels (NMG and MKT) were left on the surface of the hairless mice skin for 24 hours. Aceclofenac, the glycolic acid ester of diclofenac inhibits the cyclo-oxygenase enzyme (COX) involved in the production of prostaglandins that cause pain, swelling and inflammation [32]. Capsaicin on the other hand, selectively binds to a protein known as TRPV1 that resides on the membranes of pain and heatsensing neurons. Capsaicin helps in the treatment of inflammation by depleting the presynaptic substance P, which is one of the body's neurotransmitters for pain and heat [33]. During skin inflammatory disorders like psoriasis, various cytokines, chemokines, eicosanoids and substance P has been reported to play vital roles in the regulation of the inflammatory process. Psoriasis is an acute inflammation, characterized by classical symptoms such as heat, redness, scaling, skin thickness, swelling and pain. Further, IMQ has been reported to induce skin inflammation in mice that exhibits the classical phenotypic and histological features of human psoriasis including the up regulation of cytokines like IL-23 and IL-17 [34]. Therefore the efficacy of NMG was investigated in an IMQ induced psoriatic-like plaque model developed in mice. Note worthily, the H&E staining demonstrated all the classical symptoms mentioned above for the clinical psoriasis presentations, for the positive control (IMQ only treatment). Further, the extent of epidermal thickening, elongation of epidermal ridges and brown staining for IL-23 protein was much less for the nano formulations (NMG, NMI and NEM) and MKT than FD. The histological characteristics of skin treated with FD were comparable to the positive control with extensive brown staining of IL-23, hence demonstrating the inefficiency of the formulation to adequately deliver therapeutics amounts of ACE and CAP at the epidermal and upper dermal regions for anti-inflammatory actions. Also, all the nano formulations showed less epidermal thickening and acanthosis than MKT possibly due to the improved delivery of drug into the deeper skin layers by its multi absorption mechanisms. However, among the nano formulations, the NMG exhibited markedly reduced inflammation and histological features that were comparable to the negative control (no IMQ and no drug treatment) and much less than the NMI and NEM (p,0.05). This suggests that because the NMG is a combination of NMI and NEM, it possesses improved topical delivery properties that make it superior to the two formulations. This observation was also notable in the significant (p,0.01) decrease in spleen weights caused by the 5-day treatment with NMG following IMQ-induced inflammation, compared to MKT and all the other treatment groups. There was no significant difference in the spleen weights between the NC (Negative control) and NMG and this is suggestive of the fact that NMG was capable of delivering therapeutic amounts of both ACE and CAP into the epidermal and dermal layers of the skin to reduce the inflammation induced by IMQ treatment. Permeation through psoriatic inflamed skin is expected to be less because of the development of plaque, scaling, epidermal alterations, epidermal thickening and elongation of epidermal ridges, which create barriers to the penetration of drugs through the skin. This hence may account for the decreased transport of drugs into the skin by topical drug delivery systems during psoriasis treatment. This observation has been demonstrated by Shinji Oshima et al. [35] who have shown in their studies that flurbiprofen plasma concentration following the topical application of ZEPOLAS was suppressed in rats presenting with inflamed skin. They also observed that there was decreased permeation of the drug into the receptor compartment through the inflamed rat skin compared to the normal skin in in vitro and microdialysis experiments. The results from the present study were hence in agreement with their report because after 24 hr in vitro permeation studies with NMG through the differently treated inflamed rat skin (5 day treatment with NMG, NMI, NEM, FD and MKT following IMQ-induced skin inflammation), there was a significant (p,0.01) increase in penetration of ACE and CAP into the receptor compartment through NMG-post treated inflamed skin compared to MKT and FD. This observation further illustrates the efficiency of NMG to deliver ACE and CAP into the deep layers of the skin for the complete inhibition of inflammation of the skin to a level that was comparable to the negative control (no IMQ treatment). To ascertain whether the combination of NEM and NMI (NMG) was additive or synergistic, the combination index value was calculated for all the in vitro and in vivo tests conducted. The combination Index (CI) values calculated for the different parameters were in the range of 0.71 to 0.76 indicating the moderate synergism of NMG. The reason for this synergism could be the utilization of all the possible absorption pathways by NMG, which results in the enhanced permeation and retention of drug in skin and thereby produces a superior therapeutic effect than either NEM or NMI. Conclusion Our studies demonstrate that the NMG comprising NEM and NMI enhanced the skin permeation of aceclofenac and capsaicin by translocating the nanoparticles across the deeper skin layers by improving the skin contact time, hydrating the skin and by forming a thin layer on the skin surface (occlusive effect). Thus, the increase in skin permeation of aceclofenac and capsaicin was further responsible for the improved therapeutic response of the psoriatic plaque like model to the treatment of ACE and CAP-containing NMG suggesting the potential of this combination therapy to treat psoriasis. Further, the data of microdialysis demonstrates that the NMG was superior to the MKT. This approach can also be extrapolated to the delivery of therapeutics for treating other skin diseases like fungal, bacterial, viral infections and skin cancers like melanoma. However, more experiments are necessary to prove the absorption mechanism of NMG to find out the exact factors contributing to its synergistic effect over NMI and NEM. Future studies will be directed to the use of more potent drugs with the NMG approach for the treatment of various skin disorders like psoriasis, allergic contact dermatitis, acne etc. Bioterrorism & Biodefense Brucella: Molecular Diagnostic Techniques in Response to Bioterrorism Threat Brucellosis, a worldwide zoonosis caused by members of the genus Brucella , is responsible of a considerable human morbidity and economic losses. Although the disease is associated with low mortality and has a relative limited medical impact, Brucella spp., particularly B.melitensis and B. abortus , have been also reported as possible biological weapons. A prompt detection and identification of involved biological agents and the following discrimination between natural outbreaks and/or intentional release of micro-organism, represents the crucial point for an effective response. Furthermore, being members of the genus Brucella genetically homogeneous, the development of accurate strain typing methods is essential in order to investigate the source of an epidemic event. The aim of this paper is to provide an overview of the current molecular diagnostic tools developed as response to bioterrorism episodes. bacteria constituting an infectious aerosol dose (10-100 organism), the nonspecific clinical symptoms of brucellosis, the worldwide circulation of the infection, the onset of chronic debilitating disease, make of Brucella spp. a category B bioterroristic agent, according the Center for Disease Control and Prevention definition [22]. Although Brucella is sensitive to inactivation standard methods as heating and disinfectants, it often survives for up to two years in the environment [20]. The most virulent of several strains of the Brucella bacteria, code name US, was the most advanced and the only standardized agent fill by the end of the 1950. By the summer of 1951 the Chemical Corps Biological Department scheduled the production of B.suis and B.melitensis. In 1954, B. suis became the first agent weaponised by the USA and tested on animals. By 1955, the USA filled cluster bombs with this agent for the US Air Force at the Pine Bluff Arsenal in Arkansas [23]. In 1967 the development of Brucella as a bioweapon was stopped, and Richard Nixon on 1969 in the Statement on Chemical and Biological Defense Policies and Programs unilaterally renounced to use chemical weapons and the banned the development of all biological weapons. A preliminary treaty prohibiting the development, storage acquisition of biological weapons was completed in 1972 and ratified as Biological and Toxin Weapons Convention (BTWC) in 1975 from 144 countries, However several nations of the Middle East did not sign the treaty and Soviet Union, in spite of the Convention, expanded its biological weapons program [24]. In 1999 a case of brucellosis was reported in a 38-year-old woman who resided in New Hampshire. Even if the atypical clinical presentation and suspicious circumstances raised the possibility of a case of bioterrorism, it was impossible for criminal investigators to find evidence of biological terrorism act. [25]. In fact, the correlation of suspicious cases to a possible attack involving a biological agent represents the mainly difficult in the forensic investigations. An effective public health response to a possible biological terrorism crime or terrorism threat include 1) sensitive, specific, and rapid laboratory diagnosis of patients and characterization of biological agents; 2) early detection through improved surveillance; 3) effective communication; and 4) coordinated local, state, and federal response in the investigation of unusual events or unexplained illnesses [25]. The early detection is essential to ensure a |
prompt response to a biological terrorist event, but also the discrimination between natural outbreaks and/or intentional release of micro-organism agents is of crucial importance in the context of the bioterrorism. Therefore it is very important to have a strain typing epidemiological tool for source trace back in outbreaks. Characterization of Brucella at species and biovar level using differential microbiological approaches for phenotyping often may result in complicate interpretation where a more accurate identification is necessary [26,27]. Furthermore, these typing methods are time consuming and potentially hazardous for laboratory oper ators, as Brucella spp. need BSL3 facilities. Thus, genetic characterization using molecular DNA technology has been developed and several molecular techniques for subtyping have been proposed. Pathogenesis and clinical diagnosis Brucella species are facultative intracellular bacteria able to multiply within human or animal phagocytosis cells, survive to intracellular conditions and escape to the host's immune system [3,28,29]. Epithelial cells, placental trophoblasts, dendritic cells and macrophages are the target cells [30,31]. Transmission of brucellosis to humans is usually the result of direct or indirect contact via ingestion or inhalation, or through conjunctiva or skin abrasions [15,31,32]. The brucellosis is a typical occupational disease affecting especially farmers, veterinarians, abattoir workers and laboratory per sonnel exposed to aerosolisation [33]. Interhuman transmission is rare and only, some anectodic cases are reported as following blood exposure, primary exposure to infected tissues or after sexual contact [20]. Human brucellosis present in various forms with signs mostly non-specific and similar in patients whatever the route of transmission. The main symptoms are fever or chills, arthralgia, sweating and hepatomagaly and splenomegaly [34]. Diagnosis The "gold standard" in the diagnosis of brucellosis is bacterial isolation from blood or bone marrow specimens that requires long cultivation per iods ( 4 to 7 days up to 40 days) and often the blood cultures are unsuccessful [35]. Serological tests, as serum agglutination test (SAT), Rose Bengal test, complement fixation test, and enzymelinked immunosorbent assay are still frequently used [36]. Since the routine identification and differentiation of brucellosis suspected specimens, based on culture isolation and phenotypic characterization, requires BSL3 protocols for the high risk of laboratory-acquired infections [37,31], molecular methods have been explored in order to overcome these difficulties. Furthermore,the Polimerase Chain Reaction (PCR)-based assays have shown a higher sensitivity respect to the standard microbiological assay for the diagnosis of brucellosis. [15]. Molecular methods for Brucella spp. genotyping PCR specie-specific: PCR DNA-based methods such as gene probes and PCR utilize primers derived from different polymorphic regions in the genomes of Brucella species. Different PCR methods for the detection of Brucella spp. that utilize primers derived from different polymorphic regions in the genomes of Brucella species as i.e. (1) a gene encoding a 31-kDa B. abortus antigen which is conserved in all Brucella species (primers B4/B5) [38], (2). a sequence 16S rRNA of B. abortus (primers F4/R2) [39], (3). a gene encoding an outer membrane protein of 26-k-Da (omp-2) (primers JPF/JPR and primers P1/P2) [40,41], (4) outer membrane proteins (omp 2b, omp2a and omp31) [42], (5) proteins of the omp25/omp31 family of Brucella spp [43], the entire bp26 gene of B. melitensis 16M, encoding the BP26 protein (omp 28) (primers 26A/26B) [44] were described [45,46].However these techniques allow the differentiation of limited number of species. The comparison of PCR sensitivity for Brucella DNA detection shows different values for distinct assays, i.e the limit of sensitivity was 8 fg for B4/B5, 5 pg for F4/R2 and 20 pg for JPF/JPR [45]. Decrease of PCR sensitivity was observed in presence of human genomic DNA for primers F4/R2 and B4/B5, from 8 fg to 800 fg and from 5 pg to 50 pg respectively, while JPF/JPR were not affected. Another comparison evaluating the sensitivity of the PCR primer pairs B4/B5, JPF/JPR, P1/ P2 and 26A/26B, applied in about 5000 samples (buffy coat, wholeblood, and serum) was described [47]. The results of the study showed a detection limit for B4/B5 and JPF/JPR primers pairs of 10 to 100 fg and 25 to 250 fg, respectively while the sensitivity for P1/P2 and 26A/26B primers pairs was of 12.5 to 125 fg and 20 to 200 fg respectively. All four assays had also an excellent diagnostic sensitivity ranging from 95.5 to 100% in acute infection, depending on the PCR assay and the type of specimen. As blood is known containing inhibitory substances for PCR, the PCR detection limit was investigated [48] testing four primers pairs including B4-B5, ISP1-ISP2, F4-R2, JPF-JPR and modifying the previously reported methods [45,46]. Results indicated that the detection limit varied between 25 to 800 CFU/ml, depending on the extraction and amplification method. B4-B5 was the most sensitive primers pair (25 and 100 CFU/ml suspended in one ml water and blood, respectively) followed by ISP1-ISP2 and F4-R2, while the JPR-JPF pair was unable to detect Brucella DNA. These data were apparently in conflict with the results published by Navarro et al. [45], indicating F4/ R2 as the most sensitive primers, but the differences could be due to the different DNA sources. A PCR assay using seven individual reactions for the rapid detection of the Brucella genus, and the differentiation among six recognized Brucella species, was described [49]. This assay, that can be used in both real-time and conventional PCR, used the multiple insertion element, IS711, which is stable in both number and position in the Brucella chromosomes as a target. The PCRs for species differentiation were based on unique genetic loci of B. melitensis, B. abortus, B. suis, B. ovis, B. canis, and B. neotomae. Multiplex PCR typing Several multiplex PCRs which identify the genus Brucella at the species level and partly at the biovar level using different primer combinations, have been reported. The first multiplex PCR, called AMOS PCR assay (AMOS is an acronym from ''abortus-melitensisovis-suis''), comprised five oligonucleotide primers for the identification of selected biovars of four species of Brucella. The assay exploited the polymorphism arising from species-specific localization of the genetic element IS711 in the Brucella chromosome. Identity was determined by the size of the product amplified from primers hybridizing at various distances from the element. This method could identify three biovars (1, 2, and 4) of B. abortus, all three biovars of B. melitensis, all B. ovis biovars and biovar 1 of B. suis. An abbreviated multiplex AMOS PCR assay based on three additional primers was developed to differentiate B. abortus vaccine strains S19 and RB51 from field strains [50]. In 2005 the finding of a deletion next to one of the IS711 copies in B. abortus biovars 5, 6, 9 and in some field strains of biovars 3 of B. abortus has allowed to design and add a specific primer to the eight primer mixtures of AMOS PCR, allowing to enhance the discrimination power of this assay [51]. A RAPD-PCR (random amplified polymorphic DNA) was used in order to develop a multiplex PCR that uses the AMOS primers, additional specific loci of the insertion element IS711, and other unique insertions and deletions. This novel PCR assay differentiates between all presently recognized Brucella species, including the recently described species B. ceti (formerly named 'Brucella maris' or 'Brucella cetaceae'), B. pinnipedialis (formerly named 'Brucella maris' or 'Brucella pinnipediae'), and B. microti, including some more recently described strains of the latter species [7][8][9], and also allows accurate differentiation of certain biovars of B. abortus and B. suis [52]. A new generation of multiplex PCR assays has been developed on the basis of the knowledge arisen from the recent availability of genome data. Garcıa-Yoldi et al. [53] described a multiplex PCR assay for the identification of all six classical species, Brucella isolates from marine mammals, the vaccine strains B. abortus RB51 and S19 and B. melitensis Rev 1. The eight species-specific primer pairs amplified fragments of different sizes that showed a unique profile for each species following agarose gel electrophoresis. However, this multiplex PCR was unable to differentiate B. microti from B. suis and B. ceti from B. pinnipedialis. A similar multiplex approach based on species-specific differences was recently described as being able to distinguish the six classical species but with some problems with B. canis and B. suis differentiation [49]. In addition some single target PCRs have proven particularly useful e.g. the presence of an insertion sequence, IS711, downstream of the bp26 gene, a feature specific to the marine mammal Brucella strains [54]. An advancement of the Garcia-Yoldi protocol for the differentiation of all currently described Brucella species was published by Mayer-Scholl et al. [55]. The primer pair identifying B.microti [7] was included in the multiplex PCR described by Garcıa-Yoldi et al. [53], and the assay was set up on the DNA of Brucella reference strains and field isolates. The assay allowed the identification of all currently known Brucella, distinguishing also between the marine species B. ceti and B. pinnipedialis and identifying the recently described species B. microti and B. inopinata. Real-time PCR Real-time PCR is more rapid and more sensitive than conventional PCR. It does not require post amplification handling of PCR products, thereby reducing the risk of laboratory contamination and false-positive results. Real-time PCR assays have been recently described in order to test Brucella cells [56], urine [57], blood, and paraffin-embedded tissues [58]. Three separate real-time PCRs were developed to specifically identify seven biovars of B. abortus, three biovars of B. melitensis and biovar one of B. suis using fluorescence resonance energy transfer [56]. The upstream primers used in these real-time PCRs derived from the insertion element, IS711 whereas the reverse primer and FRET probes are selected from unique species or biovar-specific chromosomal loci. Sensitivity of B. abortus-specific assay was as low as 0.25 pg DNA corresponding to 16-25 genome copies and similar detection levels were also observed for B. melitensis and B. suis-specific assays. Light Cycler real-time PCR with SYBR Green I targeting bcsp31, a gene found in all Brucella species and biovars, was described [57]. The assay was per formed on DNA extracted by urine samples and showed a sensitivity of 10 fg corresponding to one genome copy. Another real-time PCR assay for the rapid laboratory diagnosis of human brucellosis on whole blood and paraffin-embedded tissues was developed [58] using three assays with hybridization probe detection. These assays targeted conserved and specific regions of the Brucella genome: the ribosomal 16S-23S ITS region, omp25 and omp31. The ITS-PCR clinical specifity was 100% and showed a limit of detection as low as 3 genome copies per reaction while omp25 and omp31 assays targeting only a single copy gene. Various molecular techniques differentiating Brucella at the species level and/or at the biovar level have been described [53,[59][60][61]. These methods are usually less labour-intensive, faster than biochemical typing but these techniques did not set up with the aim of obtaining clear-cut species and biovar assignment in a very short time for routine laboratory testing. A real-time 5' nuclease PCR assay specific for amplification of a 322 bp fragment of the per osamine synthetase (per ) gene, a highly conserved region present in the naturally rough Brucella species B. ovis and B. canis and spontaneously rough strains of B. abortus and B. melitensis [54], was described by Bogdanovich et al. [62]. The assay showed a detection limit ranging from 200 fg (approximately 40 CFU) to 2 pg (approximately 400 CFU) but the differences could be due to the different DNA template quality obtained from different sources. The per formances of newly designed real-time PCR assays using TaqMan probes and targeting the 3 following specific genes: (i) the insertion sequence IS711, (ii) bcsp31 and (iii) per genes for the detection of Brucella at genus level was described by Bounaadja et al. [63]. The study sshowed that the use of the IS711-based TaqMan realtime PCR assay was specific, with the sensitivity 10 times higher of the two other targets, efficient and reproducible method for the rapid and safe detection of the genus Brucella. The bcsp31 and per targets could be used as confirmatory tools, in order to optimise the diagnostic specificity. High resolution melt The development of a molecular technique which utilizes realtime PCR followed by high-resolution melt (HRM) curve analysis to reliably type members of this genus has been described by Winchell et al., [64]. The assay targeted discriminating loci within the genomes of Brucella spp and through the dissociation curve analysis allowing the accurately identification |
of Brucella isolates at the species level and of unusual Brucella isolates such as BO1 and BO2. This assay also proved successful for discriminating B. suis from B. canis, but was unable to accurately differentiate a B. suis bv4 from B. canis. However, this particular B. suis biovar has previously been reported to exhibit a genotypic pattern identical to B. canis, and it is still debated as to whether this is truly a unique biovar of B. suis [65,66] Restriction fraction length polymorphism (RFLP) based approaches Recently, PCR-RFLP has provided evidence of polymorphism in a number of genes including the outer membrane protein 2 (omp2), the heat shock protein dnaK, htr, and the erythrulose-1-phosphate dehydrogenase gene (ery). Particularly, the DNA polymorphism in omp2a, omp2b, omp25 and omp31 has been found to be useful for the differentiation between the Brucella species and their biovars, including the marine mammal Brucella isolates [59,[67][68][69][70][71]. Results of PCR-RFLP allowed to identify in omp25 a marker for B. melitensis in the form of absence of an EcoRV site though B. suis biovars 3 and 4 and B. canis could still not be distinguished [70]. Other omp genes examined include omp31, known to be deleted in B. abortus [72], but which has markers for B. canis, B. suis biovar 2 and B. ovis [73]. Single nucleotide polymorphisms (SNPs) typing Single nucleotide polymorphisms (SNPs) represent powerful markers that allow accurately describing the phylogenetic framework of a species, particularly in a genetically conserved group as Brucella. The approach is based on a series of discrimination assays interrogating SNPs that shown to be specific to a particular Brucella species. Scott et al. [74] described the use of SNPs in order to develop a multiplex SNP detection assay, based on primer extension technology, that can rapidly and unambiguously identify an isolate as a member of one of the six classical Brucella species or as a member of the recently identified marine mammal group. An alternative approach based on Minor Groove Binding protein (MGB) probes applied on a real-time PCR platform was described [75,76]. The assay distinguishes all members of the classical species, but the differentiation of B. suis and B. canis was difficult as no B. suis specific SNP has been identified. However, as a specific B. canis SNP has been identified [65], it is possible a discrimination with B. suis/B. canis specific SNP and the B. canis specific SNP [2]. A new SNP signatures for the rapid identification and biovar characterization of B. suis was described by Fretin et al. [77]. Allelic profiles unique for each B. suis biovar were defined and the most relevant signatures were determined. Biovars assigned with both present and classical methods were globally consistent except for some biovar 3 field strains which matched the allelic profile of biovar 1. An advancement of this method has represented by a novel SNPbased typing platform that, incorporating targets that define the three Brucella vaccine strains, allows the differentiation of the live Brucella vaccine strains from field isolates [78]. MALDI-TOF-MS spectra (MS spectra) Bacterial identification based on peptidic spectra obtained by matrix-assisted laser desorption ionization time-of-flight (MALDI-TOF) mass spectrometry was proposed 30 years ago [79]. This method represents a new diagnostic tool in established microbiological laboratories [80]. Databases have been developed that include the main pathogenic microorganisms, thus allowing the use of this method in routine bacterial identification from plate culture. Recently, to identify Brucella species a reference library was constructed using 12 Brucella strains. With this 'Brucella library' discrimination was not possible to the species level [81]. Tandem repeat based typing (MLVA) In the last years the availability of microbial genome sequences has facilitated the development of multilocus sequence-based typing approaches such as multiple locus VNTR (variable number of tandem repeats) analysis (MLVA). The VNTR, allelic hyper variability related to variation in the number of tandemly repeated sequences observed at several genomic loci in the Brucella genomes, were used for the discrimination of bacterial species that display very little genomic diversity. The first application of VNTR based typing to Brucella was the HOOF-Prints scheme (Hyper variable Octomeric Oligonucleotide Finger-Prints) published by Bricker [82]. The approach was based on a comparison of the newly completed genome sequences of B. suis and B. melitensis along with a draft B. abortus sequence which identified an eight base pair tandem repeat sequence at nine distinct genomic loci [2]. Eight of the nine loci were variable among the three genome sequences allowing the development of a PCR-based method to identify the number of repeat units at each locus. HOOF-Prints have been employed with entire genome sequences to identify rapidly evolving loci and are capable of differentiating Brucella isolates by the variability in the 8-bp tandem repeat [82]. HOOF-Prints has been used to investigate the clustering of Brucella isolates [83,84]. In 2006 Whatmore et al. [85] described a new scheme that included the eight of the original loci of Bricker as well as an additional 13 newly VNTR loci to give a 21 locus scheme, VNTR-21, that allowed to provide some resolution at the species level. In the same year a scheme labelled MLVA-15, based on a subset of 15 loci that comprises 8 markers with good species identification capability and 7 with higher discriminatory power, was published [86], and followed by MLVA-16, a slight modification of MLVA-15 [87]. MLVA16 has been used to recognise human outbreaks that relate to a common source or to confirm relapse [88] or for tracing back source of laboratory infection [89] as well as to demonstrate heterogeneity in profiles even in a restricted area of endemicity [90] and identify that human B. melitensis isolates from Per u form a distinct cluster from previously described European isolates [91]. MLVA-15 has also been used to assess the stability of a live vaccine [92], to show that wild boar and domestic pigs sharing localities can have identical B. suis genotypes [93], identifies clusters that are congruent with species identification [94]. The MLVA band profiles, obtained by the amplification of the different alleles, may be resolved by different techniques ranging from low cost manual agarose gels to the more expensive capillary electrophoresis sequencing systems. Recently, a more rapid and inexpensive method based on the Lab on a chip technology has been proposed [95]. This miniaturized platform for electrophoresis applications is able to size and quantify PCR fragments, and was previously used for studying the genetic variability of Brucella spp. [96]. On the basis of this equipment a new high throughput microfluidics system, the LabChip 90 equipment (Caliper Life Sciences), applied to the selected subset of 16 loci proposed by Al-Dahouk [87] for MLVA typing of Brucella strains was developed [97]. Conclusion Since Amerithrax incident in 2001, when the deliberate release of anthrax spores via mail within USA caused five deaths and panic spreads across North America, the governments focused on the enhancing of their biosecurity and biosafety programs. It was also highlighted the requirement for the acquisition of new technologies in order to detect and differentiate the biological warfare agents. Particularly, the need for discriminating between natural outbreaks or deliberate microorganism release brought to the upgrowth of new genotyping methods. Although Brucella represents more a biological warfare agent of historical significance than a real threat, it is also the etiologic agent of one of the world's major zoonotic infections, responsible for economic losses and considerable human morbidity [14,98]. Therefore, the world community was motivated to strengthen the control mechanisms for this agent. Since the first signs of a bioterrorism attack could appear in the same way of a natural outbreak, the development of methods for detection and identification of biological warfare agents has been considered a need by many countries, though the genetic homogeneity of the genus hamper s this target. For the direct detection, speciation and differentiation of Brucella spp recovered from clinical and environmental specimens, in the last years different target genes, primer pairs, PCR techniques and extraction procedures have been proposed. Further standardisation and optimisation are necessary to achieve consistency and reliability of results before they are incorporated in routine laboratory investigations because of differences of sensitivity of the assays, due to different DNA extraction methods, detection formats and limits, and to different types of specimens used. Moreover, the development of DNA technologies for the molecular subtyping of Brucella, have been described; for example over the last years wholegenome analysis provided new insights into this genus. The first Brucella genome sequence available was B. melitensis 16 M [99], followed by B. suis 1330 [100], B. abortus 2308 and 9-941 [101,102], the full sequence of the vaccine strain B. abortus S19 [103], B. canis, B. ovis, B. suis biovar 2, B. melitensis biovar 2 and an incomplete B. ceti genome [104] as well as genomes of B. suis biovar 3 and 4 and B. melitensis biovar 3 [105]. At this time, about forty genome sequences from different Brucella strains, representing all species, have been published either as complete genomes [106,107]. or as draft assemblies in the complete genome and whole genome shotgun divisions of GenBank NCBI (National Center for Biotechnology Information. http://www.ncbi.nlm.nih.gov), or can be downloaded from the Pathosystems Resource Integration Center web site (http://brcdownloads.vbi.vt.edu/patric2/PATRIC/). The genomic analysis allowed revealing the genetic divergence within Brucella spp. as the variable number tandem repeat sequences throughout the genome, hyper variability due to recombination between and within the repeats. MLVA represents the most used approach allowing distinction between genetically homogeneus species, as well as the close genetic relationship represented by biovars within a species or clade. The development of MLVA based methods on high throughput capillary gel electrophoresis or microfluidics technology provides robust highresolution typing tools for epidemiological trace-backing, suitable for high throughput analysis and inter-laboratory comparisons, offering a fair compromise among costs, speed and specificity, compared to any of the conventional molecular typing techniques. More recent advances in sequencing technology produced a new class of massively parallel next-generation sequencing platforms such as: Illumina, Inc. Genome Analyzer, Applied Biosystems Solid System, and 454 Life Sciences (Roche) GS FLX that could represent a promising epidemiological and forensic typing tools. However, the advent of bioinformatics, genome-sequencing and high-throughput genome-wide platforms has lead to an impressive enhancement of the data flow. Therefore the storing and the integration of these heterogeneous informations have become the crucial point to facilitate data exchange, analysis and identification of suspect agent. Bioinformatics programs have been adopted by several research centres in order to provide a wide integration of genomic, transcriptomic and proteomic data, which is essential for developing a system-centric resource needed for supporting the research or forensic investigations. These programs allow the storage of data, providing visualization and analysis tools. Furthermore, there are publicly available databases for storing and disseminating data, as MLVAbank for Bacterial Genotyping [108] which works as repositories of individual data types. Moving house: long-term dynamics of corticosterone secretion are unaltered in translocated populations of a rare reptile (the tuatara, Sphenodon punctatus) Translocations are an important conservation tool; however, the process is a potential stressor for most species. We evaluated effects of translocation on plasma corticosterone in a rare reptile (the tuatara). We found that translocated tuatara are resistant to cumulative stressors and show no hormonal sign of chronic stress. Introduction Translocations are human-assisted movements of living organisms from one area to another and are an important tool for conservation efforts and population restoration of species at risk (Armstrong and Seddon, 2008;Ewen et al., 2012;Seddon et al., 2014). The International Union for Conservation of Nature (IUCN) recognizes two types of conservation translocation to restore populations, namely (i) reinforcements, in which individuals are released into an existing population of conspecifics to enhance the sustainability of populations, and (ii) reintroductions, in which individuals are released in a historically occupied area in order to re-establish a population after extirpation. Although these types of movements are ultimately aimed at helping species, the translocation process is inherently stressful, because associated procedures, such as habitat disturbance, capture, handling, processing, captivity, transport and release to a novel environment, are necessary and unavoidable (Germano and Bishop, 2009;Dickens et al., 2010;Parker et al., 2012). In a recent review, Tarszisz et al. (2014) identified physiology as a key disciplinary area that is lacking attention in conservation translocations and also highlighted how physiological data can improve short-and long-term translocation success. In vertebrates, the acute stress response |
produces a rapid increase in glucocorticoid hormone secretion [corticosterone (CORT) or cortisol] to help individuals cope with immediate stressors (Wingfield et al., 1998); consequently, non-essential processes (such as reproduction and growth) are suspended until homeostasis returns. Although the stress response serves to promote immediate survival, prolonged or sustained CORT secretion (typically expected during translocations) can manifest as 'chronic stress' and is generally considered detrimental to overall health and fitness (i.e. the CORT-Fitness Hypothesis; Sapolsky et al., 2000;Wingfield and Sapolsky, 2003;Bonier et al., 2009;Parker et al., 2012;Almasi et al., 2013). In a recent review of the associations between stress and movement of animals, Teixeira et al. (2007) concluded that stress is a contributing factor to the success or failure of a translocation project. Stress induced by the initial translocation process and relocation to a novel environment increases the vulnerability of individuals to reproductive failure, disease, starvation, predation and longrange dispersal, thereby decreasing the chance that individuals will survive and that a self-sustaining population will result (Teixeira et al., 2007;Dickens et al., 2009Dickens et al., , 2010Parker et al., 2012). Measuring and monitoring CORT secretion is the most widely used method for assessing stress in vertebrates (Wikelski and Cooke, 2006;Dickens et al., 2010;Sheriff et al., 2011). Although several factors relevant to translocation efforts influence CORT secretion, studies that assess and monitor stress (by way of CORT secretion) throughout and after the translocation process are limited (Germano and Bishop, 2009;Harrington et al., 2013;Tarszisz et al., 2014). Numerous studies have shown that associated procedures commonly applied in translocation programmes (such as capture, handling and transport) stimulate a significant stress response and influence CORT secretion (Fazio and Ferlazzo, 2003;Langkilde and Shine, 2006;Narayan and Hero, 2011;Bliley and Woodley, 2012;Baker et al., 2013;Fazio et al., 2014). Likewise, altered CORT secretion has been associated with variation of environmental factors, such as exposure to humans (French et al., 2010;Taylor et al., 2014) and novel predators Rödl et al., 2007), change in food availability (Woodley et al., 2003;Kitaysky et al., 2007;Bryan et al., 2014), latitudinal differences (Silverin et al., 1997;Dahl et al., 2012;Eikenaar et al., 2012;Quirici et al., 2014) and habitat type (French et al., 2008;Zhang et al., 2011;Li et al., 2012;Bauer et al., 2013). In addition to experiencing acute stressors during the initial translocation process, translocated individuals released into new environments are faced with several survival challenges (such as finding food and shelter and avoiding predators); therefore, physiological stress is inevitable (Teixeira et al., 2007). Here, we examine the acute CORT response and long-term dynamics of CORT secretion through the translocation process in a rare reptile, the tuatara (Sphenodon punctatus). The tuatara is a protected reptile endemic to New Zealand and is the only living representative of the reptilian order Rhynchocephalia (Jones and Cree, 2012). Although the tuatara is now considered non-threatened but 'at-risk -relict' (Hitchmough et al., 2013), translocations contribute to conservation and ecological restoration efforts and serve to reestablish tuatara within their pre-human range (Hitchmough et al., 2013;Cree, 2014). In addition to easing extinction pressure, translocations also offer a chance to examine and address relevant research questions (Germano and Bishop, 2009;Miller et al., 2012;Cree, 2014). In 2012, wild tuatara were translocated to six island and mainland sanctuaries from two source populations (Lady Alice Island and Stephens Island/ Takapourewa, New Zealand; Cree, 2014), which presented an excellent opportunity to examine the CORT response to translocation in multiple populations. Previous studies have examined patterns of CORT secretion in tuatara; in general, baseline CORT in tuatara is fairly low (with plasma concentrations typically 2-5 ng/ml), a significant CORT response to capture restraint is observed, and female reproductive condition, body temperature and season (but not time of day) are influential factors (Tyrrell and Cree, 1998;Tyrrell et al., 2000;Anderson et al., 2014). In a recent study examining CORT secretion in four populations of tuatara, we found that baseline CORT was similar among all populations; however, the CORT response varied with latitude, seabird density, sex ratio and genetic diversity (L. Anderson, N. Nelson, D. Towns and A. Cree, unpublished data). Although translocations of tuatara continue to take place, CORT secretion (as an indicator of stress) during and after the translocation process has not been examined. Comparing CORT secretion simultaneously in translocated and source populations of tuatara would allow detection of altered CORT secretion that is correlated with environmental and/or habitat change and that would be an indication of chronic stress. This study had two aims. First, we examined the acute CORT response in tuatara at different stages of the initial translocation process and tested the prediction that the acute CORT response would be amplified with cumulative stressors. Second, we tested whether long-term changes in CORT secretion provide evidence of 'chronic stress' in three translocated populations (compared with the corresponding source populations as control animals). We made the following predictions: (i) baseline CORT (post-translocation) would be similar among all populations; and (ii) the CORT response (post-translocation) would be amplified in translocated populations that experienced a marked environmental and/or habitat change (e.g. a greater latitudinal shift). Moreover, body condition indices (mass relative to snout-vent length) in translocated populations (along with source populations as control animals) were examined, because chronic stress can influence energy expenditure (Romero, 2002). Stephens Island is located in the Northern South Island regional climate zone, and Cape Kidnappers Sanctuary is located in the Eastern North Island regional climate zone, both experiencing warm dry summers and mild winters (with frost). Sanctuary Mountain Maungatautari is located in the Central North Island regional climate zone, experiencing warm dry summers and cool winters (with frost and fog; NIWA, 2014). All translocation release sites offered suitable physical habitat for tuatara (with artificial burrows also provided) and social aspects, such as male:female sex ratio (Fig. 2), and tuatara densities were within the normal range. Short-term monitoring: the acute corticosterone response at different stages of the initial translocation process In translocation A only, we examined the acute CORT response (i.e. CORT secretion above baseline) through all stages of the initial translocation process, during which standard translocation protocols were followed (Towns et al., 1990;Cromarty and Alderson, 2013). In summary, adult tuatara (snout-vent length ≥170 mm) that had emerged from their underground burrows were captured by hand (between 20.00 and 04.00 h), and sex was identified by examining secondary sex characteristics, such as head size/shape, body shape, spine shape and crest development (Cree, 2014). All individuals in the translocation programme were subjected to a capture-restraint 'hold' [which involved capture of individuals and initial holding (between 40 and 60 h) in cloth capture bags], processing (which involved handling, weighing, measuring and implantation of a passive integrated transponder tag) and transfer to the release site [which involved holding (between 6 and 10 h) in perforated cardboard postal tubes (10 cm × 50 cm), movement by foot to the helicopter pick-up site, a 30 min helicopter flight, unloading and a 30 min handing-over ceremony upon arrival at Motuihe Island]. To determine the acute CORT response at different stages of the translocation programme, we collected baseline CORT samples (following the blood sampling protocol described below in the 'Sampling protocol' section) from tuatara at capture (0 h; n = 54) and collected a second sample after one of the following: (i) an 18 h hold (n = 15); (ii) a 42 h hold (n = 14); (iii) a 42 h hold + process + transfer (n = 11); or (iv) a 66 h hold + process + transfer (n = 14). Tuatara do not show significant daily variation in baseline CORT (Tyrrell and Cree, 1998); therefore, the time of day at sampling is unlikely to contribute to variation in CORT secretion in this study. Cree, unpublished data), we also analysed samples from the ST source population (obtained in a previous study; March 2012) for annual comparison. In order to determine whether the release site (within a translocated population) had a significant effect on CORT secretion, we collected post-translocation samples from two separate release-site locations on Mot (site 1 = Orchards Bush and site 2 = Von Luckner's Bush) and at MT (site 1 = Tuatarium and site 2 = Northern Enclosure). Sampling protocol In order to determine baseline CORT concentrations, a blood sample (up to 1 ml) was collected within 10 min of capture from the base of the tail with a heparinized 23-gauge needle and 1 ml syringe. After baseline samples were taken, individuals underwent capture restraint in a cloth capture bag and/or postal tube (3-66 h, depending on study), whereupon a second blood sample (up to 1 ml) was taken to determine the CORT response. Internal body temperature (T b ) was recorded with a cloacal thermocouple (Fluke ® Multimeter, model 179; specified accuracy ±0.1°C; Fluke Corporation, Everett, WA, USA) prior to taking blood samples from each individual (both baseline and CORT response). After CORT response samples were obtained, individual mass (in grams) was determined (to the nearest ±5 g) with a 1000 g spring scale (Pesola AG, Baar, Switzerland), and snout-vent length (in millimetres), tail length (in millimetres) and tail regeneration length (in millimetres) were measured with a ruler. Body condition scores were generated for each individual as standardized residuals from a regression of log tail-corrected mass (Newman et al., 1994) and log snout-vent length (Schulte-Hostedde et al., 2005). Body condition scores were generated separately for each source population (ST and LA) and sex. Depending on field conditions (i.e. electricity available or not), blood samples were separated either by centrifuge (5 min at 480 g) or under normal gravity for 6-8 h at 4°C (Reimers et al., 1983;Sheriff et al., 2011;Anderson et al., 2014). Plasma was transferred into cryogenic vials with a micropipette, stored in a cryogenic dry shipper (Thermo ScientificTM, Arctic ExpressTM Dual 10) or in a freezer at −20°C until return to the laboratory, and then stored at −80°C until assayed. Corticosterone was analysed with commercial enzyme immunoassay kits (Cayman Chemical Co., Ann Arbor, MI, USA) using a previously described method validated for tuatara (Anderson et al., 2014). Briefly, CORT was extracted from plasma samples with redistilled dichloromethane, and each sample was assayed in duplicate. For each extraction, a subset of tritiated CORT samples were analysed to measure extraction recovery. Extraction recovery was 106 ± 8% (mean ± SD) with an overall coefficient of variation of 7%. Intra-assay and interassay coefficients of variation were 9.9 and 14.2%, respectively. Statistical analyses Data analyses were carried out using R v3.2.0 statistical software (R Development Core, 2013) and Prism 6 (Graphpad Software Inc., La Jolla, CA, USA). All data were checked for assumptions of normality and were transformed if necessary. Linear mixed effects regression (LMER) models were fitted using the 'lme4' package (Bates, 2013) in R to analyse the following: (i) the acute CORT response during the initial translocation process; and (ii) the long-term dynamics of CORT secretion in translocated populations. Models were constructed through forward/backward stepwise regression procedures (Field et al., 2012). In all LMER models, a random effect of tuatara identity was included to account for repeat sampling of individuals. The 'lmerTest' package (Kuznetsova, 2013) was used to compute P-values for coefficients in final models, and significance was assumed at P < 0.05. Sex (M, F) and linear covariates of body temperature (T b ) and body condition score (log tail-corrected mass/log snoutvent length) were not significant predictors of CORT secretion in this study (LMER, P > 0.05) and were therefore not included in final models. Furthermore, the location of release site (within translocated populations) did not have a significant effect on CORT secretion in either translocation study A (P = 0.775) or B (P = 0.656); therefore, individuals from separate release sites within translocated populations were pooled for further analyses. In analysis 1 (short-term monitoring), log-transformed CORT was the response variable, and sample (baseline, 18 h hold, 42 h hold, 42 h hold + processing + transfer or 66 h hold + processing + transfer) was the input variable. In analysis 2 (long-term monitoring), we first examined whether CORT secretion varied by release site within translocated populations, with log-transformed CORT the response variable and input variables of hour (0 or 3 h), site (site 1 or site 2) and an interaction term of hour × site. Models were fitted to data from Motuihe Island and Sanctuary Mountain Maungatautari, because these translocated populations had two separate release locations. Then, we |
compared CORT secretion between translocated and source (control) populations, with log-transformed CORT the response variable and input variables of hour (0 or 3 h), sample (source pre-, source post-, translocated pre-or translocated post-) and an interaction term of hour × sample. Lastly, we compared body condition pre-and posttranslocation, with body condition score (log mass/log snout-vent length) the response variable and sample (source pre-, source post-, translocated pre-or translocated post-) the input variable. Models were fitted to data from translocation A and B. Results Short-term monitoring: the acute corticosterone response during different stages of the translocation process An acute CORT response (indicated by a significant increase from baseline CORT) was observed in all stages of the translocation process of tuatara from Lady Alice Island to Motuihe Island (Table 1 and Fig. 3). The acute CORT response peaked at 18 h hold and successively decreased (though remaining significantly higher than baseline CORT) at 42 h hold (LMER, t = −4.495, P < 0.001), 42 h hold + processing + transfer (LMER, t = −3.899, P < 0.001) and 66 h hold + processing + transfer (LMER, t = −4.118, P < 0.001; Fig. 3). Contrary to our prediction, cumulative procedures of processing + transfer did not amplify the acute CORT response, because individuals held for 42 h (without processing + transfer) showed a similar acute CORT response to individuals held for 42 h + processing + transfer (LMER, t = 0.305, P = 0.761) and to individuals held for 66 h + processing + air transfer (LMER, t = 0.371, P = 0.712). Corticosterone concentrations in animals experiencing the latter three treatments were significantly lower than in individuals held for 18 h only (Table 1 In translocation B, CORT secretion varied between translocated (CK and MT) and source (ST) populations at 6 months post-translocation. Baseline CORT was significantly lower in one translocated population (CK; LMER, t = −2.345, P = 0.020), but was similar in the other translocated population (MT; LMER, t = −0.925, P = 0.356) compared with the source (ST) population (Fig. 4b). The CORT response was similar in one translocated population (CK; LMER, t = −1.247, P = 0.213), but was significantly higher in the other translocated population (MT; LMER, t = 1.991, P = 0.048) compared with the source (ST) population (Fig. 4b). Corticosterone secretion (both baseline CORT and the CORT response) was similar between the two translocated populations (CK and MT; baseline CORT, LMER, t = 1.210, P = 0.227; and CORT response, LMER, t = 0.745, P = 0.457; Fig. 4b). Corticosterone secretion in the source (ST) population was similar between both pre-translocation samples (March 2012 vs. October 2012; Table 2b and Fig. 4b). In all populations, baseline CORT was significantly higher at 6 months post-translocation (March 2013) compared with both pre-translocation samples (March 2012 and October 2012; Table 2b and Fig. 4b). The CORT response was similar pre-and post-translocation in the two translocated populations (CK and MT), but was significantly lower post-translocation in the source (ST) population (Table 2b and Fig. 4b). Discussion Here, for the first time, we examined CORT secretion throughout the entire translocation process in a rare reptile (the tuatara, S. punctatus). Our findings were as follows: (i) plasma CORT concentrations remain elevated throughout the initial translocation process (short-term monitoring between 18 and 66 h) but are not amplified by cumulative stressors; and (ii) the long-term dynamics of CORT secretion are similar in translocated and source populations. Taken together, our results show that tuatara are generally resilient to cumulative acute stressors and to chronic translocation stress. Cumulative stressors during translocation do not affect the acute corticosterone response in tuatara To our knowledge, this is the first study to quantify the effect of cumulative stressors (routinely experienced in a translocation) on the acute CORT response in a reptile. We expected to see an effect of additive stressors on the acute CORT response, but this was not the case. The CORT response peaked at 18 h of holding/captivity restraint, and additional processing procedures of measurements, microchip insertion and air transfer did not increase CORT secretion further, suggesting resistance to cumulative stressors in this species. Some species show diel variation of CORT secretion (Breuner et al., 1999;Jones and Bell, 2004), which can confound interpretation of results if samples are not taken at 24 h intervals; however, no evidence of a diel cycle has been found in the tuatara (Tyrrell and Cree, 1998). Nevertheless, a significant CORT response was observed throughout all stages of the translocation process, and at no point did CORT returned to baseline concentrations. This observation is consistent with results from our previous study examining the acute CORT response to capture restraint in tuatara, in which a return to baseline CORT concentrations was not observed over 24 h (Anderson et al., 2014). Therefore, we recommend that animal disturbance, holding time, transport duration and post-translocation disturbance be minimized in tuatara to mitigate potentially harmful effects of sustained CORT secretion in individuals directly following translocation. In the present study, we did not examine patterns of CORT secretion in the immediate weeks following translocation (our first follow-up sampling occurred at 6 months post-translocation). Consequently, we are lacking information on the speed of recovery to baseline CORT secretion levels. Langkilde and Shine (2006) found that CORT secretion in male and female lizards (Eulamprus heatwolei) subjected to microchip implantation remained elevated at 14 days (post-treatment) and subsequently increased in response to additional stressors at that time. Likewise, tortoises (Testudo hermanni) that experienced handling plus ground transport had increased baseline CORT at 4 weeks (post-stressor), compared with a control group that experienced handling only (Fazio et al., 2014). These studies have shown that short-term CORT secretion dynamics are significantly altered by processes experienced during a translocation; therefore, obtaining supplementary information on short-term patterns (within 4 weeks posttranslocation) of CORT secretion in tuatara would shed light on the presence of a sustained CORT response and speed of recovery/negative feedback dynamics following translocation. Long-term dynamics of corticosterone secretion in tuatara are not altered by translocation We found that translocation of tuatara did not consistently result in altered CORT secretion relative to control animals (source populations) at 6 or 12 months following translocation (summarized in Table 3). Our results accord with recent studies of translocated reptiles in which CORT secretion was not altered post-translocation. For example, Drake et al. (2012) found that baseline CORT in desert tortoises (Gopherus agassizii) was similar between translocated and control groups at both 1 and 2 years post-translocation, and both Holding et al. (2014) and Heiken (2013) found that baseline CORT and the CORT response in translocated northern pacific rattlesnakes (Crotalus oreganus) were not altered post-translocation, compared with controls. In contrast, Gerber et al. (2004) found that baseline CORT in translocated Turks and Caicos iguanas (Cyclura carinata) remained significantly higher than controls at 1, 5 and 12 months following translocation; however, body condition improved and successful reproduction occurred in translocated animals. Although studies are few, our results add to the general reported trend of resilience to translocation and/or translocation stress in reptiles. In contrast, several studies in mammals and birds have reported significant long-term effects of 9 Conservation Physiology • Volume 3 2015 Research article (Franceschini et al., 2008;Dickens et al., 2009;Zidon et al., 2009;Gelling et al., 2012;Jachowski et al., 2013). However, this observation is not consistent, because other studies have reported no long-term effect (Hartup et al., 2005;Adams et al., 2013;Bosson et al., 2013;Ji et al., 2013), suggesting that adaptation to new environments (indicated by long-term CORT secretion) is species specific or context dependent (e.g. might be due to time of year, weather conditions or hard vs. soft release). Unexpectedly, through long-term monitoring in this study, we observed a significant annual increase in baseline CORT among all source and translocated populations (Table 3 and Fig. 4), probably indicating a ubiquitous environmental effect. The CORT response was unaltered in all populations, with the exception of the Stephens Island source population, where the CORT response was reduced at 12 months post-translocation (Table 3). Body condition declined in the Lady Alice source population and the Motuihe Island and Cape Kidnappers translocated populations. Moreover, these results highlight the importance of collecting information simultaneously from source populations (as a control), because without this our results of increased baseline CORT in all translocated populations and reduced body condition in two out of three translocated populations could have been erroneously interpreted as an indication of chronic stress. It is probable that we detected an unplanned/unexpected effect of drought on baseline CORT secretion in tuatara. In 2012-2013, New Zealand experienced its worst drought in 40 years, with the North Island affected more severely (Porteous and Mullan, 2013). Lance et al. (2010) observed increased plasma CORT in alligators (Alligator mississippiensis) experiencing a severe drought, and recovery of CORT concentrations (to within normal limits) was observed after substantial rainfall. Although dehydration stress was not directly measured (by way of CORT secretion), Davis and DeNardo (2009) found that water supplementation in a long-lived desert lizard (the Gila monster, Heloderma suspectum) led to greater hydration, tail-fat reserves and surface activity. In a recent experimental study, Dupoue et al. (2014) examined CORT secretion in water-deprived snakes (Antaresia childreni) and found that the CORT response, but not baseline CORT, was significantly higher in dehydrated snakes, and the loss of body mass was two to four times greater, compared with controls. The authors suggest that baseline CORT in snakes may respond only to a more severe degree of dehydration and that reduced locomotion (to reduce levels of dehydration) may explain the amplified CORT response in water-deprived snakes (Dupoue et al., 2014). Reptiles, including tuatara, can moderate water loss through behavioural adaptations, such as limiting movement/ locomotion and retreating to (or not emerging from) burrows, caves, fallen logs or undersides of rocks, where humidity is higher (Wilson et al., 2001;Bonnet and Brischoux, 2008;Davis and DeNardo, 2009;Cree, 2014). Dunlap (1995) found that lizards (Sceloporus occidentalis) that were more active (compared with less active) during a drought experienced greater changes in physiological measures (e.g. CORT, weight loss, haematocrit and osmolality). Moreover, Dunlap (1995) suggested that individual variation in behavioural responses of reptiles (e.g. remaining active during drought) can lead to biased analysis of stress in natural populations. Burrowing in tuatara reduces water loss by up to three times the rate experienced when emerged (Cree, 2014). Thus, it is possible that the increased baseline CORT observed in our study is influenced by sampling bias (capturing active individuals out of burrows rather than inactive individuals remaining in burrows). Contrary to our prediction, an amplified CORT response (post-translocation) was not observed in translocated populations experiencing a shift to warmer climates/lower latitudes, specifically the Cape Kidnappers and Sanctuary Mountain translocated populations. In previous studies, we observed an amplified CORT response in tuatara at higher temperatures (L. Anderson, N. Nelson and A. Cree, unpublished data) and at lower latitudes (L. Anderson, N. Nelson, D. Towns and A. Cree, unpublished data). The Stephens Island (40° 40′ S) source population showed a reduced CORT response (from pre-translocation to post-translocation), which was not observed in the Cape Kidnappers (39° 64′ S) and Sanctuary Mountain (38° 30′ S) translocated populations (in which the CORT response was unaltered). Likewise, individuals translocated to Motuihe Island (35° 58′ S) from Lady Alice Island (35° 53′ S) did not show an altered CORT response. It is possible that the individuals translocated from Stephens Island (to Cape Kidnappers and Sanctuary Mountain) would have shown a dampened CORT response (at 12 months) if they remained on Stephens Island or were translocated to equal or higher latitudes. Examining CORT secretion in tuatara populations translocated to equal/higher latitudes [e.g. to Orokonui Ecosanctuary (45° 77′ S) from Stephens Island; Cree, 2014], might clarify the effects of latitudinal/climate change on the CORT response. Although body condition did not statistically influence CORT secretion in our study, the sustained body condition in the Cape Kidnappers translocated population and in the Stephens Island source population suggests better hydration at these sites in the midst of a drought. Porteous and Mullan (2013) report that New Zealand's South Island (close to where Stephens Island is located) was not affected as severely as the North Island, and better hydration in the Cape Kidnappers population was probably achieved through provision of supplementary water sources (L. Anderson, personal observation). |
Reduced body condition has been observed in dehydrated/water-restricted reptiles, including snakes (Dupoue et al., 2014;Lillywhite et al., 2014), lizards (Summers and Norman, 1988;Dunlap, 1995;DeNardo, 2009, 2010), alligators (Lance et al., 2010) and turtles (Ray et al., 2004(Ray et al., , 2008van de Merwe et al., 2013). Clearly, information on relationships among water availability/dehydration, body condition, stress and CORT secretion is lacking and should be considered in light of imminent climate change. Selection of an Appropriate Left-sided Double-lumen Tube Size for One-lung Ventilation among Asians Context: Selecting an appropriate size double-lumen tube (DLT) for one-lung ventilation has always been a challenge as most choose it based on experience or using the existing guidelines based on gender and height. Aims: The aim of this study was to determine if the appropriate choice of this tube could be based on the patients’ height, weight, tracheal diameter (TD), or the left main stem bronchus diameter (LMBD) and also to determine the relationship between height and depth of insertion among Asians. Subjects and Methods: This was a retrospective review of 179 patients who were intubated with a left-sided DLT and also had a posterior-anterior view of a digital chest radiograph for tracheal and left main bronchus diameter measurements. Additional data collected included patients’ demographics and DLT size used. Results: There were 123 (68.7%) males and 56 (31.3%) females with an overall mean age of 33.3 ± 16.3 years. Majority of the males (48.8%) used a size 39 Fr while females (46.4%) used a 35 Fr. There were weak correlations between DLT size with height (male: R2 = 0.222; female: R2 = 0.193), DLT size with weight (male: R2 = 0.109; female: R2 = 0.211), DLT size with TD (male: R2 = 0.027); female: R2 = 0.016), and DLT size with LMBD (male: R2 = 0.222; female: R2 = 0.193). There was a good correlation between depth of DLT inserted with patient's height for both genders. Conclusion: The appropriate size of the left-sided DLT could not be predicted based on patients’ height, weight, tracheal or left main bronchus diameter alone in Asians; however, the depth of insertion of the tube was dependent on the height in both genders. From the records, we included only patients who were managed by one single anesthesiologist, patients whose trachea were intubated with a left-sided Mallinckrodt DLT (Mallinckrodt TM (Covidien, Tullamore, Ireland)) with a posterior-anterior view of a digital chest radiograph (Medweb Telemedicine Intranet Software version 7.0.9, 2014, San Francisco, CA) taken in our center. Patients who had a distorted trachea-bronchial tree on the chest radiograph were excluded. From the patients' records, data collected included patients' age, gender, height, weight, American Society of Anesthesiologists (ASA), type of surgery, size of the left-DLT used, and the level anchored at the incisors. Using the preoperative digital chest radiograph retrieved from the system, measurements of the tracheal diameter (TD) taken at the level between the two clavicle heads and the left main bronchus diameter (LMBD) at one cm below the carina were then performed by a single operator to reduce bias and error. These measurements were done using a digital ruler from the Medweb software as mentioned above whereby calibration has been done, pixel by pixel and zooming the images during measurements would not result in error. Statistical analysis was performed using SPSS v. 21 software (SPSS, Inc., Chicago, IL, USA). Descriptive statistics (mean and standard deviation and percentage) were used to summarize the data. Pearson regression analysis was used to determine the association between the variables. Results A total of 179 records of patients who underwent OLV requiring DLT intubation with a posterior-anterior view of a digital chest radiograph done in our center over the study period were retrieved. There were 123 (68.7%) males and 56 (31.3%) females. The mean age for all patients was 33.3 ± 16.3 years. 116 (64.8%) patients were classified as ASA I, 54 (30.2%) patients in ASA II while the remaining 9 (5%) in ASA III. Majority of the cases (45%) performed were for bilateral sympathectomy for hyperhidrosis through video-assisted thoracoscopic surgery (VATS), 16.2% for VATS bullectomy, 10% for decortication for empyema, and 9.5% for resection of malignant lung tumors. Only 3.4% of patients required a thoracotomy for tuberculosis. The remaining cases were for esophagectomy, spine tumor, scoliosis surgery, and others. All the 179 chest radiographs reviewed had a visible tracheal air column, and only 176 (98.3%) had a clearly visible left main bronchial air column which could be measured. Demographic characteristics and airway dimensions of the patients are shown in Table 1. We further categorized the patients' height into three different height ranges for all patients to compare with the conventional recommendation (based on gender and height range). [3] For the males, 59 (48%) were from the height range of 160-170 cm and only 33 (55.9%) of these patients were suitable for a 39 Fr DLT as recommended [ Figure 2]. As for those measuring >170 cm in height, only 23 (41.8%) out of 55 patients in this category received the recommended size of 41 Fr DLT. Majority of these patients (n = 25, 45.4%) used a size 39 Fr instead. For females with height ranges of 150-160 cm and >160 cm, 30 out of a total of 53 (56.7%) patients were intubated with the recommended size of DLT according to height range [ Figure 3]. TD was successfully measured on all 179 chest radiographs. The average TD for 123 males was 17.2 ± 2.2 mm (range: 10-24 mm) and 15.1 ± 2.0 mm for 56 females (range: 10-20 mm) as shown in Table 1. The LMBD could not be measured from the chest radiograph in three patients who were all males. We found a positive correlation between TD with LMBD for males (R 2 = 0.142, R = 0.38) and females (R 2 = 0.375, R = 0.612), and therefore, unknown values of LMBD can be calculated based on the equation as shown in Figure 4. When comparison was made between DLT size with TD and LMBD, there was weak correlation between DLT size with TD (male: R 2 = 0.027, R = 0.163; female: R 2 = 0.016, R = 0.128) and DLT with LMBD (male: R 2 = 0.222, R = 0.174; female: Pearson regression analysis showed that the depth of DLT was positively correlated to patient's height (male: R 2 = 0.220, R = 0.47; female: R 2 = 0.114, R = 0.34) as shown in Figure 5. The regression lines for both male and female were almost identical and were y = 13.95 + 0.09x and y = 14.92 + 0.08x, respectively. We found that the average depth of DLT insertion in male patients <160 cm in height was 28.1 ± 1.4 cm, between 160 and 170 cm was 28.6 ± 1.3 cm, and for patients >170 cm was 29.8 ± 1.2 cm. For female patients of height <150 cm, the average depth was 26.0 ± 1.7 cm, between 150 and 160 cm was 27.1 ± 1.4 cm, and >160 cm was 27.8 ± 1.3 cm. Discussion In general, one of the difficulties encountered by most anesthesiologists performing OLV is the selection of an optimal size DLT which would best fit the patient. In our center, this selection has always been based on the patients' height and gender as recommended by Slinger which was calculated approximately based on the median values. [3] However, this recommendation may not be appropriate for all patients, especially Asians who are generally smaller. Mackey et al. found that the average tracheal widths for Asians and non-Asians were 19 ± 2 mm (men), 16 ± 2 mm (women) and 21 ± 2 mm (men), 17 ± 3 mm (women), respectively. [4] This was consistent with our findings among Asians whereby the mean TD was 17.2 ± 2.2 mm for males and 15.1 ± 2.0 mm for females. In our study, we found that most of our male population were from the height range of 160-170 cm but only 34 (56.6%) of these patients were suitable for a 39 Fr DLT and those >170 cm in height, a 41 Fr DLT could only fit 41.8% patients as recommended. Fifty (43.5%) of these two categories of patients required a downsizing of the DLT [ Figure 2]. In the female category, only about half of the patients in the category of 150-160 cm and >160 cm used a DLT size based on the recommendation [ Figure 3]. This was not surprising as the TDs in Asians are generally Other methods have been proposed to determine the appropriate size of the DLT which included measurement of TD and LMBD. Brodsky et al. found that direct measurements from the posterior-anterior chest radiograph of the tracheal width at the interclavicular plane can be used as a guide. In their study, the mean tracheal width for men was 20.90 ± 0.32 mm and 16.90 ± 0.25 mm for women and that majority of the men used a 41 Fr DLT regardless of their height or weight with the tracheal width being a better predictor for the selection of the appropriate size DLT. [5] Knowledge of the patient's LMBD could also provide a useful guide to choose the appropriate DLT size. [6][7][8][9] Therefore, if the LMBD is measured from either a chest radiograph or a CT scan and also measurements of the external diameter of the bronchial tip of the DLT are performed, we could then select a DLT with a bronchial lumen which has a slightly smaller diameter. Unfortunately, the anatomy of left main bronchus is not always visible in a chest radiograph. Hannallah et al. could only visualize and directly measure the width of the left main stem bronchus in half of their patients [7] and up to 69% by Hampton et al. [10] Interestingly, we were able to measure the LMBD in 98.3% (95% confidence interval: 93%-97%) of our patients. This may be due to the more effective filmless digital system in our center where the contrast of the image is adjustable which improved the left air bronchogram visibility to help with measurement. In adults, the left bronchial width is directly proportional to the tracheal width, and the ratio between LMBD and TD was found to be 0.69 in men and 0.68 in women at autopsy. [11] Based on the relationship between TD and LMBD, the LMBD could then be calculated by multiplying the TD measured at the level of the clavicles by 0.68. [5] In our study, we found that the ratio between the mean LMBD and TD was 0.69 ± 0.11 in males and 0.70 ± 0.09 in females which was consistent with the autopsy findings. We would also be able to calculate the LMBD values from the measured TD on the chest radiograph using the following equations: LMBD mm = 6.58 + 0.3 (TD mm ) for males and LMBD mm = 3.13 + 0.49 (TD mm ) for females obtained from the Pearson regression analysis as shown in Figure 4. In our study, the mean LMBD for males was 11.8 ± 1.8 mm and females was 10.6 ± 1.6 mm. Chow et al. measured the LMBD using CT scan on Asian-born adults and found that the mean bronchial diameter was 11.6 ± 1.4 mm and 9.6 ± 1.0 mm in male and female patients, respectively, which were very similar to our findings. They selected the size of the left-sided DLTs based on the LMBD measurements and found that they could predict the sizes of the DLT fairly accurately, especially the smaller DLTs. [12] We, however, could not find a strong correlation between the DLT size and LMBD in our patients (male: R = 0.174; female: R = 0.215). The gold standard for the final accurate placement of the DLT with the blue endobronchial cuff just visible below the carina can only be achieved using a fiberoptic bronchoscope. The correct depth of this placement has been reported to be correlated with the patient's height and is taken at the level of the upper incisors or gingival margin. Brodsky et al. found a significant correlation between depth of insertion with patients' height in both genders. As the height increases, the optimal depth of DLT insertion increases. The average depth of insertion was 29 cm for patients 170 cm tall, |
and for each 10 cm increase or decrease in height, average placement depth was increased or decreased by 1 cm. [13] This corresponded closely to our findings as we found a significant correlation between the depth of insertion with height for both male and female patients [ Figure 5]. The regression slope of 0.09 (for males) and 0.08 (for females) showed that a 10 cm change in height predicts a 9.0 mm and 8.0 mm change in the depth of insertion, respectively. Conclusion We found that there was a poor correlation between DLT size with height, weight, TD, or LMBD for both genders. Only about half the patients for both genders used a DLT size predicted according to height range. Most of the remaining patients required a downsizing. All patients had good lung isolation which enabled the surgery to be performed successfully. The depth of insertion was, however, dependent on height. This was a retrospective study, and to ensure that the technique of insertion of the DLT was standardized, we confined our data from cases conducted by one anesthesiologist. One of the main limitations of this study was that the measurements of the TD and LMBD from the digital chest radiograph must not be taken as the absolute values as the images were magnified. Magnification of chest radiograph is based on triangular mathematics, depending on the distance between X-ray source to object and X-ray source to image. Our X-ray source to object was 180 cm and the distance between the plate and the bronchus was about 15 cm. From here, we were able to calculate the magnification percentage which was estimated to be 8.0%. The type of DLT studied was the left-sided Mallinckrodt DLT (Mallinckrodt TM (Covidien, Tullamore, Ireland)); therefore, the findings may not be applicable to the DLTs by other manufacturers. Financial support and sponsorship Nil. Conflicts of interest There are no conflicts of interest. Experiences, distress and burden among neurologists in Norway during the COVID-19 pandemic Background The ongoing COVID-19 pandemic has caused rapid changes in the healthcare system. Workforce reorganization, reduced standard of care and a lack of personal protection equipment (PPE) for health care workers were among the concerns raised in the first wave of the pandemic. Our aim was to explore the experiences, distress and burden among Norwegian neurologists during the first weeks of the pandemic. Methods Hospital-based neurologists in Norway (n = 400) were invited to a web-based survey in April 2020. The study focused on patient management, organizational changes and personal stress during the first weeks of the pandemic lockdown. Work-home interface stress was assessed by the Cooper Job Stress Questionnaire. Results In total, 135 neurologists participated. Seventy-three% experienced a change in their personal work situation, and 67% examined patients with suspected COVID-19 infection and neurological disease. Changed access to resources, and the perception that medical follow-up was unsatisfactory, were associated with a high degree of burden and stress. Neurologists were also worried about the potential lack of PPE and the fear of spreading SARS CoV-2 to close family members. The mean score of work-home interface stress was 2.8 with no significant differences between gender or specialist status. Reduced standard of care was reported for all neurological conditions, and in particular for non-emergency treatments. Conclusion The vast majority of neurologists in Norway experienced a change in their personal work situation during the first phase of the pandemic. The fear of becoming infected and ill was not a major contributor to burden and stress. Introduction The coronavirus disease of 2019 (COVID- 19) was declared a pandemic by the World Health Organization in March 2020 and led to challenges in the delivery of medical care worldwide [1,2]. Initially, there was much uncertainty, and different countries and regions chose different approaches. The major common goal was to reduce cross-infection with Severe Acute Respiratory Syndrome CoronaVirus 2 (SARS-CoV-2) and thus to protect patients and hospital staff in order to avoid critically overloading the healthcare system [3]. Workforce reorganization, changing and increased shift load and downscaling of usual care were among the initial efforts made [3]. In many countries this led to a reduced availability of in-hospital appointments and delayed treatment for chronic diseases [2]. A rapid shift in favor of telemedicine provided continuous access to care despite infection control measures for some patients, but for others this resulted in suboptimal consultations without the possibility of a proper clinical examination and treatment [4,5]. While efforts were made on all levels of the healthcare system, physiciansincluding neurologists-remained on the frontline. Early reports of lack of personal protective equipment (PPE) and deaths among exposed physicians contributed to uncertainty and fear employing about 400 neurologists (approximately 50% residents and 50% senior consultants). Norwegian hospitals are almost exclusively publicly financed. Patients need a referral from a general practitioner in order to visit the hospital, except in emergencies. Norway has an allcovering national health insurance. Thus, all patients are referred on the same conditions and with the same threshold for further investigations, treatments and follow-ups. The first Norwegian national lockdown due to the COVID-19 pandemic was declared by March 12 th , 2020 and the schools and universities gradually re-opened from April 27 th , 2020. The study was conducted as an anonymous online questionnaire survey among Norwegian neurologists about neurological diseases during the primary stage of the COVID-19 pandemic. The survey was distributed by e-mail in April 2020 to hospital-based Norwegian neurologists. The neurologists had approximately three weeks to answer the questionnaire, and the study were concluded in May 2020. Questionnaire and outcomes The questionnaire was based on the authors' clinical experiences from the first weeks of the pandemic in addition to their general knowledge and experience in neurology and health services research. The questionnaire (S1 File) included background variables including age, sex, training status (resident/ senior consultant) and type of hospital (university hospital/ non-university hospital). Furthermore, the participants answered various questions regarding their personal considerations and their own handling of patients with neurological disorders during the pandemic. There were both general questions and questions specific for certain neurological conditions. All questions were asked in relation to the first weeks of the pandemic lockdown. Perceived work-home interface stress was measured using a modified and previously validated version of the Cooper Job Stress Questionnaire [13,14]. To the question "To what extent are the following situations/factors burdening (stressful) for you?", three statements were provided: "The job has a negative effect on my family life / on striking a balance between work and private life / The job has a negative effect on my social life". The statements were scored from 1 (no burden) to 5 (very high burden). The average scores for the three dimensions were calculated into a work-home interface stress score [15]. Statistical analyses For descriptive data, proportions, means, and standard deviations (SD) or 95% confidence intervals (CI) are given. The total number of responders to each question vary, as all follow-up questions were not relevant for all participants, e.g. a neurologist who did not do teleconsultations or treated headache patients during the pandemic did not receive further follow-up questions on this topic. These missing data points were handled by reducing the total number of respondents as applicable. Groups were compared using the t-test (continuous data) or the χ 2 test (categorical data). Non-parametric tests were used as appropriate. Statistical significance was defined by p<0.05, using a two-sided test. Statistical analyses were performed using IBM SPSS Statistics for Windows, Version 26.00 (SPSS Inc., Chicago, IL, USA). Ethics In accordance to the Norwegian law on medical research, the project did not require an approval from the Regional Committee for Medical Research Ethics. The Data Protection Officer at Akershus University Hospital approved the study. Informed consent was obtained by agreeing to participate. All responses were collected anonymously with no identifiable information gathered on the respondents by the study team. The NorPan study is registered in the COVID-19 trial registration at the Norwegian Clinical Research Infrastructure Network. Overall management at neurological departments Overall, 73% reported a change in their personal work situation and 50% had their work schedule changed ( Table 2). More residents than senior consultants had their work schedules changed (60% vs. 42%, p = 0.044) or had extended working hours (25% vs. 9%, p = 0.021). Sixty-seven % had assessed patients with clinical suspicion of COVID-19, with a significant higher proportion among residents than among senior consultants (77% vs. 60%, p = 0.038). There were no significant differences between residents and senior consultants or between women and men concerning work satisfaction during the lockdown. Major changes in the clinical practice were reported (Tables 3 and 4). Fifty-three % reported that the number of neurological in-patient beds was reduced. The majority (85%) agreed that fewer patients had been referred to the emergency ward with neurological conditions during the pandemic lockdown. Further, 80% agreed that fewer patients than usual had been admitted from the emergency ward to the neurological ward. Patient delay was common (72%). When patients with acute neurological disorders were admitted, most were treated as usual, but in-patient planned evaluations and treatments were postponed. Only 8% maintained a regular out-patient clinic with in-person appointments. Eighty-seven % reported a shift towards more telemedicine, with significantly more use of telephone than video consultations for both newly referred patients (54% vs. 30%, p<0.001) and follow-ups (99% vs. 50%, p<0.001). Only 30% agreed that the healthcare authorities in Norway collaborated to find good solutions for neurological patients, but 62% agreed that the academic community collaborated on these topics (Table 3). Twenty-four % agreed that the standard of care was reduced for patients with acute neurological conditions whereas 64% agreed that the standard of care was reduced for patients with chronic neurological conditions ( Table 3). Multiple sclerosis We have maintained the regular multiple sclerosis care 11 (18) Treatment of acute attacks is less available 12 (20) I have asked patients to stay away from the hospital because they are a vulnerable group 37 (63) We have chosen different types of treatment for newly diagnosed patients than we normally do 35 We have changed the current treatment for individual patients due to the pandemic 12 (22) Out-patient follow-up are primarily conducted by telephone consultation 54 (93) Movement disorders We have maintained the regular movement disorders care 15 (28) Out-patient follow-up are primarily conducted by telephone consultation 39 Out-patient follow-ups are postponed 39 I have patients whose advanced Parkinson's treatment has been postponed as a consequence of the pandemic 12 (28) Amyotrophic lateral sclerosis We have maintained regular ALS out-patient clinic 9 (60) We offer telephone consultations to ALS patients 13 I have recommended that ALS patients avoid hospital because they should not be exposed to potential Covid-19 infection at the hospital 13 (81) I have patients whose home respirator has been delayed 1 (7) I have patients who has been denied home respirator 0 (0) Patients with ALS have a decreased lifespan due to worse options from the health services during the pandemic 2 (13) Glioblastoma Out-patient follow-ups are postponed 7 (32) Out-patient follow-ups are postponed because these patients should not be exposed to potential Covid-19 infection at the hospital 6 (29) Treatment with temozolomide is given as planned 18 Diagnostic examination of glioblastoma patients takes longer time 0 (0) There is longer waiting time before surgery for newly diagnosed glioblastoma patients 0 (0) Radiation therapy are postponed 1 (7) We avoid admitting patients with glioblastoma even when their condition is worsening because they should not be exposed to potential Covid-19 infection at the hospital 1 (5) Patients with glioblastoma have a decreased lifespan due to worse options from the health services during the pandemic 1 (6) Immune-mediated polyneuropathies Patients receive immunological treatment with regular intervals 23 for dystonia and spasticity. Table 4 shows a more detailed description of the reported management of ten different neurological disorders during the initial phase of the lockdown. Overall, the acute management was mostly reported to be unchanged, while there were several changes to planned activities, in particular out-patient follow-ups which were postponed or conducted by teleconsultations. Regarding stroke patients, 57% of the respondents reported that out-patient clinic followups were postponed, and 64% reported that rehabilitation options were reduced. Acute treatment of seizures, and EEG in the acute phase, were not strongly impacted by the pandemic. Access to elective |
EEG and sleep-deprived EEG was reduced and follow-ups were mainly conducted by telephone for epilepsy patients. More than half (57%) reported that migraine patients were not treated by regular intervals with botulinum toxin A (BTX) injections. Almost half of the respondents (43%) were more likely to put patients on CGRP antibodies rather than BTX. However, BTX was offered at regular intervals to patients with dystonia or spasticity. Only 18% reported that regular multiple sclerosis care was maintained. More specifically, 65% reported that they had chosen different types of treatment for newly diagnosed patients than they normally would do. Twenty-two % of the neurologists had changed the current treatment scheme for individual patients due to the pandemic. Telephone consultations were the main form of follow-up, and 63% had asked patients to avoid the hospital because they considered patients with multiple sclerosis a vulnerable group to COVID-19. Management and care of movement disorders were changed in 72%, and 28% had patients whose advanced Parkinson's treatment was postponed. Sixty % maintained regular ALS out-patient clinics, but 81% had recommended the ALS patients to avoid the hospital in order to reduce the risk of being exposed to COVID-19 infection. No neurologists reported that patients with ALS were denied home respirator treatment due to the pandemic. Glioblastoma patients were treated as normal. Almost all reported that diagnostic examinations, treatment with temozolomide, surgery and radiation therapy were given as planned. Self-perceived sleep problems, depression and distress among neurologists Thirteen % (n = 17) of the neurologists reported that their sleep had suffered due to the pandemic and 15% (n = 19) had felt depressed for more than 14 out of the last 28 days. There were no significant differences between residents and senior consultants, or between those who assessed and did not assess potential COVID-19 infected patients with regards to suffering from sleep problems and feeling depressed. A high degree of burden and stress was associated with changed access to resources and that patients were not given the follow-up they should have received (Table 5). Furthermore, perceived stress was associated with the potential lack of PPE and the fear of spreading SARS CoV-2 to close family members. The danger of becoming infected and ill was not perceived as an important contributor to burden and stress, and almost 80% were satisfied with their work situation during the pandemic. Neurologists with self-perceived depressed thoughts reported significantly higher burden/ stress for the following statements: The fear of contracting SARS CoV-2 at work impacts negatively on my quality of life (p = 0.031), I am afraid that I will spread SARS CoV-2 to close family members (p = 0.001), Potential lack of personal protection equipment in my clinical work (p = 0.04), Patients are not given the follow-up they should receive (p = 0.032) and Changed work routines and access to resources (p = 0.001). Discussion There were major changes in clinical neurological practice during the first weeks of the pandemic lockdown in Norway. Almost three out of four neurologists experienced a change in their personal work situation, and the majority examined patients with suspected COVID-19 infection and neurological disease. The pandemic affected various aspects of their personal and professional life. A high degree of burden and stress was associated with changed access to resources and to the fact that patients were not given necessary follow-up. Neurologists were also worried about the potential lack of PPE and the fear of spreading SARS CoV-2 to close family members. Interestingly, the danger of becoming infected and ill was not reported as an important contributor to burden and stress, and almost 80% were satisfied with their work overall situation during the pandemic. Non-urgent follow-ups, such as out-patient clinic appointments were postponed or carried out as telephone consultations, and rehabilitation options were reduced. Delivering sub-optimal health care could impact on the short-term quality of life for patients with conditions such as headache and epilepsy, and on long-term outcome in other neurological conditions, e.g. stroke or multiple sclerosis. On the other hand, necessary urgent treatment of serious conditions, such as acute stroke, acute seizures, glioblastoma and immune-mediated polyneuropathies, was reported to be virtually unchanged during the first weeks of the pandemic. In Norway, as in Germany, a decrease in the absolute thrombolysis treatment numbers has been reported, but with similar treatment rates compared with previous months for those with ischemic stroke that presented to hospital within the thrombolysis time window [16][17][18]. The shift towards more teleconsultations is similar to what has been reported elsewhere [19][20][21]. Some neurological disorders may be better suited for telemedicine than others, and overall, this shift may be less problematic within non-emergency neurology than it is in other medical fields [22][23][24][25][26][27][28]. Many neurological patients have increased risk for severe COVID-19 due to their age, comorbidities or conditions that require immunosuppressive treatments. A large proportion of the neurologists in the present study had asked patients with multiple sclerosis, glioblastoma and ALS to stay home and postpone follow-ups in order to avoid being infected with COVID-19 at the hospital. For these patients, teleconsultations may provide a safe way to maintain follow-up care. Some academic and scientific communities were making international protocols and consensus statements rapidly available while others used more time. A recent consensus statement on good clinical practice regarding patients with neurological disease during the COVID-19 pandemic will hopefully provide guidance for neurologists and be an important contribution in the second wave of the pandemic [29]. The global community was clearly not sufficiently prepared for such a pandemic [30][31][32][33]. In most places it was local initiatives and heroic efforts from health care workers, more than governmental plans, that were responsible for the rapidly implemented changes needed to treat a high number of COVID-19 patients [3]. These fast re-organizations show that it is possible to implement fast-track changes if deemed necessary from a professional point of view. However, unplanned, fast and large scale changes may come with a cost for both patients and health care workers [10]. Only 30% of the respondents agreed that the health authorities in Norway collaborated to find good solutions for neurological patients, but 62% agreed that the academic community collaborated on these topics. This is in line with findings from other countries, where a lack of coordination and preparedness has been found and criticised [30,31,33]. It may be considered a paradox that 78% of the respondents reported that they were satisfied with the overall work situation in a period with much uncertainty and rapid shifts in availability of resources. One explanation may be that healthcare workers find it meaningful to contribute in the time of crisis, and may find the situation more satisfying than usual care despite relocations to other wards, downsized ward teams, less rest between shifts, longer shifts and more patients per shift. The Norwegian neurologists experienced stress from the potential lack of PPE and the fear of spreading SARS CoV-2 to patients or close family members. The Norwegian government expressed early a shortage in PPE and required health care workers to reuse PPE. Furthermore, health care workers were told to use PPE only in cases of a high suspicion of COVID-19 despite widespread community spread, uncertainty about air transmission and concerns about asymptomatic transmission [34,35]. A recent study among US neurologists found that during the initial weeks of the pandemic, most institutions had expressed shortages in PPE, with almost half (45%) of neurologists having to reuse their PPE [21]. Interestingly, the thought of becoming infected and ill was not associated with a substantial burden and stress among Norwegian neurologists, even though media reports from other countries suggested a high incidence, and even mortality for doctors on the frontline [36]. A recent meta-analysis found pooled prevalence of 23, 23 and 34% for anxiety, depression and insomnia among health care workers during the initial phase of the COVID-19 pandemic [10]. A study of frontline doctors in Pakistan reported a 43% prevalence of anxiety/depression in the initial weeks of the pandemic [37]. Further, a majority of the physicians dealing with suspected or confirmed COVID-19 cases in an Iraqi Kurdistan study reported high levels of insomnia and stress [38]. Two studies among health care workers in highly affected areas in Italy both reported high levels of burnout and psychological symptoms during the COVID-19 emergency, but with no significant differences between physicians and nurses [39,40]. We found lower rates, with 13% reporting sleep problems and 15% reporting depressive thoughts due to the pandemic, but as different questions and comparisons were used in previous reports, these may not be directly comparable. The reasons for such high psychological distress may be diverse, but a study of Canadian emergency physicians reported that the following factors had an impact on physician well-being during the initial phase of the pandemic: personal safety, academic and educational work, PPE, the workforce, patient volumes, work patterns, and work environment [41]. These reported factors are in line with our findings. Some of the stress related to an unknown and rapidly evolving situation is probably unavoidable. However, it may be hypothesized that an increased preparedness from the government could reduce the stress and burden placed on many health care workers during a pandemic with an unknown agent such as SARS-CoV-2. First of all, access to adequate PPE, but also realistic plans on how to strategically use the workforce including work schedules, academic and social support, and reasonable time to recover must be in place. Interestingly, we found no significant differences in sleep problems and depressive thoughts between residents and senior consultants, or between those who assessed and did not assess potential COVID-19 infected patients. It may be hypothesised that residents on the frontline assessing potential COVID-19 patients have a higher chance of experiencing quarantine leading to isolation, loss of social support and negative thoughts. In addition, younger residents are more inexperienced, their educational activities halted, and they received reduced supervision from more senior physicians. In total, this could have led to more distress among residents than senior consultants. However, our findings do not point to any of these associations. The combination of i) work related factors such as rapid and unpredictable changes in the work situation, the potential lack of PPE, the fear of spreading SARS CoV-2 to close family members, with ii) private life factors such as reduced contact with close family or caring for family members in distress, reduced social contact and suspension of personal travels may have led to work-home interface stress [15]. The mean score of work-home interface stress was 2.8 for the total sample with no significant differences between residents (2.9) and senior consultants (2.6). Neurologist with self-reported depressive thoughts or sleep problems had significantly higher work-home interface stress. The cross-sectional design in the present study does not permit any conclusions about causality or direction of the association between these outcomes. A study of Norwegian physicians from 2019 reported lower mean scores of 2.5 and 2.3, 4 and 15 years after graduation [42]. Based on this it is reasonable to assume that neurologists experienced slightly more work-home interface stress during the initial phase of the pandemic. However, this comparison should be made with caution as previous studies have shown that neurologists have a higher degree of stress, less job satisfaction and more burnouts than other physicians [11,12]. Thus, our findings may reflect this previously described difference between neurologists and other physicians. Respondents represented all neurological departments in Norway, which should ensure good representativeness. There are approximately 400 neurologists in Norway, of which 135 answered the questionnaire. However, a proportion of the non-responders will not have been in clinical work during the relevant period because of research or education terms, rotation to other wards or being in quarantine, on sick leave, or parental leave. Thus, we consider the 34% responder rate a conservative estimate. Given that all data were collected during three hectic and uncertain weeks when many changes were implemented for neurologists at the hospitals, we consider the responder rate satisfactory. Further, based on the age and gender diversity of the sample we do not suspect major biases in responder rate. Questionnaire studies such as this may introduce recall bias, however, we do not consider a reason to suspect systematic bias. The survey was sent out at the beginning of the pandemic in Norway when most of the hospitals made rapid changes in |
protocols and there was much uncertainty. Thus, the fact that the first wave of the pandemic ended up being better controlled in Norway than in many other countries may not have influenced the answers at this point, and we believe the results to some extent may be generalized to other countries with similar health care systems. As the world is facing the second wave of the pandemic, now with much more knowledge about COVID-19, it would be interesting to repeat the questionnaire to describe changes in the delivery of health care and in the impact on neurologists. Conclusion Three out of four neurologists in Norway experienced a change in their personal work situation during the first phase of the COVID-19-pandemic. Reduced standard of care was reported for all neurological conditions, particularly for non-emergency treatments. Changed access to resources, and the perception that medical follow-up was unsatisfactory, were associated with a high degree of burden and stress among neurologists. The fear of becoming infected and ill themselves was not reported as an important contributor to burden and stress. Supporting information S1 File. The complete questionnaire of NeuroPan. (PDF) Reducing Conditions Favor Magnetosome Production in Magnetospirillum magneticum AMB-1 Magnetotactic bacteria (MTB) are a heterogeneous group of Gram-negative prokaryotes, which all produce special magnetic organelles called magnetosomes. The magnetosome consists of a magnetic nanoparticle, either magnetite (Fe3O4) or greigite (Fe3S4), embedded in a membrane, which renders the systems colloidaly stable, a desirable property for biotechnological applications. Although these bacteria are able to regulate the formation of magnetosomes through a biologically-controlled mechanism, the environment in general and the physico–chemical conditions surrounding the cells in particular also influence biomineralization. This work thus aims at understanding how such external conditions, in particular the extracellular oxidation reduction potential, influence magnetite formation in the strain Magnetospirillum magneticum AMB-1. Controlled cultivation of the microorganisms was performed at different redox potential in a bioreactor and the formation of magnetosomes was assessed by microscopic and spectroscopic techniques. Our results show that the formation of magnetosomes is inhibited at the highest potential tested (0 mV), whereas biomineralization is facilitated under reduced conditions (-500 mV). This result improves the understanding of the biomineralization process in MTB and provides useful information in sight of a large scale production of magnetosomes for different applications. INTRODUCTION Magnetotactic bacteria (MTB) are a diverse group of prokaryotes able to synthesize magnetic nanoparticles in organelles called magnetosomes. Magnetosomes are composed of a magnetic core, nanocrystals of either magnetite or greigite, encapsulated in a lipid bilayer (Uebe and Schuler, 2016). The nanoparticles are organized in chains due to a protein-based backbone stabilizing the alignment and enabling the bacteria to align along the Earth's magnetic field lines Faivre, 2015). Living cells have found applications as drug carriers or as biohybrids thanks to their magnetic behavior (Taherkhani et al., 2014;Felfoul et al., 2016;Stanton et al., 2017). In addition, the unique characteristics of magnetosomes enable applications such as contrast agent in magnetic resonance imaging. Magnetosomes can also be functionalized with proteins or antibodies and used as biocatalysts in various chemical applications (Yoshino et al., 2009;Ginet et al., 2011;Taukulis et al., 2015). The chemical process controlling magnetite formation in magnetosomes is under study, specifically because size and morphology control as well as colloidal stability is difficult to obtain in purely chemical syntheses. The chemical stability of magnetite is restricted to a limited domain in the oxidationreduction potential (ORP)/pH space where magnetite formation typically takes place at ORP values from −200 to −400 mV and pH > 8 . Nevertheless, MTB are able to provide chemical conditions suitable for the production of the appropriate iron phase under different physiological conditions and therefore, to biomineralize magnetite under adverse circumstances. While there are still a lot of uncertainties concerning their physical and chemical conditions leading to biomineralization, the genetic background has been the subject of numerous studies highlighting the genetic control on the process (Uebe and Schuler, 2016). A genomic island, called Magnetosome Island (MAI) was identified containing the genes (mam and mms genes) responsible for crystal formation (Ullrich et al., 2005). The process can also be summarized as follows: (1) Iron is taken up both as Fe (II) and Fe (III) (Faivre et al., 2007); (2) The cytoplasmic membrane forms an invagination, which will then acts as the chemical reactor or the formation of magnetosomes (Komeili et al., 2006). (3) Iron is transported inside the membrane by the iron transport proteins such as MamM, MamB, FeoA and FeoB (Uebe and Schuler, 2016). (4) Iron is then partially reduced and precipitated as magnetite (Baumgartner et al., 2013;Siponen et al., 2013;Amor et al., 2018). In this process, some magnetosome-associated proteins, such as MamE, MamP, MamT, and MamX, take an active part in the control of the Fe 2+ to Fe 3+ ratio (Siponen et al., 2013;Jones et al., 2015;Barber-Zucker et al., 2016). In addition to iron, oxygen is also a particularly important redox-active player affecting bacterial growth and magetosome formation (Heyen and Schüler, 2003;Liu et al., 2010;Yang et al., 2013). Firstly, O 2 coming from the air was speculated to take an active part in the process, due to the competition between biomineralization process and respiration (Blakemore et al., 1985) before isotope analyses demonstrated that the oxygen involved in the magnetite biomineralization comes from water (Mandernack et al., 1999). Here, we use Magnetospirillum magneticum ABM-1, a facultative anaerobe capable of growing in the presence or in the absence of oxygen. However, different respiration pathways are being used while the final electron acceptor is the O 2 during aerobic growth. The general relationship between MTB and their extracellular environment have been explored aside from redox parameters. Typically, these experiments have investigated potential changes in the morphology of the cells grown in different physical and chemical conditions, as well as changes in the biomineralization process caused by extracellular disturbances. While differences in physical parameters such as temperature or magnetic field have shown only marginal and weak effects, changes in the extracellular chemical environment seem to have a larger influence on the formation of magnetosomes Moisescu et al., 2014). For example, change in the pH of the growth medium impacts the iron uptake, and resulted in altered morphologies of the produced crystals (Moisescu et al., 2014). The initial Fe availability causes change in crystal size distribution, morphology and aspect ratio . Oxygen inhibits the formation of magnetosomes at high O 2 partial pressures (Heyen and Schüler, 2003). Accordingly, smaller magnetosomes are formed such that the magnetic properties of magnetosomes and MTB indirectly dependent on dissolved oxygen concentration (Li and Pan, 2012). Oxygen concentration is ecologically relevant for MTB, since their natural habitat is found at the oxic-anoxic transition zone (OATZ), where chemical gradients are present (Lefèvre and Bazylinski, 2013). In natural waters, such as lakes and ponds, a chemical gradient is created in the sediments layer by the diffusion of oxygen from the water surface downward, and hydrogen sulfide produced by the sulfate reducing bacteria, which diffuses upward from the anaerobic zone. MTB are thought to have developed their complex magnetic apparatus in order to find the exact position in these chemical gradients (Lefèvre et al., 2014). Both the observed effect of oxygen on biomineralization and the ecological behavior of MTB suggest a strong connection between ORP and magnetosome formation. Oxidation -reduction potential is an indicator of electrochemical activity of substances in environmental conditions. ORP is defined as a measure of the intensity of its oxidizing and reducing properties. So far, the effect of ORP was only indirectly assessed by changing the partial pressure of oxygen during cultivation (Heyen and Schüler, 2003). Decorrelating redox conditions from oxygen concentration is yet to be done. Here, we thus conducted a study where the cells were grown in a bioreactor, where pH, temperature, ORP and pO 2 were continuously monitored and kept stable trough feedback systems. We used titanium (III) citrate as reducing agent and potassium ferricyanide as oxidizer. Titanium (III) citrate was already utilized as a reductive compound for microorganism growth (Jones and Pickard, 1980). Potassium ferricyanide, in turn, was also used in previous redox control of Escherichia coli anaerobic fermentative growth (Bagramyan et al., 2000). In order to check the influence of ORP on magnetosome formation, we investigated the anaerobic growth of Magnetospirillum magneticum strain AMB-1 (Matsunaga et al., 1991) at neutral pH (7), and varied the ORP from reduced (ORP = -500 mV) to neutral (ORP = 0 mV) compared to a control experiment where ORP was not regulated but continuously measured. MATERIALS AND METHODS Bacterial Strains, Flask Growth, Cell Density Measurements, and Toxicity Tests M. magneticum AMB-1 strain was purchased from the American Collection of Cell Cultures (No. 70264). The bacteria were grown in a modified flask standard medium (FSM) containing (per liter deionized water) 0.1 g KH 2 PO 4 (Acros organics, Belgium), 0.15 g MgSO 4 × 7 H 2 O, 2.38 g Hepes, 0.34 g NaNO 3 (Sigma-Aldrich, United States), 0.1 g yeast extract, 3 g soya peptone, and 100 µM iron (III) citrate (Alfa Aesar, Germany). If not otherwise specified, all chemicals were purchased from Carl Roth (Karlsruhe, Germany). The medium contained 27 mM sodium pyruvate (Alfa Aesar, Germany), modified trace elements, and an iron-depleted mineral solution (Widdel and Back, 1992). Afterward, the flasks were sealed with open ended caps and rubber stoppers and autoclaved. Inoculation was done with 10% (v/v) subculture and incubated under gentle shaking (100 rpm) at 28 • C for 24-48 h. All of the experiments were performed with axenic conditions, with regular microscopic controls. A UVspectrophotometer (Nanophotometer, IMPLEN, Germany) was used to determine the cell growth by measuring the optical density at 565 nm (OD 565 ). The optical density was measured with two bar magnets (Supermagnete, Germany) placed parallel (OD min ) and perpendicular to the light beam (OD max ). The coefficient of magnetically induced differential light scattering (C mag ), describing the magnetic orientation of the cells was calculated following a published procedure (Schüler et al., 1995). In order to correlate OD 565 values with number of cells, cell-counting experiments were performed using a Neubauer Chamber (Brand, Germany) with an optical microscope (AXIOmicroscope, Zeiss, Germany) and anaerobic cultures. The cell density was measured to be 9.07 × 10 7 cells mL −1 , at an OD 565 of 0.277. A linear correlation of cell count with the optical density was assumed, as done in other studies on Magnetospirillum (Heyen and Schüler, 2003) such that cell density of 3.27 × 10 8 cells × mL −1 corresponded to an OD 565 of 1. This cell density value was used for further normalization of iron content in cells and it is similar to cell density reported earlier (Heyen and Schüler, 2003). The results were compared to the cell dry mass using the value provided in the literature (Heyen and Schüler, 2003), which reported a culture density at OD 565 = 1, a dry weight of 0.28 g. We used titanium (III) citrate as reducing agent and potassium ferricyanide as oxidizer. For Titanium (III) citrate, no major toxicity was reported for an anaerobic bacterium such as Clostridium formicoaceticum or for a facultative anaerobic bacterium such as Pseudomonas denitrificans (Zehnder and Wuhrmann, 1976). As toxicity tests, the concentration of titanium (III) citrate was tested on M. magneticum AMB-1 cells in flasks with concentrations varying from 0.22 to 15 mM. Potential toxic effects of potassium ferricyanide were also measured at different concentrations (0.1 mM < [C 6 N 6 FeK 3 ] < 3.6 mM). The experiments were conducted in 250 ml flasks filled FSM and treated as described above. To quantify the cell growth, OD 565 was measured directly after the inoculation, and after 24 h. Growth in the Bioreactor and ORP Determination Before testing the ORP conditions, non-magnetic precultures were prepared as previously described (Faivre et al., 2007), with the following modifications: from a fully grown microaerobic FSM iron-rich culture, 3% (v/v) was inoculated into 100 mL of fresh iron-depleted culture medium. The entire culture was then used as subsequent preculture for flasks containing 1 L of iron-depleted fresh medium. The final, 1 L subculture was used for the experiments performed in the bioreactor. The age of the preculture was tested in order to better understand the following cellular growth in the bioreactor. Experiments did not show differences in |
growth between 1, 2, and 3 days-old inoculum (see Supplementary Figure S1). Forty eight hours old precultures were consequently chosen as inoculum for further experiments in the bioreactor. All of the precultures and taken samples were controlled for any potential contaminants by regular checks, using the optical microscope. Moreover the C mag was measured to distinguish between magnetic and non-magnetic cells. The bacteria were grown in the bioreactor with a volume of 16 L (L1523, Bioengineering, Switzerland). The bioreactor was able to keep constant the following parameters: temperature, oxygen concentration, pH and the steering speed (Supplementary Figure S3). The bioreactor was filled with 12 L of FSM medium, omitting Hepes; sodium pyruvate was substituted by aqueous solution of 0.3% (v/v) of L-(+)-lactic acid (Sigma-Aldrich, United States). The pH of the culture was controlled by the addition of either 1 M NaOH or 0.5 M H 2 SO 4 . The ORP control was obtained by the addition of either 0.8 M of titanium (III) citrate or 0.3% (w/v) potassium ferricyanide solution. The oxygen concentration was continuously measured using an oxygen probe (Bioengineering, Switzerland). After sterilization of the filled reservoir (121 • C for 35 min), the syringe providing nitrogen source was plugged in during the cooling process to sparge nitrogen through the medium and obtain oxygen-free conditions. After the desired temperature was reached, the pH and the ORP values were adjusted. As such, the bioreactor was prepared to be inoculated and to perform the experiments. All experiments were done in triplicates and error bars are reported as standard error of the mean. Samples were taken after 4, 14, 16, 18 20, 22, and 24 h after the beginning of the growth in the bioreactor. Sampling was done in the proximity of a flame utilizing a sterile syringe. 20 mL of culture was removed each time, from which the OD 565 was first measured. Sixteen microliter cultures were then centrifuged at 8000 rpm for 10 min at 4 • C (Avanti J-E centrifuge, Beckman Coulter, United States), pellets and supernatant were divided and stored at −20 • C until further analysis. The remaining 4 mL were used for analyses using electron and optical microscopies. The ORP value of the suspension was measured using platinum redox electrode (Pt4805-DPAS-SC-K8S METTLER TOLEDO). Iron Concentration Measurements Iron measurements were made using an inductively-coupled plasma optical emission spectrometer (ICP OES, Optima 8000, PerkinElmer Inc., United States). For iron analysis, the samples containing bacterial pellets were dissolved in 500 µL of an aqua regia solution [65% w/v HNO 3 and 35% w/v HCl, in a ration 1:2 (v/v)]. The samples were incubated overnight in glass vials at 40 • C in order to dissolve the iron species. Optical Microscopy The cells length was measured at time 0 and at time 24 h for each experiment. Images were taken using an optical microscope (Imager A2 AXIO-microscope, Carl Zeiss, Germany) at 39.69 X magnification. The length of 100 cells per condition was then measured using the software "Image J" (National Institute of Health, United States). A straight line was drawn from the one end to the other end of each cell, and this line was then measured (Supplementary Figure S2A). TEM Analysis For electron microscopy, M. magneticum cells were concentrated by centrifugation (5000 rpm, 5 min, 4 • C), adsorbed onto a cooper grid, rinsed once with buffer (10 mM Hepes, 5 mM EDTA) and once with distilled water. The samples were imaged using a Zeiss 912 Omega transmission electron microscope at 120 kV. The transmission electron microscopy images were analyzed with "ImageJ" in order to determine the size of magnetosomes, their amount per cell and the cell length ( Supplementary Figures S2B-F). The magnetosomes dimensions were estimated by calculating the best fit of ellipse to the contours (Devouard et al., 1998). Magnetosome Size Analysis Using X-Ray Diffraction To complement the magnetosomes size evaluation by TEM, X-ray diffraction (XRD) analyses were performed. XRD measurements were performed at the µ-Spot beam line of the BESSY II synchrotron facility of the Helmholtz Center for Energy in Berlin, Germany (Paris et al., 2007). The energy was set up to 15 keV (λ = 0.82656 Å) with a Si 111 monochromator. For measurements, a thin Kapton foil (Breitlander GmbH, Hamm, Germany, Cat. No. CH-440) was clamped on a special custom-made sample holder (Schmitz, 2010;Fischer et al., 2011). Ten microliter of each sample solution were pipetted on the thin Kapton film. The samples were mixed with the internal α-quartz standard (Standard Reference Material, 1878, NIST) prior to drying. After drying on the Kapton foil, the samples were measured with a 100 µm beam. The diffraction data were acquired on a MarMosaic 225 detector (Mar USA, Evanston, United States), consisting of nine independent CCD cameras. The detector resolution was 3072 × 3072 pixels, whereas the pixel size was 73.242 µm. For monitoring the XRD, the Spec program (SPEC, 2010) was used. Exposure times varied between 5 and 300 s, depending on the intensity of the signal. Each sample was measured three times. The sample to detector distance was set to 150-170 mm. The software DPDAK was used to fit the peaks with a pseudo-Voigt function. The internal α-quartz standard was used to calculate the exact sample-to-detector distance and the instrumental peak broadening. The crystallite size of the magnetosomes was then calculated from the corrected full width at half maximum (FWHM corr ) of the (311)-peak of magnetite (Benecke et al., 2014) based on the Scherrer equation (Equation 1) (Widdrat et al., 2017). (1) Toxicity Tests Toxicity test were necessary in order to see if the ORP-setting chemicals influence the growth of AMB-1 cells. At this first stage we did not concentrate on the effect of ORP on magnetosome formation, but on the maximal concentration of the chemical that can be used without inhibiting cellular growth. We observed that both titanium (III) citrate and potassium ferricyanide inhibit but do not stop cell growth at high concentration (Figures 1, 2). After the addition of 0.22 mM titanium (III) citrate, we observed a reduction of the growth rate of 17% after 24 h, in comparison to the untreated control. At the concentration of 15 mM of titanium (III) citrate, the growth rate decreased by 51%. In contrast, for the oxidizer, no major effects were measured with the addition of 0.1 mM potassium ferricyanide after 24 h whereas; whereas a 60% decrease was recorded at the concentration of 3.6 mM. The cells grown either with potassium ferricyanide or titanium (III) citrate did not exhibit any change in the external cellular morphology. In both cases, the cells of M. magneticum exhibited FIGURE 1 | Toxicity test for a range of Titanium (III) citrate concentrations used during M. magneticum cultivation. The AMB-1 cultures were grown in 500 mL flasks bubbled with N 2 in order to sustain anaerobic conditions. Ten milliliter of the preculture was used as an inoculum for each flask. Cultures without Titanium (III) citrate served a control. Optical density (OD 565 ) was measured for each concentration immediately after the inoculation (t0, brown bars) and after 24 h of cell cultivation (t24, gray bars). Error bars represent the standard deviation of the three repetitions. a short spiral shape, typical for this strain (see Figure 3) and they all actively swim under the microscope. Effects on magnetosome formations were not analyzed during these experiments. During experiments performed in the bioreactor, titanium (III) citrate and potassium ferricyanide were added continuously via titration systems. The final culture volume was 13 L. During the 42 h cultivation, around 200 mL of either titanium (III) citrate or potassium ferricyanide was necessary to keep ORP either at either −500 or at 0 mV. According to that, the maximum and final concentration set at 0.45 mM of titanium (III) citrate and 0.1 mM of potassium ferricyanide. Comparing these values with our tests, we can therefore assume that the concentration of ORPsetting chemicals will not affect bacterial growth in a large extent. Control Experiments Without ORP Adjustment The ORP was controlled by a feedback system of the bioreactor. The ORP change in the growth medium was compared between the medium without cells and with the inoculated medium, the latter being further cultivated for 42 h. The reported ORP values of control experiment in the medium with cells (shown in Figure 4, dark blue line) were first decreasing from -250 to -300 mV in about 1 h, settling for about 15 h before increasing back to -250 mV over time. Below, we will call this condition -250 mV even if the ORP is not controlled by the feedback loop of the bioreactor. The measurement of the ORP value in the medium without cells settled around -300 mV (Figure 4, green points). Additionally, it was observed that the ORP in the blank medium was probably influenced by the daily light cycle with an increase of around 20 mV for daylight conditions with respect to darkness (i.e., night) indicating potential photochemical reactions taking place in the medium. For our purposes, these changes were considered, when compared to the difference in ORP at which the experiments were performed and thus not further considered: the variation of ORP observed due to light cycle (±20 mV) is around 10 times lower than the experimental controlled ORP variation (±250 mV). Studying the effect of photoactive compounds on the growth of the bacteria and/or studying phototrophic properties of MTB will be of interest in the future. Cell Properties After running the control experiments, we assessed different experiments run at -500 mV, control experiments (around -250 mV as shown above), and 0 mV. Starting with similar cell density, the final OD 565 was higher for ORP -500 mV in comparison to 0 mV (Figure 5). Results of the cell length measurements (Table 1) show an abnormal cell elongation for the two tested conditions (0 and 500 mV) compared to regular lengths showed by the control experiments with unclear origin. It is speculated that bacteria become longer in conditions of stress as shown in the literature (Young, 2006). The pellets of cells were collected by centrifugation and the iron content was measured by ICP-OES in order to understand how much iron was accumulated in the cells during the reaction. The results of the iron content in pellets indicated that the iron concentration in the cells (mg g −1 dry weight) increased during bacterial growth. The highest intracellular iron accumulation after 42 h was reported at ORP −500 mV, while the lowest at 0 mV, which confirms the hypothesis that AMB-1 strain FIGURE 4 | ORP continuous measurements of two controls. One was performed with cells (ORP -dark blue) while the second one was done without cells (ORP -green points). Growth of cells is shown as optical density measurement at the wavelength 565 nm (OD 565 ). OD 565 is shown as red triangles connected with dashed red line as a guide for the eye. forms magnetosomes preferably under reduced conditions (Yang et al., 2001). The differences obtained for optical density, cell magnetism and iron content per dry weight under reduced and oxidized conditions are summarized in Table 1. Magnetosome Properties With TEM imaging, it was possible to count the number of magnetosomes per cell and to measure the size of the magnetosomes. The results are expressed as number of magnetosome per µm of cell to avoid bias due to differences in the cells length as reported above. For particle size measurements, the major and minor axes of the crystals were analyzed and the surface of the ellipse was calculated. A clear trend in magnetosomes amount and size is visible among tested conditions; at ORP −500 mV the highest magnetosome size and number was observed, at ORP −250 mV magnetosomes exhibited similar values, while at ORP 0 mV the lowest size of particles and lowest amount of magnetosomes was observed. The results are summarized in Table 2. In Figure 6, representative TEM images of magnetosomes chains are shown. Briefly, the magnetosome sizes we obtained at ORP −500 and −250 mV are 50.8 ± 13.6 and 46.7 ± 15.7 nm respectively, in agreement with a previous study (Li and Pan, 2012). However, it is important to mention that many small particles might have been excluded from the measurements due to the filtering the particle area at |
225 nm 2 particle area and at a roundness of 0.7. This issue had a strong influence, especially at ORP 0 mV, where the average Measurements of the number of magnetosomes, obtained via analysis of transmission electron microscope images. size of magnetosomes was 34.5 ± 12.3 nm and where in particular many magnetosomes were underdeveloped and certainly have not reached mature size. The mean magnetosome size was also analyzed with XRD. The trend in the decrease of particles size obtained by both methods is similar for ORP 0 mV. The diameter of nanoparticles measured via X-ray is 31.5 ± 1.3 nm while size of nanoparticles measured from TEM images is 34.5 ± 12.3 nm. The size of nanoparticles synthesized at −250 mV (46.7 ± 15.7 nm) and at ORP −500 mV (50.8 ± 13.6 nm) showed higher values than the sizes obtained via XRD; ORP −500 mV: 37.2 ± 0.6 nm; ORP −250 mV: 37.2 ± 0.8 nm. This confirmed the hypothesis that ORP 0 mV caused a decrease in the size of magnetosomes. The comparison between these two techniques is shown in Table 2. We performed an analysis of variance (ANOVA) on the magnetosome sizes. The tests yielded a significant variation among conditions [F(2; 1197) = 72.84, p < 2e-16]. A post hoc Tukey test showed that all the groups differ significantly at p < 0.001. DISCUSSION In this work, the formation of magnetosomes was monitored during cultivation in a bioreactor at different ORP values. ORP was varied between −500 and 0 mV. Previous studies only analyzed the effect of oxygen partial pressure as an oxidizer, where variations of ORP values were the consequence of different oxygen concentrations (Li and Pan, 2012;Moisescu et al., 2014), making this study the first one to disentangle ORP and oxygen concentration. Since the experiments were performed under strictly controlled anaerobic conditions, it was possible to exclude oxygen coming from the air and use other chemical agents to modify the redox state of the medium. The redox control during the cultivation of AMB-1 cells under reduced and oxidized conditions was successfully achieved by using either titanium (III) citrate or potassium ferricyanide. Effect of Redox-Active Chemicals on Cells Morphology and Growth In order to disentangle the effect of ORP from the effect of redox-active chemicals (titanium (III) citrate and potassium ferricyanide), we performed preliminary experiments and toxicity tests, which showed that the concentration of these chemicals, at which we later use them, is not dramatically affecting the cells. Even at the highest, tested concentrations, 15 and 3.6 mM for titanium (III) citrate and potassium ferricyanide respectively, cells remained motile. They exhibited slower growth but it was still observable. Other studies done by Li and Pen showed that under anaerobic conditions and after 40 h growth, the cell amount calculated as cells OD was much lower (OD 600 ca. 0.05), when compared to cells grown in our bioreactor for 42 h (Li and Pan, 2012). Even aerobic conditions, which usually induce cell growth, reached lower OD values (OD 600 < 0.2) than cell grown in our system at all used conditions; at 0 mV, OD 565 = 0.238; at −250 mV, OD 565 = 0.292, at −500 mV, OD 565 = 0.324). More research will thus be necessary to clarify the origin of the recorded cell elongation. Effect of ORP on Iron Accumulation Iron accumulation measurements showed an inhibition of the biomineralization process at ORP 0 mV, while it was enhanced at −500 mV. Possible reasons for such behavior could be connected to the chemical behavior of iron in aquatic solutions. Magnetite synthesis requires a 2:1 Fe(III):Fe(II) ratio. At lower levels of ORP (the −250 and −500 mV conditions), the synthesis of magnetite (Fe 3 O 4 ) at pH around neutrality is thermodynamically favored. At higher ORP levels, other iron oxide phases like Fe 2 O 3 are advantaged (Bell et al., 1987). The values of iron uptake appeared to be in the same range as previous studies on M. magneticum AMB-1, when compared on the laboratory scale. Heyen and Schüler showed that in large 4 L scale fermenter the iron content, after AMB -1 growth, reached 9.8 mg × g −1 dry mass. Our experiment at four time bigger scale (16L) showed that the amount of iron was much higher, reaching 12.8 mg × g −1 dry mass. This increase showed that AMB-1 strain is even better candidate for the production of magnetosomes than already, widely used MSR -1, on the technical scale, under reduced fermentation conditions. biomineralization inhibited at high oxygen concentrations, while restored with the addition of a reducing agent. In Table 3, a comparison between data obtained in this study with results obtained by Heyen and Schüler (2003) is shown. It exemplified that at lower oxygen concentration (0.25 mbar partial pressure); an iron accumulation is observed, which origin remains unclear. Once the oxygen partial pressure increased, iron accumulation was reduced with values similar to those in our experiment. Interesting would be to know which ORP values correspond to the values expressed by Heyen and Schüler (2003), in order to have fully comparable results. A logical extension of the here presented study would be the investigation of the mechanism behind the observed phenomenon. Although M. magneticum has been shown to possibly take up both Fe (II) and Fe (III) (Faivre et al., 2007), the first is assume to be assimilated by passive absorption, while Fe (III) are actively taken up (Schüler and Baeuerlein, 1996;Amor et al., 2018). As a result, we can speculate that the effect of ORP could be to modify the ratio between the two oxidation states, thereby influencing the action of the MTB redox proteins, resulting in a slower and/or more energetically-costly biomineralization process. Effect of ORP on Size and Amount of Magnetosomes Li and Pan (2012) showed that the average magnetosome size in AMB-1 cells grown under anaerobic conditions was 41.5 ± 15.0 nm with the average shape factor, determined from TEM images, of 0.78. Comparing our results of image analysis of magnetosomes, those done at 0 and −250 mV appear to be in line with those measured by Li and Pan. But magnetosomes obtained at the ORP -500 mV are 10 nm bigger. The amount of magnetosomes obtained per cell is also higher when compared to studies of Yang et al. (2013) or Moisescu et al. (2011). The average amount of particles per cell reached 28 and 39 magnetosomes per cell, respectively. In our case, the amount of particles was calculated per µm due to enourmous cells elongation and it reached 6 magnetosomes. When calculated per cell, we reached a number of 94 magnetosomes per cell at -500 mV. In TEM size analysis, only a few particles were measured, whereas in the x-ray data analysis bulk dimension was determined, which gives a better overview about the actual size distribution in the whole sample. TEM observations are used as a double proof of the size and can additionally give raise of the magnetosome shape and the structure. Other Parameters Related to ORP and Their Potential Effects Beside the cell growth, magnetosome formation and iron dynamics, possible effects could potentially arise from the chemicals used too. Moreover, citric acid in bacteria has been reported to act as an exogenous siderophore (Silva et al., 2009) and M. magneticum have been reported to possibly secrete siderophore 3,4-dihydroxybenzoic acid for Fe (III) complexation (Calugay et al., 2004). The ability to bind ferric iron leads to the formation of citrate correlated siderophores such as rhizoferrin and staphyloferrin A, where the structure of the chelating unit preserves the citric acid structure (Johnstone and Nolan, 2015). This molecular mechanism would cause higher iron accumulation on the surface of the cells and may also be partly responsible for the enhancement of magnetosome formation at −500 mV. With the addition of potassium ferricyanide, the magnetosomes formation was disrupted at 0 mV. A previous study of Bagramyan et al. (2000) showed how a change in ORP in the growth medium of bacteria, specifically E. coli, influences the membrane transport mechanisms, in particular H + and K + fluxes trough the membrane. We hypothesize, that the oxidative stress might also reduce the iron flux across the membranes. It should be also taken into a consideration that the ORP of the medium is the outcome of a chemical mixture. That is why it is important to remember that there might me other reducing species affecting the ORP and in consequence influencing the ions flux. For example organic species, such as phenols or other metal ions could influence intracellular states, especially as environmental pollutants (Enache and Oliveira-Brett, 2011;Liu et al., 2013). It was shown that AMB-1 can mineralize tellurium nanorods separately from its magnetite crystals as a new method for recovering this rare element from the environment (Tanaka et al., 2010). Other organisms use biomineralization to detoxify pollutants, such as Rhodanobacter A2-61 that forms intracellular uranium -phosphate complexes (Sousa et al., 2013). SUMMARY AND CONCLUSION In summary, the results here highlight the differences between the bacteria grown at different ORP. These results are useful for the elucidation of the relationship between physico-chemical conditions of the growing medium and the biomineralization processes within the bacteria, showing that it is a variation in the external ORP, not specifically of oxygen, which ultimately is responsible for the observed effect on biomineralization. Finally, we show that ORP monitoring and tuning is of interest for the production of magnetosome in high yields. This is a prerequisite for the application of these biological particles with high added value in nano-and biotechnologies such as cancer treatment and hyperthermia (Alphandéry et al., 2011), for MPI research (Kraupner et al., 2017) or as contrast agent in magnetic resonance imaging (Taukulis et al., 2015). However, all these applications require several mg of magnetosmes. With current yields obtained from bioreactors (about 1 mg per L of culture) and the necessary amount requested for the above cited application (tens of mg of magnetosome per patient), there is an urgent need for the development of biotechnological processes leading toward the high yield production of magnetosomes. We show here that a high amount of particle is obtained at low ORP. Therefore; coupling current processes where high cellular yield can be obtained first, with our approach showing that low ORP are favorable for high yield production of magnetosomes, we expect that the development of magnetosome based applications will be simplified. AUTHOR CONTRIBUTIONS AO-W and DF conceived the project. GS, AO-W, and VR conducted the experiments. AO-W and GS contributed equally to this work. DF supervised the work. All authors discussed the results of the experiments and contributed to the final manuscript that was initially drafted by AO-W and GS. All authors approved the submitted version. FUNDING This work was supported by the Max Planck Society. Fourteen polymorphic microsatellite markers for the widespread Labrador tea (Rhododendron groenlandicum) Premise Microsatellite markers were developed for Labrador tea (Rhododendron groenlandicum, Ericaceae) to facilitate downstream genetic investigation of this species and the extremely closely related, circumboreal Rhododendron subsect. Ledum. Methods and Results Forty‐eight primer pairs were designed using Illumina data and screened for excellent amplification. Sixteen successful pairs were developed as microsatellite markers using fluorescently labeled amplification to generate chromatogram data. These data were evaluated for intrapopulation and interpopulation variability in three populations from Alaska and Maine, USA, and the Northwest Territories, Canada. Fourteen polymorphic markers genotyped reliably, each with one to eight alleles. Cluster analysis indicates that across the range, populations can be easily discriminated. Cross‐amplification in other Rhododendron subsect. Ledum species shows broad application of the developed markers within this small, well‐supported clade. Conclusions These microsatellite markers exhibit significant variability and will be useful in population genetics within R. groenlandicum and for investigation of species boundaries across Rhododendron subsect. Ledum. Rhododendron groenlandicum (Oeder) Kron & Judd (Labrador tea) is one of eight named species within Rhododendron subsect. Ledum (L.) Kron & Judd (Ericaceae). Rhododendron groenlandicum is widespread across northern North America in damp habitats such as bogs and rocky alpine slopes. Although the related species commonly known as Labrador tea were long considered closely related to Rhododendron, Kron and Judd (1990) first demonstrated, using morphological cladistic analyses, that these species should not be maintained as the separate genus Ledum, but included within Rhododendron. Hart et al. (2017) |
confirmed the monophyly of subsect. Ledum in a molecular phylogenetic study. However, this study also demonstrated clear conflict between the nuclear and chloroplast genomes, suggesting likely recent hybridization involving multiple species within this lineage. Indeed, the named species in subsect. Ledum have a complex nomenclatural history that mirrors this reticulate evolutionary history, with little consensus about what taxa should be recognized. Therefore, the evolutionary history of this lineage remains unclear, particularly at the population scale. Löve and Löve (1982) reported a sporophytic chromosome count of 2n = 26 for R. groenlandicum; however, recent flow cytometry data (K. T. Theqvist, unpublished) suggests that at least some populations may be tetraploid. A close relative, R. tomentosum Harmaja, was reported by Lantai and Kihlman (1995) to have populations of mixed ploidy (2n = 26, 52). Therefore, the possibility of tetraploid R. groenlandicum populations is reasonable. Currently, there are no microsatellite markers available for use in any member of Rhododendron subsect. Ledum. The absence of rapidly evolving markers for this lineage limits our ability to investigate boundaries among these recently diverged and likely reticulate species. Because of the young age of this lineage and the high likelihood of hybridization, it is appropriate to investigate relationships among species at the population level by documenting population-level ploidy, zones of hybridization, and genetic diversity alongside phylogenetic investigation. Development of microsatellite markers for R. groenlandicum, the most widespread species within subsect. Ledum, will likely provide novel tools for use across this entire closely related lineage. METHODS AND RESULTS All bioinformatics aspects of this project followed Gillespie et al. (2017). DNA from one R. groenlandicum individual (Appendix 1) was extracted following a modified cetyltrimethylammonium bromide (CTAB) approach (Doyle and Doyle, 1987) followed by CsCl 2 purification (Palmer, 1986). A microsatellite sequencing library using the MiSeq v2 protocol and 2 × 250-bp paired-end sequencing was performed on an Illumina MiSeq at Cornell Life Sciences Sequencing and Genotyping Facility (Ithaca, New York, USA). Out of 3,882,418 raw sequence reads (GenBank Sequence Read Archive no. PRJNA577479) that were trimmed of vector and low-quality sequence using the BBduk 1.0 plugin within Geneious 11.1.5 (Kearse et al., 2012;Biomatters Ltd., Auckland, New Zealand), 605,089 reads included microsatellite regions. Of this subset of reads, 16,420 permitted design of unique primers using MSATCOMMANDER (Faircloth, 2008) with mostly default settings, but mononucleotide motifs were excluded, primer length was 20-22 bp, and primer GC maximum content was 50%. A PIG-tail sequence (Brownstein et al., 1996) was added to reverse primers for stability. Details of both amplification and polymorphism screens followed Kasireddy et al. (2018). DNA from seven silica-preserved R. groenlandicum individuals (Appendix 1) was extracted using a QIAGEN Plant Mini Kit (QIAGEN, Hilden, Germany) modified for use with herbarium material (Drábková et al., 2002). These seven DNAs were used to screen 48 markers representing diverse motifs and repeat numbers via PCR amplification (1× GoTaq Flexi Buffer, 2.5 mM MgCl 2 , 800 μM dNTPs, 0.5 μM of each primer, 0.5 units GoTaq Flexi DNA Polymerase [Promega Corporation, Madison, Wisconsin, USA], and ~20 ng DNA, in a 10-μL reaction). Touchdown PCR (94°C for 5 min; followed by 13 cycles of 45 s at 94°C, 2 min at touchdown temperature [68-55°C], and 1 min at 72°C; followed by 24 cycles of 45 s at 94°C, 1 min at 55°C, and 1 min at 72°C; and followed by 5 min at 72°C) was employed. After the amplification screen, 16 primer pairs ( Table 1) that amplified exactly one distinct amplicon were genotyped at the Georgia Genomics and Bioinformatics Core (University of Georgia, Athens, Georgia, USA) and scored for polymorphisms using DNA of 68 well-spaced individuals from three populations representing the broad range of R. groenlandicum (Sitka, Alaska, USA; Northwest Territories, Canada; and Washington County, Maine, USA). For PCR reactions used to genotype individuals, 50% of forward primer was replaced with fluorescently tagged (6-FAM, VIC, NED, or PET; Life Technologies, Grand Island, New York, USA) M13 universal primers. Resulting chromatograms were manually scored using Geneious 11.1.5. We employed strict criteria for calling peaks. First, a peak was called only if the relative fluorescence unit (RFU) was ≥3000 and exhibited little background noise relative to signal. Additionally, a second peak (i.e., a heterozygote) was called only if the secondary peak's RFU was ≥90% of the first peak. Consequently, our measurements of genetic diversity are conservative. Descriptive statistics, including Hardy-Weinberg equilibrium (HWE) deviations, multilocus matches analysis (MMA) and principal coordinate analysis (PCoA) (Orloci, 1978), were calculated using GenAlEx version 6.503 Smouse, 2006, 2012). Two markers, RGROE019 and RGROE042, did not genotype consistently and were not developed further. Although some past studies have allowed the possibility that R. groenlandicum is polyploid, 14 loci revealed chromatograms with one to two peaks per individual. Our scoring of peaks is conservative in terms of genetic diversity, and therefore may underscore alleles associated with dosage differences. Although there was very little background noise/stutter in our data set, failure to detect polyploidy using this methodology is acknowledged. Overall, however, we conclude that individuals sampled here are diploid. Fourteen polymorphic loci exhibited one to eight alleles per population (mean 2.81) ( Table 2). No more than two peaks per individual were observed. Observed heterozygosity ranged from 0.000-0.636 (mean 0.125). HWE expectations were not met for 11 loci (78.6%) in at least one population including RGROE045, which violated HWE assumptions in all three sampled populations. The 14 polymorphic loci easily differentiated the populations, demonstrated by genetic distance followed by PCoA (not shown). The first three axes of the PCoA explained 52.61% of the variation and showed a clear division between the Sitka, Alaska, USA, population and the other two populations, which were moderately differentiated. The MMA of the 14 polymorphic loci revealed two sets of identical individuals within the Sitka population, suggesting limited clonality. The MMA and PCoA results together suggest considerable population structure within R. groenlandicum. The 14 developed markers were cross-amplified within a phylogenetic context following Hart et al. (2017). This included 12 individuals from Rhododendron subsect. Ledum (five R. columbianum (Piper) Harmaja, three R. tomentosum, and one each of R. diversipilosum (Nakai) Harmaja, R. hypoleucum (Kom.) Harmaja, R. palustre (L.) Kron & Judd, and R. tolmachevii (Tolm.) Harmaja). Amplification of all developed markers (Table 3) was successful in all species Asterisks (*) indicate statistically significant deviation from HWE (*P < 0.05; **P < 0.01; ***P < 0.001). M = monomorphic marker; NS = not statistically significant. CONCLUSIONS These newly developed microsatellite markers represent the first such tool for use in Labrador tea and its close relatives. The markers will allow population-level investigation within R. groenlandicum but are likely to also aid in clarifying the evolutionary history of Rhododendron subsect. Ledum, including investigation of species boundaries and putative hybridization events. The markers presented here are collectively able to demonstrate considerable genetic structure in just three populations of R. groenlandicum and genotype well in all sampled species within Rhododendron subsect. Ledum, likely because of inter-species similarity resulting from recent and ongoing divergence of these species. ACKNOWLEDGMENTS The authors acknowledge Ms. Gail Beaulieu and Ms. Suzanne Carriere (Government of Northwest Territories) for collections from Northwest Territories, Canada. Startup funding to E.L.G. was provided by Butler University and Marshall University. We acknowledge iNaturalist (www.iNatu ralist.org) for publicly available observation data that were critical in identifying localities and collaborators to accomplish field collections for this widespread species. AUTHOR CONTRIBUTIONS K.L.L. and E.M. conducted all fieldwork (but see Acknowledgments). E.L.G. carried out all bioinformatics and project design aspects and analyzed the data. M.L.S. conducted the majority of the lab work with assistance from E.L.G. M.L.S. drafted the manuscript for submission, and all co-authors commented on and edited the manuscript. DATA AVAILABILITY The raw sequence reads are deposited in the National Center for Biotechnology Information (NCBI; GenBank Sequence Read Archive accession no. PRJNA577479). Sequence information for the developed primers has been deposited to NCBI; accession numbers are provided in Table 1. Prophage protein RacR activates lysozyme LysN, causing the growth defect of E. coli JM83 Prophage enriched the prokaryotic genome, and their transcriptional factors improved the protein expression network of the host. In this study, we uncovered a new prophage-prophage interaction in E. coli JM83. The Rac prophage protein RacR (GenBank accession no. AVI55875.1) directly activated the transcription of φ80dlacZΔM15 prophage lysozyme encoding gene 19 (GenBank accession no. ACB02445.1, renamed it lysN, lysozyme nineteen), resulting in the growth defect of JM83. This phenomenon also occurred in DH5α, but not in BL21(DE3) and MG1655 due to the genotype differences. However, deletion of lysN could not completely rescued JM83 from the growth arrest, indicating that RacR may regulate other related targets. In addition, passivation of RacR regulation was found in the late period of growth of JM83, and it was transmissible to daughter cells. Altogether, our study revealed part of RacR regulatory network, which suggested some advanced genetic strategies in bacteria. annotated as a lysozyme gene, which is highly probable to participate in the process of cells lysis. Subsequently, LysN was overexpressed via pBAD-lysN (plysN) in JM83. The growth curve showed that the turbidity of culture decreased significantly, while the number of living cells only decreased slightly (Fig. 3a,b). In addition, LysN can also lead to lysis of the other four strains (Fig. 3a), suggesting that LysN may be indeed the direct functional protein associated with lysis. As expected, the fractal pattern due to LysN overexpression observed in JM83, DH10B, MG1655 cells was quite similar to pracR/JM83 overexpression case (Figs 1c and 3c). Combined together, we hypothesize that LysN is one of the direct effectors in the RacR regulatory loop to damage the cells. inactivation of lysN rescues JM83 from lysis. To further confirm whether lysN is under control of RacR and accordingly lead to cell lysis, the lysN mutant strain was tested, of which the region from +91 to +201 was replaced by a linker plus 3 × Flag-tag (111 bp) and ended with the stop codon TGA (Fig. 4a). Compared to JM83, cell density of lysN mutant strain (ΔlysN) stayed at a medium level (Fig. 4b), and SEM observation showed that cell debris almost disappeared (Fig. 4c). Taken together, these results showed that inactivation of lysN rescues JM83 from lysis. Meanwhile, when introducing the lysN expression cassette into pracR, as expected, cells would lyse, which indicated that RacR indeed regulates lysN in vivo ( Fig. 4b and Supplementary Fig. S1). We also noticed the cell elongation phenomenon by RacR overexpression in wild type, ΔlysN strains via SEM, no matter with or without lysN expression (Figs 1c and 4c). This suggests that there may also exist targets other than lysN regulated by RacR, which is responsible for cell elongation instead of cell lysis. While although the elongated cells overexpressing RacR didn't show significant culture turbidity decrease (purple line in Fig. 4b) or cell debris (Fig. 4c) in the first 2-5 hours, the living cell number was decreased significantly (purple bars in Fig. 4d). The decrease of living cells overexpressing RacR was even more serious than lysN ovexpressing cells in the first 3 hours but ultimately became close after 5 hours. This may imply that the other regulatory pathway other than lysN by RacR is also quite important for cell survival. RacR triggers transcription of the lysN directly. We then measured the transcription level of lysN in pracR/JM83. The accumulation of RacR led to a more than 4000-fold increase in lysN mRNA level, which is almost silent without RacR induced expression (Fig. 5a). It suggests that RacR has the ability to trigger the transcription of lysN. Then, electrophoretic mobility shift assay (EMSA) was performed with purified His 6 -RacR and the DNA probe (designated lysN * : from −235 to +7 relative to start codon of lysN). The top of Fig. 4a showed the potential binding motifs in lysN * that derived from half of the palindromic sequence 5′-GCCTAA-3′ and 5′-TTAGGC-3′ 27 . As shown in Fig. 5b, His 6 -RacR was observed to bind to lysN * probe in a concentration-dependent manner. This interaction was nearly completely blocked by addition of 150-fold unlabeled lysN * , while not by addition of the 150-fold unlabeled unspecific DNA. Furthermore, the upstream region and 90 bp sequence at the 5′ end of lysN was fused to the reporter gene lacZ, and the mutations were introduced into the potential binding |
motifs, as shown in Supplementary Fig. S2. The result showed the activity of controls were extremely low, while overexpression of RacR led to a 140-fold increase (Fig. 5c), which was highly consistent with the transcriptional levels of lysN in the genome (Fig. 5a). In addition, base substitution mutation significantly decreased the activity of LacZ, although not completely inhibited ( Supplementary Fig. S2). These results demonstrated that RacR activates transcription of lysN by directly binding to the promoter region. JM83 restored after 18 h of induction due to passivation of regulation. However, we were soon puzzled by the subsequent behavior of pracR/JM83, since the cell density of culture would begin to recover after 8 h of induction (Fig. 6a). What's more, replenishment of strong inducer at 9 h point would not interfere with the original growth trend (Fig. 6a), indicated that the recovery of growth was not due to insufficiently induced expression of RacR. We then checked the morphology of pracR/JM83, plysN/JM83, and pracR/ΔlysN after 18 h of induction, as we expected, Fig. 6b confirmed the recovery of all these strains. To answer this question, we performed qRT-PCR. We found that the mRNA level of lysN decreased significantly at 18 h after induction (Fig. 7a). Hence, we inferred that there is a strong connection between lysN mRNA level and cell status. We assumed the possibility that the decline of RacR or the passivation of its regulation causes a lower transcriptional level of lysN, and detected cellular RacR protein level of each time points. As shown in Fig. 7a, the His 6 -RacR protein level did not change by culture, which excluded the possibility of cascaded decrease of RacR and LysN. Even so, we decided to introduce another plasmid in those recovered cells to produce RacR, in case there are any undetectable changes in pracR. We selected monoclonal recovered pracR/JM83 and renamed it JM83-Anti (resist the overexpression of RacR, Anti for short). After checked the availability in JM83 www.nature.com/scientificreports www.nature.com/scientificreports/ (Fig. 7b), pCA24N-racR was selected to produce pCA24N-racR/Anti (contain pracR and pCA24N-racR). We did not observe a obvious difference in the expression levels of RacR between JM83 and Anti strain, however, the latter cell growth was no longer inhibited (Fig. 7b). This result strengthened the hypothesis of passivation of RacR regulatory, and suggested this negative effect is permanent. Indeed, continuous cells culture imply the heritable peculiarity of passivation, since all daughter cells grew normally in the presence of L-Arabinose (Fig. 7c, G2 to G5). To sum up, although this particular third factor was not identified, we have uncovered the mechanism of cell growth recovery. In the following research, we will explore more deeply about the passivation of RacR regulatory. Discussion RacR was previously predicted as a transcriptional repressor that belongs to MerR superfamily, whose inhibitory effect on Rac prophage toxin YdaS has been demonstrated recently 27 . In our study, RacR overexpression triggered ϕ80 lysozyme LysN and resulted in cell lysis. This kind of relationship is highly similar to the effect of RerR on the toxin genes in Clostridium difficile 28 . The phage ΦCD119 regulator RerR has been shown inhibit to the distant toxin genes tcdA and tcdR by directly binding of their promoter, and the PaLoc (pathogenicity locus) of these toxin genes is commonly thought to belong to mobile genetic elements 29 . The activation and inhibition to LysN and YdaS mean that the host must provide a tight mechanism to adjust the concentration of RacR. In fact, the irregular palindrome in the racR-ydaS intergenic region raised the possibility that RacR might negatively regulate its own transcription ( Supplementary Fig. S2). This assumption is consistent with the characteristic of regulators of MerR superfamily, which bind to palindrome sequences and were reported to regulate their own expression 30 . Another characteristic of these proteins is the N-terminal located HTH domain and irregular C-terminal domain, the latter is commonly used to bind metal ions, such as Hg 2+ for Tn501 protein MerR 31 . The lysN transcription decreased significantly at 18 h, even reached the same level as the wild-type JM83, while the RacR was constant (Figs 5a and 7a). We propose a hypothesis that the N-terminal of RacR binding the promotor of lysN at the early stage, with the assistance of an unknown metal ion. While at the late stage, the concentration of these ions decreased and the free RacR no longer activates the expression of lysN. However, only three imperfect conserved motifs were found in the upstream region of lysN, and their base composition and spatial arrangement neither can be compared to the regulatory sequence of ydaS. More than that, obvious enzyme activity was detected when lacZ under the control of a mutant lysN promoter ( Supplementary Fig. S2), which suggested that these "CCTA" containing motifs were not the key sequences for RacR regulation, and the promoter region of lysN contain some more important motifs. Although we have uncovered the causes of RacR overexpression leading to JM83 lysis, the growth of ΔlysN strain cannot recover completely (Fig. 4b), indicated that RacR may also influence other genes. In the early stages of our research, the famous phage regulators CI, CII, and Cro were also considered as potential targets for RacR. We detected their mRNA in ΔlysN after 2 h and 18 h of RacR overexpression, as shown in Supplementary Fig. S3. It seems that RacR tried to break the balance of these three regulators in E. coli to establish a phage lytic state 32 , since the concentration of primary repressor CI was reduced and its negative regulator Cro was increased. However, we did not find active phages in our solid or liquid medium, and the cultures containing broken JM83 cells were not infectious. We speculate that the defects of ϕ80dlacZΔM15 prophage prevent the assembly and packaging of ϕ80, but whether CII or Cro plays a role in lysN expression and growth defect of ΔlysN is unknown. On the other hand, the elongation of cells is probably related to division genes, as the proteins that constitute the divisome 33 , the ZapA-ZapB complex 34 , and the Tol-Pal system 35 . It has been reported that the interference with cell division leads to an elongation phenotype in E. coli 36,37 , which is extremely similar to the cell morphology of RacR overexpressing JM83 (Figs 1c and 4c). In summary, the regulation of RacR to lysN is special, since they belong to two different prophages. While in the typical phage lytic cycle, the S holin and R transglycosylase are under the strictly controlled of their own major phage regulators, which activated in the late stage to release the phage 38 . Although the physiological significances for lysozyme activated by foreign regulator still unclear, the model of the cell response to LysN and eventual recovery (Fig. 6a) reveals a diversity of bacterial genetic strategies. We attempted to find clues of the temporal expression of LysN, by introducing a Flag-tag in JM83 genome (Fig. 4a). However, the target protein (LysN_30N-Flag, Figure 7. The regulation of RacR is invalid in all daughter cells. (a,b) qRT-PCR analysis of pracR/JM83 and growth curves of JM83 and Anti after induction. Western blotting indicates the His 6 -RacR level in each strain, and the full-length blot is presented in Supplementary Fig. S4. The Anti strain was depicted in this study. (c) The continuous cell culture of pracR/JM83. The whole process contains five generations (G1, G2, G3, G4, and G5). Data represent means ± standard deviations of results from three independent experiments. (2019) 9:12537 | https://doi.org/10.1038/s41598-019-48690-4 www.nature.com/scientificreports www.nature.com/scientificreports/ 8.2 kDa) was not detected successfully, and we supposed that the short half-life period of a small artificial protein affects its detection, since the mRNA of LysN_30N-Flag was transcribed (Supplementary Fig. S5). Growth curves and spotting assay. Growth was monitored by measuring the optical density (OD) at 600 nm (OD 600 ). A single colony of each strain was inoculated in LB and grown at 37 °C overnight. Then, the strains were transferred to 500 mL flasks containing 100 mL of LB medium and were cultured at 37 °C in a shaking incubator (190 rpm). 0.2% L-Arabinose and/or 1 mM IPTG was added when OD 600 reached about 0.6. We recorded the optical density of these strains at an hour intervals for 8 hours at 28 °C. Meanwhile, 100 μL bacterial suspensions after 1 hour, 3 hours and 5 hours of induction were harvested, 10-fold gradient diluted in fresh LB medium and spread on LB agar plates. The plates were incubated at 28 °C for 24 hours followed by calculating the average colony-forming units (CFU) per milliliter according to the formula [(viable count from each concentration × dilution fold × 10)/n]. Above assays were repeated in triplicate. Scanning electron microscope (SEM). Equivalent cell densities of different E. coli strains were collected through centrifugation (2300 × g for 5 min at 4 °C) and washed three times with phosphate buffer (PBS, 0.1 M, pH 7.5). Then, the cell pellets were fixed with 2.5% glutaraldehyde at 4 °C for 5 hours. After washing three times at 4 °C, these samples were dehydrated for 10 min each in increasing concentrations of ethanol (30%, 50%, 70%, 80%, and 90% (V/V)). Subsequently, the samples were frozen at −80 °C for about 24 hours, dried with a vacuum freeze dryer, and then observed with S-3400N scanning electron microscopy. construction of lysN in-frame deletion mutant ∆lysN. All primers used in mutant construction are listed in Supplementary Table S2. PCR amplifications were performed to generate the upstream fragment of lysN with primer pair 19SY-1/19SY-2 and the downstream part with primer pair 19XY-1/19XY-2. Otherwise, we introduced a linker plus 3 × Flag-tag sequences which replaced the in-frame deletion region from +91 to +201 in lysN (Fig. 4a). The PCR product containing a site-directed deletion of lysN was obtained via overlap PCR with primer pair 19SY-1/19XY-2 and ligated into the NheI/XbaI site of the suicide vector pDMKE (the insB deleted derivative of pDMK 39 ). The resulting plasmid, pDMKE-lysN, were duplicated in E. coli DH5α(λpir) and then electrotransformed into E. coli JM83. Single colonies selected on LB plate with kanamycin and chloramphenicol suggest that the plasmid was integrated into the chromosome by homologous recombination. The double-crossover recombination was selected on LB plate with 10% sucrose. The lysN in-frame deletion mutant was designated as ∆lysN and confirmed via PCR and sequencing. Quantitative real-time PCR (qRT-PCR). RNA from E. coli JM83 or ∆lysN strains frozen at −80 °C was extracted using Pure RNA Isolation Kit according to the manufacturer's protocols. For removal of the remaining DNA, total RNA was incubated with RNase-free DNase I at 28 °C for 1 hour. 1 μg total RNA was used to generate cDNA using Reverse Transcription M-MLV (RNase H-) kit. Subsequently, quantitative real-time PCR was performed according to SYBR Green PCR Master Mix and each sample was made in triplicate. rpoD acts as the internal reference gene. To normalize data, transcription levels of the rpoD gene in all samples were set to 1.0. Relative mRNA levels were analyzed using the 2 −ΔCt (ΔCt = Ct tested genes − Ct rpoD ) method. The primers for qRT-PCR are listed in Supplementary Table S2. Overexpression and purification of RacR protein. The racR gene was PCR-amplified from E. coli JM83 and cloned into the NdeI/EcoRI site of the pET-28a (+) vector to yield pET28a-racR with an N-terminal His 6 -tag. The RacR expression plasmid was transformed into E. coli BL21(DE3). The E. coli strain was induced with 1 mM IPTG until OD 600 reached about 0.6 and grown at 16 °C for 16 hours. Then, the strain was harvested by centrifugation (5900 × g for 5 min at 4 °C) and washed three times with phosphate buffer. Pellets were resuspended to a final concentration of 10 OD/mL, sonicated on ice, and centrifuged at 5900 × g for 5 min at 4 °C. The protein was then purified via nickeliminodiacetic acid-agarose chromatography and desalinated into 1 × binding buffer (10 mM Tris-HCl [pH 8.0], 1 mM EDTA, 0.1 M NaCl, 0.1 mM dithiothreitol, 5% glycerol, and 10 μg/mL bovine serum albumin 27 ). Purified protein was analyzed by 12% SDS-PAGE, and the protein concentration was determined by the Bradford assay. Electrophoretic mobility shift assay (EMSA). EMSAs were carried out using |
the purified His 6 -RacR and PCR-amplified DNA probes. The biotin-labeled probes were obtained by PCR with primer head-biotin in Supplementary Table S2, then purified and quantified. Increasing amounts of RacR were added to the 1 × binding buffer that containing target lysN * probes (5 ng) and 50 μg/mL poly(dI·C), and incubated at 28 °C for 40 min. Samples were run on a 6% polyacrylamide gel in 0.5 × TBE buffer at 130 V for 1 hour, then transferred to a nylon membrane at 380 mA for 55 min, subsequently analyzed using chemiluminescent EMSA kits. A Novel DNAzyme Signal Amplification-Based Colorimetric Method for RNase H Assays A simple visual strategy was developed for the RNase H colorimetric measurement using DNAzyme-mediated signal amplification. When RNase H was presented, the RNA strand of the duplex formed by the G-rich DNA sequence (G-Rich) and its complementary RNA sequence (cp-RNA) was digested, releasing G-Rich to form HRP-mimicking DNAzymes of the G-quadruplex/hemin complexes in the presence of hemin. These DNAzymes catalyze the oxidation reaction of the substrate of 2,2′-azino-bis (3-ethylbenzothiazoline6-sulfonic acid) (ABTS) to produce green color products of ABTS•, allowing for the detection of RNase H. A horseradish peroxidase (HRP)-mimicking DNAzyme of the Gquadruplex/hemin complex was used to mediate the signal amplification in the sensing strategy, resulting in high selectivity and sensitivity. This proposed colorimetric method shows a low detection limit of 0.04 U/mL, with a detection range of 0.1 to 3 U/mL. Moreover, this colorimetric method has been successfully used for RNase H assays in complicated biosamples, such as cell lysates. These results indicate that our colorimetric method not only detects RNase H in an ideal system but also in real samples. Introduction RNase H (Ribonuclease H), a highly conservative endogenous ribonuclease that attracts significant attention and is studied worldwide due to its biological functions, has been implicated in diseases 1-3 , primarily because RNase H participates in biological processes involving replication 4, 5 , repair 6 , and reverse virus transcription 7,8 , RNase H can only specifically hydrolyze the phosphodiester bonds in the RNA strand of a heterogonous DNA/RNA duplex 9,10 . In past decades, detection and quantification methods of RNase H, like gel 11 or capillary 12 electrophoresis, HPLC 13 , and fluorometry 14 have been mainly constructed of cumbersome and infrastructure-heavy analyses, thus limiting the application of these methods in resource-poor environments. A simple robust method for RNase H measurement is still urgently needed in both biomedical research and clinical diagnostics. Since RNA was found to have catalytic functions, the theory that all enzymes are proteins has been discarded 15 . Subsequently, many nucleic acids with catalytic functions were not only found in nature but also by systematic evolution of ligands by exponential enrichment (SELEX) 16,17 . As one kind of nucleic acid enzyme, DNAzymes are capable of catalyzing biochemistry reactions [18][19][20] . DNAzymes are more popular with researchers than protein enzymes, due to their cost-effectiveness, activity, and stability 21,22 . The DNA sequence containing rich guanine can fold to form a G-quadruplex (G-qua) 23 . Introducing hemin forms a G-quadruplex/hemin conjugate, a new kind of DNAzyme that mimics horseradish peroxidase (HRP) to catalyze the oxidation reactions of the substrates 24 . 40 mM KCl, 8 mM MgCl2, and 0.05% Triton X-100) was first placed in 90 o C for 5 min, and then incubated at 37 o C for 2 h to guarantee complete hybridization of G-Rich with the cp-RNA. Before use, the prepared G-Rich/cp-RNA duplex probe system was stored at RNase H activity assay with the probe system Feasibility study of the colorimetric sensing method In this colorimetric sensing method for RNase H assay, cp-RNA strands of G-Rich/cp-RNA duplexes were efficiently hydrolyzed by RNase H. The released G-Rich sequences fold into G-quadruplexes, later combining with hemin to form G-quadruplex/hemin conjugates mimicking HRP to catalyze the redox reaction of ABTS and producing green products of ABTS· -. To investigate the strategic feasibility, the ABTS-H2O2 system was first used to study the prepared probe system. As shown in Figure 1A, when only hemin was presented in the ABTS-H2O2 system solution, its absorbance at 418 nm was at a lower level (curve 1) and the solution was colorless (picture 1), due to hemin having little activity to catalyze the ABTS-H2O2 system. When both G-quadruplex and hemin were presented in the ABTS-H2O2 system solution, a significant increase in absorbance at 418 nm was observed (curve 2) and the solution's color was dark green (picture 2), because the hemin combined with a G-quadruplex to form the HRP-mimicking DNAzyme of the G-quadruplex/hemin conjugate. However, when cp-RNA was presented and the G-Rich/cp-RNA duplex was formed, the ABTS-H2O2 system solution containing "hemin + G-Rich/cp-RNA duplexes" was colorless (picture 3) and its absorbance at 418 nm was still at a lower level (curve 3) because there were no free G-quadruplexes to combine with hemin to form the conjugate. When the G-Rich/cp-RNA duplexes were first treated with RNase H, the ABTS-H2O2 system solution containing "hemin + G-Rich/cp-RNA duplexes treated with RNase H" was also dark green (picture 4), and its absorbance at 418 nm also showed a significant increase (curve 4), which was mainly due to the fact that RNase H can hydrolyze the cp-RNA strand of the G-Rich/cp-RNA duplexes. Then Polyacrylamide gel analysis was performed to verify method feasibility. Figure 1B Optimization of the colorimetric sensing method One can see from the designed sensing strategy for RNase H assay (Scheme 1) that the sensing performance of the colorimetric method is affected by background signals, which are dependent on free G-DNA sequences, as well as hemin. To complete the hybridization of G-Rich with cp-RNA, the amount of cp-RNA should be larger than that of G-Rich. However, too much cp-RNA can decrease the sensitivity of the colorimetric method. To guarantee high sensing performance, an appropriate molar ratio of G-rich/cp-RNA should be utilized. The molar ratio of G-rich/cp-RNA was optimized by using an ABTS-H2O2 system. As shown in Figure 2A system solution decreases if the cp-RNA amount is excessive. Therefore, an appropriate molar ratio of G-rich/cp-RNA was determined to be 1:1.2 in this colorimetric sensing system and was adopted in the assay. Hemin, alone, has some enzymatic character, which can slightly catalyze the oxidation reaction of the substrates 26 . Hence, a significant concentration of hemin will lead to a high background signal, while a low concentration of hemin is not enough to make the G-Rich sequences generate complete G-quadruplex/hemin complexes, thus decreasing the method sensitivity. Accordingly, hemin concentration was also optimized to improve the performance of the colorimetric sensing method for RNase H assay. As shown in Figure S1 and Figure 2B, the absorbances at 418 nm of both the "G-Rich/cp-RNA duplexes" sample solution and the "G-Rich/cp-RNA duplexes treated with RNase H" sample solution were increased with increasing hemin concentration ( Figure S1). The absorbance change (A1-A0 at 418 nm) reached a maximum at 400 nM hemin ( Figure 2B), corroborating that an appropriate concentration of hemin aids the colorimetric sensing method. 400 nM of hemin was employed in the subsequent sensing assay. Sensitivity and specificity of the colorimetric sensing method This method was first used under optimal conditions for the determination of RNase H at different concentrations. One can see from Figure 3 that the absorbance at 418 nm ( Figure 3A) and the intensity of the green color ( Figure 3B) of the sample solutions gradually increased with increasing concentrations of RNase H (0 ~ 15 U/mL). The concentration of RNase H is in the range of 57~96 U/mL in normal human serum or the patient's serum 30 , several microlitres (μL) of serum sample will be enough to realize the RNase H measurement in real applications. These results demonstrate that there is sufficient sensitivity in this method. Five different kinds of enzymes, including EXO Ⅰ, EXO Ⅲ, DNase Ⅰ, S1, and RNase A were tested to further investigate method selectivity. One can see from Figure 4 that the absorbance of the sample solution significantly increased when 5 U/mL of RNase H was used, while only slight increases of the absorbance intensity were observed for the other enzymes (25 U/mL), whose concentrations were five times higher than that of the RNase H. Results indicate that the proposed colorimetric sensing method demonstrates high specificity for RNase H detection over other interference enzymes. RNase H assay in cell lysates with the colorimetric sensing method To investigate the feasibility of the proposed method for RNase H measurement in real Conflicts of interest There are no conflicts to declare. Table 2 and Figure 5A). Figure S2 in "Supporting Information"). Vertebral artery and posterior inferior cerebellar artery aneurysms: Results of microsurgical treatment of eighty patients Background: The choice of surgical approaches and options for the microsurgical vertebral artery (VA) and posterior inferior cerebellar artery (PICA) aneurysms repair remains controversial. Methods: A retrospective analysis of the clinical, surgical, and angiographic data of 80 patients with VA and PICA aneurysms treated from 2012 to 2018 was performed. Results: The aneurysms were saccular in 50 cases (62.5%) and fusiform in 30 cases (37.5%). The median suboccipital craniotomy was the most common approach (73.8%). Retrosigmoid craniotomy was performed in 25% of patients. There were the following types of microsurgical operations: neck clipping (61.25%), clipping with the artery lumen formation (13.75%), trapping (10%), proximal clipping (5%), and deconstruction with anastomosis (10%). Fifty-seven (71.3%) patients were discharged without worsening of the clinical signs after surgery. The most common postoperative neurological disorder was palsy of IX and X cranial nerve revealed in 14 (17.5%) patients. No fatal outcomes or patients in vegetative state were identified. The complete occlusion of PICA and VA aneurysms according angiography was in 77 (96.3%) cases. Conclusion: Microsurgical treatment is an effective method for VA and PICA aneurysms. The majority of VA and PICA aneurysms do not require complex basal approaches. A thorough preoperative planning, reconstructive clipping techniques, and anastomoses creation, as well as patient selection based on the established algorithms and consultations with endovascular surgeons, may reduce the number of complications and increase the rate of complete microsurgical occlusion in VA and PICA aneurysms. INTRODUCTION e incidence of the aneurysms of the vertebral artery (VA) and posterior inferior cerebellar artery (PICA) is 2-4.5% of all brain aneurysms. [5,8,20] e most common presenting symptoms include subarachnoid and ventricular hemorrhage associated with a high risk of disability and death without surgical treatment. [20,22,27] e choice of treatment for VA and PICA aneurysms (microsurgical or endovascular surgery) remains controversial. [3,5,21,23,25,26] is manuscript presents the results of the microsurgical treatment of VA and PICA aneurysms, types of surgical approaches, and options for aneurysm repair according to location and anatomical features. MATERIALS AND METHODS A retrospective analysis of the clinical, surgical, and angiographic data of sequential series of patients with VA and PICA aneurysms treated in the N. N. Burdenko National Medical Research Center of Neurosurgery from 2012 to 2018 was performed. During this period, a total of 4106 patients with intracranial aneurysms underwent surgery in the Center of Neurosurgery. Among them, 130 (3.2%) patients had VA or PICA aneurysms. is group did not include 17 patients who had a combination of aneurysms with arteriovenous malformations of the posterior cranial fossa, as this group of patients had different management and outcomes. e choice of the surgery type was based on the institutional algorithms for cerebral aneurysms, [9,10] including: 1. Preference for endovascular surgery in patients with aneurysms of the intracranial VA segment 2. Preference for microsurgery for all PICA aneurysms and VA aneurysms with PICA originating from the aneurysm neck or sac. Contraindications for endovascular surgery included 1. e acute subarachnoid hemorrhage (SAH) requiring stenting and antiplatelet therapy 2. Resistance to or intolerance of the antiplatelet therapy 3. Limited endovascular approach to the aneurysm (VA kinking and atherosclerosis). In patients with contraindications to endovascular surgery or with unsuccessful endovascular attempts, microsurgery is recommended. Contraindications for microsurgical treatment included severe decompensated medical conditions and systemic hypocoagulation. A total of 80 microsurgical operations and 57 endovascular operations were performed. e outcomes of endovascular surgery are out of the scope of this manuscript. Clinical characteristics of patients in the study group Among all patients, 29 (36.3%) were males, and 51 (63.8%) were females. e age of patients ranged from 7 to 68 years (mean, 46.4 years). A history of |
SAH was reported in the majority of patients -75 (93.8%); however, only 12 patients underwent surgery within the first 14 days. Five patients (6.3%) had no history of hemorrhage: three patients had asymptomatic aneurysms that were incidental findings, and in two patients with giant aneurysms signs of the disease were associated with an increase of the mass effect. Eight patients (10%) had multiple aneurysms. PICA aneurysms caused hemorrhage in 5 of 7 cases when combined with aneurysms of the anterior circulation. Anatomical and topographical characteristics of aneurysms in the study group e intracranial part of the VA can be divided into the three segments: VA proximal to the PICA (VAprox); VA in the area of the PICA origin; and VA distal to the PICA (VAdist). Peerless and Drake classified all PICA as proximal (about 1 cm from the PICA origin) and distal. [20] Similarly, we classified VA aneurysms in the area of the PICA origin and aneurysms of the anterior medullary PICA segment as proximal PICA aneurysms (PICAprox), and all PICA aneurysms in the p2-p5 segments were classified as distal aneurysms (PICAdist). us, we identified four main segments of VA and PICA aneurysm localization: (1) VAprox; (2) VAdist; (3) PICAprox; and (4) PICAdist. Table 1 shows the aneurysms distribution by segments according to their shape and size, as well as the presence of intra-aneurysmal thrombosis. Surgical approaches in the study group ree types of approaches [ Figure 1] and four positions of patients on the operating table [ Figure 2] were used to repair the aneurysms. SURGICAL RESULTS e median suboccipital craniotomy (MSC) with lateralization toward the cerebellum hemisphere according to the location of the aneurysm was the most common approach (73.8%) [ Table 2]. In general, a linear incision was appropriate for this craniotomy (n = 51). In 8 cases, MSCs involved a hockey stick incision for lateralization and approach to the medial VA-PICA complex (to the anterior approach was made in the lateral position from a hockey stick incision [ Figure 2]. Retrosigmoid craniotomy from a linear or arcuate skin incision was performed in 25% of patients [ Table 2]. Commonly (n = 16), this approach was performed in the supine position with 90-degree head rotation and a slight additional contralateral rotation of the operating table [ Figure 2]. In general, this approach was used for the PICAprox aneurysm surgery [ Table 2]. e choice of the aneurysm surgery [ Table 2] was based on the aneurysm location, shape, size, and the presence of intraaneurysmal thrombosis. In saccular aneurysms, the clipping of the aneurysm neck was performed in the vast majority of patients (49 [98%] of 50) regardless of location. is type of aneurysm surgery was more common in the PICAprox aneurysms [ Table 2]. surface of the medulla oblongata). In several patients, hockey stick incision was used to isolate the occipital artery. e posterior arch of the cervical vertebrae was also resected through MSC in 40.7% [ Table 2]. In 47 patients, the MSC was performed in the sitting position. In 12 cases, where the aneurysm repair with anastomosis creation was considered before the surgery, the MSC was performed in the prone position. Far-lateral transcondylar approach with resection of the posterior arch of the atlas was performed only in one case with the VAdist aneurysm close to the VA confluence. e In the fusiform aneurysms, complex clipping with the artery lumen formation (CALF) was relatively more common. is operation was aimed at the clipping of the eccentric part of the fusiform aneurysm sparing the blood flow in the parent artery [ Table 2]. An example of a CALF is shown in Figure 3. Only in one of 4 cases with partially thrombosed saccular aneurysms, preoperative thrombectomy from the aneurysm cavity was required for the neck clipping. In fusiform partially thrombosed aneurysms (n = 15), preliminary thrombectomy before CALF was required in 2 cases with PICAdist aneurysms. In 8 cases of large and giant partially thrombosed PICAdist aneurysms, thrombectomy was performed after aneurysm clipping for decompression of the adjacent brain and nerve structures. In 5 cases with small fusiform aneurysms, thrombectomy was not performed. Deconstructive surgery (proximal clipping or trapping) was performed in 12 fusiform aneurysms [ Table 2]. CALF was unachievable or considered inappropriate due to the significant aneurysm extent or the uniform wall dilation. e decision about aneurysm occlusion with the PICA segment was made after confirmation of the satisfactory retrograde blood flow through the distal PICA based on the intraoperative video angiography with indocyanine green (ICG) [ Figure 4, I]. In the absence of retrograde blood flow, revascularization with the anastomosis creation was performed [ Figure 4, II]. Other criteria when PICA deconstruction should be combined with anastomosis may include: large diameter of the involved PICA, hypoplasia of the contralateral PICA or VA, as well as poor enhancement of the ipsilateral PICA and SCA according to angiography. In most cases (6 of 8), revascularization involved local (in situ) anastomoses: "side-to-side" in one patient [ Figure 5], "endto-side" in two patients, and "end-to-end" in three patients. In two patients, PICA revascularization was performed using the occipital artery [ Figure 6]. Deconstruction of the artery without anastomosis was possible in the VAdist fusiform aneurysms. e occlusion of this segment through proximal clipping was performed in two patients and trapping in one patient. e requirements for the VAdist occlusion with the aneurysm included PICA sparing (the clip was applied distal to the origin of the artery) and the presence of a second VA with comparable diameter to the main artery. Our attention settled on the fact that in fusiform aneurysms in the VA segment from the PICA origin to the VA confluence there were no large perforating arteries to the medulla oblongata. CALF waiving was due to the close association of this VA segment with the caudal cranial nerve group, the medulla oblongata, and the narrow surgical corridor [ Figure 7]. Two patients with PICAprox aneurysms and one patient with PICAdist aneurysms had no neurological complications after surgery. One patient had moderate dysfunction of IX and X cranial nerve after clipping the PICAprox aneurysm. In one female patient, the surgery was complicated by major intraoperative bleeding requiring clipping of the aneurysm with the right PICA origin that resulted in the postoperative ischemia in this region. Due to severe bulbar impairment, tracheostomy was performed. e patient was discharged from the hospital 90 days after surgery Glasgow outcome scale (GOS 3). In one patient with PICAprox aneurysm, a right PICA thrombosis in the area of its origin occurred after repeated clip reposition. An attempt of PICA reimplantation into the proximal VA segments was made; however, this anastomosis did not work. After surgery, the patient had right XII cranial nerve paresis. Postoperative computed tomography (CT) revealed no areas of brain ischemia. e patient was discharged on day 7 after surgery in good condition (GOS 4). Another patient with a partially thrombosed fusiform aneurysm of the tonsillomedullary (p3) segment of the left PICA developed PICA thrombosis after a CALF attempt. Intraoperative attempts to restore blood flow in the distal PICA were unsuccessful. After surgery, the patient developed massive ischemia in the left hemisphere and the vermis with severe edema in the posterior cranial fossa. Two days after the initial operation, decompressive craniectomy of the posterior cranial fossa was performed. e patient had prolonged sedation with artificial ventilation through a tracheostomy tube. He was discharged on the 47 th day after the surgery (GOS 3). At the time of discharge, the patient was conscious, had no movement disorders in the extremities; however, moderate bulbar disorders and severe cerebellar symptoms persisted. Air embolism during craniotomy was reported in two (3.8%) of 52 patients operated in the sitting position. After hemostasis with occlusion of large veins in both cases, the operation was continued. In the postoperative period, one patient developed bilateral pneumonia in combination with moderate dysfunction of IX and X cranial nerves that required the placement of a tracheostomy tube. Postoperative complications Fifty-seven (71.3%) patients were discharged without worsening of the clinical signs after surgery [ Table 3]. e mean duration of in-hospital stay in patients without neurological complications was 9.4 days. Among patients with postoperative neurological disorders, the most common sign was palsy of IX and X cranial nerve (dysphonia, dysphagia, etc.) identified in 14 (17.5%) patients. In 3 cases, these were severe and required the placement of a tracheostomy tube during the recovery period. In 11 patients, dysfunction of IX and X cranial nerves was observed with PICAprox aneurysms [ Table 3]. us, among 46 patients with PICAprox aneurysms, the incidence of bulbar disorders due to the caudal cranial nerves palsy was 23.9%. In three patients with signs of IX and X cranial nerve palsy, XII cranial nerve was also dysfunctional with tongue deviation and dysarthria. In 3 cases of PICAprox aneurysms, an isolated palsy of the ipsilateral XII cranial nerve was reported. Postoperative ischemic complications were observed in six patients. In three patients, aneurysm clipping was complicated by the PICA thrombosis and ischemic lesions in the cerebellar hemisphere. In three patients, MRI showed small ischemic lesions in the medulla oblongata manifested with moderate hemiparesis in one patient and with hemihypesthesia in two others [ Table 3]. No postoperative hematomas and other hemorrhagic postoperative complications were reported. Complicated postoperative wound healing occurred in three patients. In one case, external lumbar drainage was placed for 5 days due to the subcutaneous accumulation of cerebrospinal fluid in the postoperative wound area with subsequent recovery of this complication. Two patients had wound liquorrhea after surgery; revision surgery with the closure of the dura mater defects was performed. Subsequently, meningitis was diagnosed in one of these patients and etiotropic antibiotic therapy was administered with a good outcome. Postoperative complications associated with SAH were uncommon due to the relatively small number of patients in the acute period (12 patients). In one patient, operated 4 days after SAH, the focal hemispheric neurological signs developed 3 days after the surgery. Angiography revealed severe vasospasm in both MCA vascular regions that caused ischemic lesions in the cerebral hemispheres [ Figure 8]. Posthemorrhagic hydrocephalus was reported in 14 (18.7%) of 75 patients with SAH. In six patients, a ventriculoperitoneal shunt was implanted during the hospitalization. Complete aneurysm occlusion was confirmed in 43 (93.5%) cases, neck enhancement was revealed in 2 cases (4.3%), and enhancement of the neck and fundus -in one (2.2%) patient. Patients with a partial enhancement of the aneurysm neck were closely monitored. Endovascular aneurysm coiling was performed in one patient with an enhancement of the aneurysm fundus. In the other 34 patients without angiography, the complete aneurysm repair was verified by the ICG video angiography followed by opening the aneurysm sac. us, the complete occlusion of PICA and VA aneurysms during microsurgery was achieved in 77 (96.3%) cases. DISCUSSION ere are several points of view regarding surgical approaches to VA and PICA aneurysms. Some authors claim that large basal approaches with partial resection of the occipital condyles provide better aneurysm exposure and reduced risk of postoperative dysfunction of the caudal cranial nerves (DCCN). [2,4,8] In contrast, transcondylar approaches increase the total duration of the operation and increase the risk of poor wound healing, as well as craniocervical instability and neck pain. [8,28] We agree with those authors who believe that the resection of condyles is unnecessary in the vast majority of cases. [24,28] In some cases, the optimal surgical view may be achieved through intradural jugular tuberclectomy. [2,17,27] Indeed, the major challenge in the surgery of PICA and VA aneurysms involve neurovascular structures that obscure the neck of the aneurysm rather than bony prominences. [27] We believe that the need for routine resection of the posterior arch of the atlas to approach the aneurysms of VA and PICA proposed by several authors [4,8,23,24] is somewhat dramatized. In fact, it necessary for the microsurgical approach to VAprox aneurysms and low PICAprox aneurysms. For VAdist and PICAdist aneurysms, resection of the posterior arch of the atlas, in general, is not warranted. Our recent experience has shown that retrosigmoid craniotomy is an adequate approach to the small PICAprox and VAdist aneurysms that are consistent with other authors. [22,27,28] e benefits of the supine retrosigmoid craniotomy with contralateral head rotation include is the medial and upward gravity-facilitated traction of the |
cerebellum and brain stem from the base and VA elevation above the skull base. Due to the narrow surgery corridor, we do not use this approach in the large and fusiform aneurysms, especially when anastomosis may be required. Tjahjadi et al. [27] recommend a standard retrosigmoid craniotomy (simple lateral suboccipital approach) for PICAprox aneurysms when the PICA-VA complex is 10 mm or more above the great occipital foramen. Otherwise, they recommend to make the craniotomy medial and down to the lateral part of the great occipital foramen. Actually, even significant bleeding from the VA or PICA aneurysms rare causes microsurgical difficulties. Al-khayat et al. [2] noted that this approach results in the relatively small traction of brain structures, and after SAH the involved structures (brain stem and cranial nerves) become swollen less than the cerebral hemispheres. One of six patients who underwent surgery within 14 days from the SAH event developed ischemic lesions in the cerebral hemispheres due to vasospasm. Estimated risk of angiospastic supratentorial cerebral ischemia in subtentorial SAH originating from the VA or PICA is 1.9-7%. [2,7,15] Intraoperative removal of blood clots from the basal cisterns of the anterior and middle cranial fossa during microsurgery for VA and PICA aneurysms is not possible. erefore, a number of patients with high linear blood flow velocity in the middle cerebral arteries and progressive clinical deterioration may require cerebral angiography and intra-arterial treatment of vasoconstriction with vasodilators. [17] Postoperative mortality in the surgical series of patients with VA and PICA aneurysms is 1.8-3.7%. [2,24,28] Poor outcomes were reported to be directly associated with the severity of the patient's condition at admission. [5,23] In our study, the majority of patients were stable at surgery, and no fatal outcomes or vegetative state events were reported. Ischemic lesions in the cerebellum and brainstem in patients with PICA and VA aneurysms are primarily associated with surgical issues. In one study, postoperative ischemic cerebellar complications were diagnosed in 19% of patients. [27] In general, these were associated with occlusion of the PICA trunk, and the choice of a PICA segment where relatively safe occlusion may be performed is unclear. Many investigators state that ischemia rarely occurs when PICA is occluded distal to the lateral medullary (p2) segment because the stem perforating arteries are more proximal, and the distal blood supply is provided by the collateral branches from the anterior inferior cerebellar artery, the superior cerebellar artery, and contralateral PICA. [6,11,14,19] Our experience has shown that severe ischemic abnormalities in the cerebellum may develop with PICA occlusion at the tonsillomedullary (p3) segment. However, ischemia may not develop even when the PICA origin is cut. Chalouhi et al. [7] reported postoperative cerebellar infarction only in 4 (36.4%) of 11 patients with complete PICA clipping. Unfortunately, there are no accurate predictive criteria of the risk for PICA occlusion yet. We take guidance from the test with temporary PICA clipping and its retrograde enhancement in the ICG video angiography. However, its evaluation is simple only in extreme cases: the rapid intensive retrograde PICA enhancement along with other intact arteries or the complete absence of the enhancement. Currently, prolonged, mild, or moderate retrograde PICA enhancement may not predict the risk of ischemia. erefore, we believe that in undetermined cases, revascularization of the target artery may be useful. Seoane et al. [24] have shown that anastomosis in VA and PICA aneurysms should be created in 5.4% of cases. We performed such operations at 10%. In most patients, we preferred to bypass the VA or PICA aneurysm through anastomosis in situ. Other clinicians share a similar strategy. [1,12,16] When local anastomosis creation is impossible, an alternative includes revascularization using the occipital artery as a donor. [13,18,24] e main disadvantages of the occipital artery as a "donor" include multiple branches and a significantly smaller diameter in the distal segments. Due to the significant tortuosity, isolating the occipital artery from soft tissues is quite challenging and may result in its damage. Ipsilateral postoperative DCCN remains the main concern of microsurgery for VA and PICA aneurysms. According to the published data, the frequency of these complications after clipping of the aneurysm of VA and PICA varies from 7.4% to 29%. [5,27,28] According to our data, the incidence of DCCN in the general group of patients with VA and PICA aneurysms was 17.5%, and with PICAprox aneurysms -23.9%. ere are encouraging data that the rate of DCCN recovery within 6 months is over 76%. [2] To reduce the risk of DCCN, sharp vasoneural dissection is recommended to prevent the tension of the IX, X, and XII cranial nerve, shorten the time for preventive clipping (up to 6 min) and avoid temporary VA trapping, if possible. [2,27,28] e risk of DCCN increases with intraoperative ruptures. e control of intraoperative bleeding in VA aneurysms includes the approachability of the proximal, and, in some cases, distal VA. e distal temporary clipping, especially in the segment between PICA and VA confluence, often requires medial traction of the medulla oblongata and carries the risk of damage to the caudal cranial nerve group. Our study has shown that microsurgery provides an opportunity to achieve a high rate (95.5%) of the complete occlusion of VA and PICA aneurysms. According to the published data, the microsurgical technique has a high rate of complete aneurysm repair: 90-97.1%, [8,21,27] and in some publications, it reaches 100%. [24,25] In clinical series, including both microsurgical and endovascular methods, comparable clinical outcomes for aneurysms of VA and PICA were observed. [5,21,23,25] However, comparison of our data with clinical and angiographic outcomes of endovascular treatment in the VA and PICA aneurysms is inappropriate as we initially included in the study patients selected for microsurgical treatment according to the above criteria rather than consecutive patients. erefore, it seems promising to use a combined technique in multiple aneurysms, as well as in giant VA aneurysms where the endovascular deconstruction of the parent artery is possible in combination with microsurgical revascularization and elimination of the mass effect. e limitations of our study included: a retrospective design, the fact that in most patients, SAH was treated on a delayed basis, and the lack of follow-up data beyond discharge. CONCLUSION Microsurgical treatment is an effective method for VA and PICA aneurysms. e majority of VA and PICA aneurysms do not require complex basal approach with resection of the condyles and cervical vertebrae. A thorough preoperative planning using advanced neuroimaging methods, reconstructive clipping techniques, and anastomoses creation, as well as patient selection based on the established algorithms and consultations with endovascular surgeons, may reduce the number of complications and increase the rate of complete microsurgical repair in VA and PICA aneurysms. EEG Correlates of Learning From Speech Presented in Environmental Noise How the human brain retains relevant vocal information while suppressing irrelevant sounds is one of the ongoing challenges in cognitive neuroscience. Knowledge of the underlying mechanisms of this ability can be used to identify whether a person is distracted during listening to a target speech, especially in a learning context. This paper investigates the neural correlates of learning from the speech presented in a noisy environment using an ecologically valid learning context and electroencephalography (EEG). To this end, the following listening tasks were performed while 64-channel EEG signals were recorded: (1) attentive listening to the lectures in background sound, (2) attentive listening to the background sound presented alone, and (3) inattentive listening to the background sound. For the first task, 13 lectures of 5 min in length embedded in different types of realistic background noise were presented to participants who were asked to focus on the lectures. As background noise, multi-talker babble, continuous highway, and fluctuating traffic sounds were used. After the second task, a written exam was taken to quantify the amount of information that participants have acquired and retained from the lectures. In addition to various power spectrum-based EEG features in different frequency bands, the peak frequency and long-range temporal correlations (LRTC) of alpha-band activity were estimated. To reduce these dimensions, a principal component analysis (PCA) was applied to the different listening conditions resulting in the feature combinations that discriminate most between listening conditions and persons. Linear mixed-effect modeling was used to explain the origin of extracted principal components, showing their dependence on listening condition and type of background sound. Following this unsupervised step, a supervised analysis was performed to explain the link between the exam results and the EEG principal component scores using both linear fixed and mixed-effect modeling. Results suggest that the ability to learn from the speech presented in environmental noise can be predicted by the several components over the specific brain regions better than by knowing the background noise type. These components were linked to deterioration in attention, speech envelope following, decreased focusing during listening, cognitive prediction error, and specific inhibition mechanisms. How the human brain retains relevant vocal information while suppressing irrelevant sounds is one of the ongoing challenges in cognitive neuroscience. Knowledge of the underlying mechanisms of this ability can be used to identify whether a person is distracted during listening to a target speech, especially in a learning context. This paper investigates the neural correlates of learning from the speech presented in a noisy environment using an ecologically valid learning context and electroencephalography (EEG). To this end, the following listening tasks were performed while 64-channel EEG signals were recorded: (1) attentive listening to the lectures in background sound, (2) attentive listening to the background sound presented alone, and (3) inattentive listening to the background sound. For the first task, 13 lectures of 5 min in length embedded in different types of realistic background noise were presented to participants who were asked to focus on the lectures. As background noise, multi-talker babble, continuous highway, and fluctuating traffic sounds were used. After the second task, a written exam was taken to quantify the amount of information that participants have acquired and retained from the lectures. In addition to various power spectrum-based EEG features in different frequency bands, the peak frequency and long-range temporal correlations (LRTC) of alpha-band activity were estimated. To reduce these dimensions, a principal component analysis (PCA) was applied to the different listening conditions resulting in the feature combinations that discriminate most between listening conditions and persons. Linear mixed-effect modeling was used to explain the origin of extracted principal components, showing their dependence on listening condition and type of background sound. Following this unsupervised step, a supervised analysis was performed to explain the link between the exam results and the EEG principal component scores using both linear fixed and mixed-effect modeling. Results suggest that the ability to learn from the speech presented in environmental noise can be predicted by the several components over the specific brain regions better than by knowing the background noise type. These components were linked to deterioration in attention, speech envelope following, decreased focusing during listening, cognitive prediction error, and specific inhibition mechanisms. INTRODUCTION The human brain is remarkably capable of focusing on one specific sound while suppressing all others (Alain, 2007). Nevertheless, processing of relevant information largely depends on the specific interaction of the acoustic features of speech and noise signals, their informative content, attention, state, and the prior knowledge (familiarity with the presented topic) of the listener (Szalma and Hancock, 2011). To understand the underlying mechanisms of this diverse phenomenology in human sound interaction, short-term features of distracting events, state of the listener, information flow, and loss of efficiency need to be studied. One key aspect of the study design is ecological validity (Chaytor and Schmitter-Edgecombe, 2003), meaning that realistically complex stimuli and conditions are included possibly in addition to artificially designed stimuli. In a learning context, the ability to acquire and retain vocal information strongly affects the overall learning performance. This is even more challenging when this occurs in the presence of environmental noise. One of the effects involved in this ability is known as the cocktail party effect (Cherry, 1978), and this refers to the ability of the brain to direct attention to a target sound despite the presence of distracting sounds. Although the underlying mechanisms are indispensable to learn from information presented in an acoustically rich environment (Lehmann and Schönwiesner, 2014), they are far from fully understood. Attention directs both cognitive and sensory resources to the target sounds (Schneider and Shiffrin, 1977). In general, such resources |
are limited in capacity based on the bottleneck (Pashler, 1984) and capacity sharing (Kahneman, 1973) theories. Most of the observed effects of noise on learning (Alain, 2007) can be explained by attention, including unlocking undesired attention focus as well as an increased cognitive load when listening to speech in noise (Rudner, 2016). Moreover, listening performance and speech intelligibility in background noise can be impaired by distracting attention away from the narrative and hampering relevant sounds (Ljung et al., 2009;Clark and Sörqvist, 2012). However, attention focusing and appropriate gating of (ir)relevant stimuli are not only the matter of cortical processing but also peripheral neurophysiological stages of auditory analysis are involved. Attention can be modulated by bottom-up factors (referring to external stimulus-driven responses that guide the attention due to inherent properties of salient events relative to the background) as well as topdown task-specific functions (referring to internal modulation of attention that is driven by cognition based on prior knowledge, expectations, and learned schemas) (Katsuki and Constantinidis, 2014;Kaya and Elhilali, 2017). Auditory attention-related research (especially bottom-up attention) mostly adopts an event-related potential (ERP) design (Alain, 2007). However, a classical ERP design with repeated stimuli conflicts with the idea of ecologically valid stimuli and studying top-down attention. In the current paper, the single-trial EEG experiment was used to study how auditory-related neural responses vary depending on acoustical stimulus and listening condition. The power spectrum of EEG signal exhibits peaks in different frequency ranges reflecting different underlying mechanisms (Buzsáki et al., 2013;He, 2014). Therefore, one of the most common methods to process the single-trial EEG signals is spectral analysis, which relies on partitioning the signal into the different frequency subbands (Clayton et al., 2015). Previous studies using spectral analysis have shown different frequency bands contribute to the various underlying mechanisms during listening to speech in noise, such as top-down attention (Gazzaley and Nobre, 2012), cortical inhibition (Uusberg et al., 2013), language processing (Pulvermüller et al., 1997), neural entrainment to speech (Riecke et al., 2018), and excitation-inhibition balance . The roles of the different frequency bands in these mechanisms are discussed separately below. Low-frequency EEG signals (1 − 8 Hz) can be modulated by attention (Kerlin et al., 2010;Braboszcz and Delorme, 2011). Two important mechanisms may be associated with the low-frequency EEG. The first one is the mismatch between current and desired levels of attention (Clayton et al., 2015) and the transition of the fatigue state (Borghini et al., 2014), which is observed as a continuous increase of low-frequency power with time on task [unlike the alpha-band activity (8 − 13 Hz) (Mierau et al., 2017)]. Frontomedial theta-band (4 − 8 Hz) activity has been linked to both enhanced attention over short time-scale cognitive tasks and reduced attention (increased attentional fatigue) following long time-scale cognitive tasks (Wascher et al., 2014;Clayton et al., 2015). Moreover, it has been shown that the delta-band (1 − 4 Hz) absolute power is higher in the mind wandering compared to the focused state over the fronto-central region (Braboszcz and Delorme, 2011). The second mechanism is the information and attention selection (Schroeder and Lakatos, 2009;Herrmann et al., 2016). This means that the attention can use a mechanism of selection leading to oscillatory entrainment to a taskrelevant stimulus (Schroeder and Lakatos, 2009). However, neural entrainment is a broader concept and refers to the temporal alignment of neural signals with regularities in an exogenously occurring stimulus, such as speech (Obleser and Kayser, 2019) and even aperiodic (speech) signals (Obleser et al., 2012;Goswami and Leong, 2013). Speech following (and speech envelope following/tracking) as one the manifestation of the neural entrainment refers to the relation between the neural and sound signals (Obleser and Kayser, 2019). Although it has been measured in various frequency bands (Obleser and Kayser, 2019), its impact on low-frequency EEG (delta and theta bands) has been shown in several electrophysiological experiments (Luo and Poeppel, 2007;Doelling et al., 2014;O'Sullivan et al., 2014;Kayser et al., 2015). The basic hypotheses of these studies are the following: (1) entrainment occurs also at other frequencies, but this effect is obscured by stronger signals; (2) the low-frequency speech envelope entrainment of brain activity could be robust against different background noises ; and (3) the speech envelope is constituted by slow temporal modulations, which contribute to speech recognition despite different background sounds (Houtgast and Steeneken, 1985;Rosen, 1992;Kerlin et al., 2010;Ding and Simon, 2013;Ríos-López et al., 2017). It has also been shown that attended and unattended stimuli could be decoded by low-frequency single-trial EEG in a cocktail party scenario based on a stimulus-reconstruction algorithm (O'Sullivan et al., 2014). This stimulus-reconstruction method indicated the slow amplitude envelope of attended speech (≤ 8 Hz) is tracked more strongly by the low-frequency EEG (2−8 Hz) compared to the unattended speech. Furthermore, it has been shown that in the multi-talker speech perception, the attended speaker is represented over the non-primary auditory cortex (AC) while the individual speakers are represented over the primary AC (O'Sullivan et al., 2019). Alpha-band activity (∼ 8 − 13 Hz) is also often modulated by auditory attention, especially by the inhibition function (Strauß et al., 2014). The term "alpha-as-inhibition" is used to highlight that alpha-band activity, beyond resting state, could reflect inhibition of the distracting sound (Clark, 1996;Uusberg et al., 2013). Increased alpha-band activity over the taskirrelevant brain regions reflects less involvement of those regions. Hence, comparison of alpha power between taskrelevant and task-irrelevant cortical regions can be an indicator for inhibition (Pfurtscheller and Da Silva, 1999;Chang et al., 2010). In fact, alpha event-related synchronization (ERS) reflects inhibition and alpha event-related desynchronization (ERD) releases from inhibition (Foxe et al., 1998;Snyder and Foxe, 2010;Klimesch, 2012). Not only absolute alpha power over a fixed frequency band but also alpha peak frequency (APF) and its corresponding power can be associated with attention, inhibition, memory, and cognitive demand (Klimesch, 1997;Clark et al., 2004;Haegens et al., 2014;Gulbinaite et al., 2017). APF (Doppelmayr et al., 1998) and individual alpha frequency (IAF) (Klimesch, 1999) indicate the actual frequency limits of alpha activity, which exhibit variability within and between subjects (Haegens et al., 2014). APF is also linked to the number of spiking neurons or the input level (Mierau et al., 2017). If the input level increases with respect to the baseline level, APF increases until the oscillation becomes unstable and then it is replaced by a lower frequency (Mierau et al., 2017). Although APF increases with a higher allocation of attentional resources, it decreases with lower attentional demand and cognitive load due to unstable state and overloaded attention capacity (Hutt et al., 2016;Mierau et al., 2017). Higher APF can be accompanied by lower alpha power resulting in task-relevant regions that exhibit increased APF during task performance (Hutt et al., 2016). Studies focusing on power-related frequency shifts have suggested a rather complex relationship between alpha frequency and power (Kawabata, 1972). Other studies have shown that APF decreases with increasing attentional demand and task difficulty (Angelakis et al., 2004;Haegens et al., 2014), which could be explained by unstable state and overloaded attention capacity. Enhanced APF might reflect a state of cognitive preparedness and the attentional switch between wandering and focused states of mind (Braboszcz and Delorme, 2011). In addition to the peaks at the frequency ranges, a predominant " 1 f " component in the EEG power spectra leads a power-law function, i.e., p ∝ 1 f a , where p is power, f is frequency, and a is the scaling exponent (He, 2014). Therefore, the EEG time series exhibit scale-free dynamics and do not have a characteristic scale (He et al., 2010;He, 2014). Furthermore, the ongoing EEG signals hold a memory of their own dynamics on time-scales, which could be linked to the scale-free dynamics and the self-similarity concept in fractal geometry (Palva et al., 2013). Long-range temporal correlations (LRTC) are the most common measures with which to quantify how slowly the autocorrelations of the signal decay in power-law function (Linkenkaer-Hansen et al., 2001;Nikulin and Brismar, 2005;Palva et al., 2013). Alpha-band LRTC could reflect an optimal balance between excitation and inhibition states . Decreased alpha-band LRTC compared to the resting state correlates with better attentional performance (Colosio et al., 2017). Higher alpha-band LRTC during resting-state could predict high performance in decision making (Colosio et al., 2017), working memory (Mahjoory et al., 2019) and attention tasks (Irrmischer et al., 2018). Increased beta-band (∼13 − 30 Hz) power over the frontolateral region has been observed in the mind wandering compared to focused state (Braboszcz and Delorme, 2011). Furthermore, the beta-band activity can be related to the maintenance of current sensorimotor or cognitive task (Engel and Fries, 2010;Weiss and Mueller, 2012;Zhao et al., 2012). A quasi-harmonic relationship has been suggested between the beta and alpha peaks or central frequencies only during rest (Van Albada and Robinson, 2013;Haegens et al., 2014). The lack of a strict relationship between the beta and alpha peak frequencies during task-based conditions may reflect independent networks being activated (Jones et al., 2009;Haegens et al., 2014). Localized gamma-band activity (∼30 − 45 Hz) has been found in task-relevant cortical regions (MacDonald and Barth, 1995;Cervenka et al., 2011;Siegel et al., 2011). Gamma-band activity plays a central role in attention, perception and language processing (Pulvermüller et al., 1997). Furthermore, gammaband activity in sensory cortices has often been linked with enhanced attention to these particular sensory inputs (Ahveninen et al., 2013). It has also been shown that gamma-band power in auditory areas increases during extended auditory attention tasks (Kaiser and Lutzenberger, 2005;Ahveninen et al., 2013). According to popular theory, gamma waves may be implicated in creating the unity of conscious perception and semantic processing (Buzsaki, 2006). In this study, we aimed to investigate the different mechanisms involved in learning from the speech presented in noise using single-trial EEG and mimicking an ecologically valid context. To this end, 23 participants were exposed to the following listening tasks while 64-channel EEG signals were recorded: (1) attending to a lecture in the background noise (LA), (2) attending to the background noise alone (BA), and (3) not attending to the sound while still being exposed to the background noise (BUA). For the background noise, realistic environmental sound fragments from continuous highway noise (HW), fluctuating traffic (FT), and multi-talker babble (MT) were used. A written exam on the lecture was taken after 13 sets of 5-min lectures and the BA task for assessing the amount of information that participants have actually acquired and retained from the lectures. We hypothesized several neural mechanisms, such as cortical inhibition, auditory attention, neural entertainment, and predictive coding, can be affected by the listening conditions we have designed. Therefore, five qualitative hypotheses were considered: (1) alpha-as-inhibition, (2) excitation-inhibition balance reflected in the alpha band, (3) low-frequency envelope following, (4) maintenance of current cognitive task, and (5) semantic processing and cognitive prediction violation or error. The alpha-as-inhibition hypothesis (Uusberg et al., 2013) implies that alpha-band activity mediates inhibition of taskirrelevant cortical areas. The excitation-inhibition balance hypothesis relates the task performance and optimal information processing to the long-range temporal correlations of alpha-band activity. The low-frequency envelope following hypothesis (Luo and Poeppel, 2007;Kerlin et al., 2010;Obleser and Kayser, 2019) implies the neural entrainment and tracking of speech (and background sound) envelope can be reflected in the low-frequency bands, i.e., delta and theta frequency bands. However, here, the relation between the EEG and sound signals has not been analyzed (which is the main tool to measure the envelope following) due to our unsupervised approach. In fact, we have assumed that the strong representation of low-frequency EEG signals (i.e., changes in spectral characteristics) may be related to the envelope following. Although, the neural entrainment and envelope following occurs also at higher frequencies but this effect is obscured by stronger signals (note that no source reconstruction was used in this paper). The hypothesis of maintenance of current cognitive task (Spitzer and Haegens, 2017) implies that the preservation of the current brain state and the long-range communication can be associated with the beta-band activity. Finally, the last hypothesis suggests that semantic or higher-level processes (specifically semantic violations) due to speech processing induce power changes in the gamma-band activity (Braeutigam et al., 2001;Buzsaki, 2006;Hald et al., 2006;Penolazzi et al., 2009). Moreover, the generative models for the perception, such as the predictive coding (Sedley et al., 2016) assume the precision of prediction, changes to predictions, and violations (errors) in |
predictions are encoded with the alpha, beta, and gamma frequency bands, respectively. These assumptions can be in accordance with our hypotheses. Since a few EEG indicators, such as alpha peak frequency and power, alpha long-range temporal correlations, and delta absolute power were evaluated in a recent work by our group (Eqlimi et al., 2019), a wider range of EEG features (see below) was estimated for investigating our hypotheses. More precisely, the following features were estimated: spectral features and peak frequency of the alpha-band activity (hypothesis 1), the alpha-band LRTC (hypothesis 2), the spectral features of the delta and theta (hypothesis 3), the beta (hypothesis 4), and the gamma (hypothesis 5) frequency bands. To group these features, the hypothesis that different listening tasks (LA, BA, and BAU) create a variance in the EEG features that will also be responsible for at least part of the observed differences in learning from speech in noise, was used. Variance in the EEG features between participants is likewise expected to be informative for the observed differences in learning from speech. Different techniques are available for data-driven aggregation of the broad collection of EEG features. Principal Component Analysis (PCA) of the z-score for each feature is the lowest order approach. It could be extended to higher-order statistical methods and machine learning (e.g., using deep learning auto-encoders). Because of the amount of data available and the advantage of explainable results, we decided to use PCA based on z-score normalized data. To explain the meaning of the EEG-PC scores (the representation of EEG features in the PC domain), they were compared between the listening tasks (LA, BA, and BUA) and background noises (MT, HW, and FT) using linear mixed-effect modeling (Bates et al., 2015). Assuming that the EEG PCs grasp the main variance between listening conditions observed through the different listening tasks, a supervised analysis was performed to relate them to the information acquiring and retaining zscores (the exam results) in the lecture attended (LA) task using linear fixed and mixed-effect modeling. Also, for this predictive model, higher-order statistical approaches and machine learning techniques could have been used, yet we again opted for reducing the degrees of freedom in the model in view of the available data. Participants Twenty-three young healthy adults (mean age: 27 years, SD: 3.18, 13 females, 20 right-handed), all English speakers, participated in the experiment. Participants had normal hearing measured by pure-tone audiometry. All participants signed the informed consent and received modest financial compensation for their participation. Based on self-reports, none of them had a history of psychiatric or neurological disorders. A full battery of audiological tests was conducted including tonal audiometry, tympanometry, stapedial reflex measurement, speech in noise, and otoacoustic emissions (OAE) with contralateral suppression. No participants were excluded on the basis of this extensive testing of the auditory periphery. Our test population was young adults and therefore their hearing capabilities were fully developed (Klatte et al., 2013). Tasks and Stimuli The main stimulus was about 1 h of English lectures mixed with realistic background noise and presented through a loudspeaker while 64-channel EEG signals were recorded. Participants were instructed to pay attention to the lectures and were informed that there would be a written exam afterward. This task is hereafter referred to lecture attended (LA). The lectures were read by a male speaker and recorded in an anechoic room. To level out participants' particular interests, 13 different 5-min topics were presented over one long lecture. The lectures were related to topics for which prior knowledge is expected to be minimal in order to facilitate the focusing of attention during the presentation. For the background noise, three 5-min realistic environmental sound fragments from continuous highway noise (HW), fluctuating traffic (FT), and multi-talker babble (MT) sounds were used. Within these fragments, a few discrete instances of very salient sounds were added. In addition, four lecture fragments were presented in silence with a low level pink noise (PK) (a.k.a 1 f noise) at a level of 35 dB(A). The signal-to-noise ratio (SNR) of lectures and background noise was set to 5 dB, with lectures at a level of 68 dB(A) and overall background noise level at 63 dB(A). This assured that the background noise did not mask the lecture energetically. The sound levels reported here refer to the A-weighted equivalent continuous sound levels in decibels (LAeq) which were measured over about 360 s. For the multi-talker babble sound, recordings were made at a cocktail party where about twenty people were having conversations. The recorded speech was not intelligible. A few 3s phone ringing sounds were added to the multi-talker sound at certain times. For the highway sound, the noise of dense traffic was recorded, for which no individual car passages could be recognized. A few 5-s emergency vehicle sounds were added to the highway sound at certain times. For the fluctuating traffic sound, recordings were made at the corner of a one-way car lane with a bicycle lane next to it, close to a park. Car passages were added to the quietest periods of the fluctuating traffic noise. In addition, at certain times, a few 1-s sounds of honking car were added. The level of the salient sounds (emergency siren, phone ringing, and car's horn) was not high enough to mask the lectures energetically. The order of presentation was completely random in both lecture and background noise while assuring the two lectures in silence were not presented in succession. The written exam was presented after the BA condition (see below), which ensured that there was a time span of 45 min between the lectures and the exam. The purpose of presenting the exam is to quantify the amount of information that participants have actually acquired and retained from the lectures. The type of questions and evaluation of the exam is explained in section 2.9. A sufficiently long time interval between the learning phase and the memory retrieval during the exam was chosen for two reasons: (1) to avoid that the last lecture would be more prominently in short term memory; (2) to avoid sequential recall as much as possible. Testing the memory and learning in a timescale of minutes and hours was discussed in Tetzlaff et al. (2012) and Kelley and Whatson (2013). For example, memory retention was tested after 30 min (Menzel et al., 2001). The choice of 45 min was a compromise between the duration of the experiment and assuring the above. To increase the range of monitored listening conditions and to allow to implicitly calibrate for inter-person differences, the participants were exposed to two additional tasks. Firstly, as a reference for top-down attention-driven listening, 12 different 3-min fragments of background noise were presented with equivalent levels of 63 dB(A) and participants were asked to pay attention to the background noise by focusing on the number of salient events, such as phone-ringing, emergency vehicle, and honking car sounds. However, this was only to make them focus on the background sound and no questions were asked about this afterward. This task hereafter is referred to as background attended (BA). Finally, 12 different 3-min background noise fragments were presented and the participants were instructed not to pay attention to any sound, which hereafter is referred to background unattended (BUA). The BUA task is definitely different from the resting state because not paying attention to the low-level characteristics of the sounds is inevitable. Unlike the BA task, the participants during BUA were instructed not to focus on the information related to the salient events. BUA task was presented after the exam which made the participants very aware that no further attention was needed at this point, and they could relax. By listening task (or simply task) along with this paper, we mean the tasks that the participant had to perform during the experiment, i.e., LA, BA, and BUA. By listening condition (or simply condition), we mean the conditions that the subjects were flooded with the listening tasks and the stimuli. In total, all subjects were exposed to ten different listening conditions depending on the task and noise: LA-PK, LA-MT, LA-HW, LA-FT, BA-MT, BA-HW, BA-FT, BUA -MT, BUA-HW, BUA-FT. For instance, LA-MT refers to the condition that the task is LA and the background noise is MT. The experimental protocol is schematized in Figure 1. Figure 2 depicts the sound level fluctuations as a function of time (line plots) and standard spectrograms (heatmaps) for one of the sound fragments presented during the LA and BUA listening tasks. From Figure 2, the FT background noise stands out in terms of sound level fluctuation. For the HW background noise, the sound level is quite stable. Finally, the MT background noise exhibits somewhat more fluctuations in the sound level than the HW noise, but the differences between the loudest and the quietest sounds levels are higher in the FT. Note that the background sounds used in the LA and BUA tasks were the same (except the time duration). Furthermore, the type and order of background sounds presented in BA and BUA were identical for all participants. The only difference between the stimuli presented during BA and BUA is that additional salient sounds were added in the last three fragments of BA due to the increased chance of focusing on the background noise sound in the BA task. EEG Recording EEG signals during the different listening conditions were acquired continuously using a BioSemi System (Amsterdam, NL) from 64 active electrodes placed according to the standard 10−20 layout (Oostenveld and Praamstra, 2001) at a sampling frequency of 2, 048 Hz. Subjects were asked to keep their eyes open and focus on a dot located in the center of the monitor to minimize eye movement. Signals from seven external electrodes were also recorded which were applied to the nose, neck, two left & right mastoids (M1 and M2), left (HEOGL) & right (HEOGR) outer canthi, and below the left eye (VEOGD). In addition, two external channels were used for capturing the sound signals (SoundL and SoundR) together with EEG signals. EEG Pre-processing The EEG data were offline re-referenced to the nose electrode (channel 65th) and re-sampled to 512 Hz using an anti-aliasing finite impulse response (FIR) low-pass filter. The EEG data were then filtered using an FIR bandpass filter (Hamming windowed sinc) of order 3,380 from 0.5 to 134 Hz to remove extremely slow drifts and sharp oscillations. (3) Background unattended (BUA). In first task, in addition to multi-talker, highway and fluctuating traffic sounds as the background noises, the lectures were also presented in pink noise and without any background noise. After second task, a written exam was asked to complete about vocal information in the first task. Equivalent levels of background noise and lecture were ∼ 63 and 68 dB(A), respectively. The lectures were are shown by L.i, i = 1, . . . , 13, and the noises are distinguished by different colors in the figure. EEG signals were cleaned up in two steps. At first, nonrepeating big artifacts were removed based on visual inspection. In a second step, infomax independent components analysis (ICA) (Bell and Sejnowski, 1995) with EEGLAB version 13.1.1b (Delorme and Makeig, 2004) using default settings was applied to identify and remove eye blink and movement artifacts. To identify the ICA components related to eye artifacts, some rules of thumbs were applied: (1) no more than three ICA components were removed; (2) both temporal and spatial plots should confirm the diagnosis of eye artifact, meaning frontally located components and a typical blink or nystagmus pattern; (3) in case of doubt, the temporal pattern of the supposed ICA component was compared with the temporal pattern of the Electrooculography (EOG) channels to make sure that the incidence of potential eye artifacts coincide; (4) only eye artifacts were removed. Since playing audio files typically has a latency of a few milliseconds, the sound was recorded together with the EEG on a free channel which could be used to synchronize with the presented audio signal. For this purpose, at first, the presented audio files were re-sampled to 512 Hz (using an anti-aliasing FIR low-pass filter) and then the cross-correlation between resampled audio signals and recorded sound signals together with the EEG was calculated. The lag corresponding to maximum cross-correlation is the delay in audio files with respect to EEG measurement. |
To compensate for this delay, all 64-channel EEG signals were shifted with estimated delays. For the analysis in this manuscript, this synchronization is less important. Finally, the power spectrum plots of all EEG channels were visually inspected, and the fragments whose all channels were extremely noisy were excluded. In addition, using the power spectrum and a combination of visual inspection and automatic method (median-based criteria), the channels that were extremely noisy were excluded. EEG Signal Processing First, the continuous EEG signals were split into separate fragments corresponding to the 3 or 5 min exposures, based on sound signal recorded as extra EEG channel. Each EEG fragment was then analyzed per channel. Three types of EEG feature were estimated: (1) low-frequency-based features, such as absolute and relative powers, bandwidth, central frequency and spectral edge frequency for delta and theta activities, (2) alpha-band based features, such as alpha peak frequency and power, individual alpha frequency, absolute and relative powers, and alpha-band scaling exponent value as a dynamic measure to quantify LRTC, and (3) high-frequency-based features, such as bandwidth, central frequency, and spectral edge frequency for the beta and gamma signals. Moreover, wide-band absolute power, theta/alpha ratio power, and absolute power for lower and upper alpha were estimated. The subsequent sections describe how a broad range of EEG features was estimated. Absolute and relative powers (AP and RP) were calculated from the 64 scalp locations in the mentioned frequency bands. Relative power was computed as the ratio of power in a given band to sum of power from 1 to 45 Hz. Moreover, the θ α power ratio (RPTA) was computed. For the frequency band 1 to 45 Hz, only the absolute power was computed. In addition to these power-based features, the following frequency-based features (Szeto, 1990;Drummond et al., 1991;Estrada et al., 2004;Vural and Yildiz, 2010) were computed for the different frequency bands using the definitions in Vural and Yildiz (2010): (1) central frequency, (2) bandwidth, and (3) spectral edge frequency 95%. The central frequency (CF) is defined as the center of gravity for frequency between the lower and upper cutoff frequencies of the power spectrum. The bandwidth (B) quantifies the width of the power spectrum over a specific central frequency. The spectral edge (SE) frequency 95% is defined the frequency below which 95% of the total power (in a specific frequency band) are located (Szeto, 1990). Alpha Peak Frequency Based on Root-MUSIC To estimate the alpha peak frequency and power, we used the root-multiple signal classification (root-MUSIC) algorithm (Barabell, 1983). The root-MUSIC as a subspace-based method estimates the frequency content of a signal using an eigenspace method. The root-MUSIC algorithm has been described in recent work from our group (Eqlimi et al., 2019). In this paper, the preprocessed EEG signals were band-pass filtered at 7 − 13 and 8 − 13 Hz (using Butterworth band-pass filter of order 2) for two reasons: (1) there is no consensus on the alpha range (like other frequency bands) and both lower cutoff frequencies (7 and 8 Hz) have been used in literature (Freeman and Quiroga, 2012;Clayton et al., 2015); (2) it has been shown that there is a 2.8 Hz between-subject variability (mean = 10.3 Hz) for the alpha peak frequency (Haegens et al., 2014). The root-MUSIC algorithm was performed on each filtered EEG channel with P = 2 as the dimension of the signal subspace. The maximum powers in µV 2 and corresponding frequency in Hz were found. MP2713 and MP2813 terms (which are used in the following sections) stand for MUSIC-based alpha peak power which are estimated in alpha frequency ranges of 7 − 13 and 8 − 13 Hz, respectively with P = 2 components. The corresponding alpha peak frequencies are denoted by MF2713 and MF2813. Individual Alpha Frequency Based on Fitting Process Individual alpha frequency (IAF) could also be estimated based on the Gaussian fit approach (Nikulin and Brismar, 2006;Van Albada and Robinson, 2013;Haegens et al., 2014). We employed the algorithm which has been suggested in Neurophysiological Biomarker Toolbox (NBT) version 0.5.5 to quantify IAF. Firstly, PSD (p) and its corresponding frequencies (f) of each EEG signal with a 0.1 Hz resolution were estimated. The peak amplitudes and locations of p in the range of 8-13 Hz were found (using Matlab function "findpeaks"). A polynomial (y 0 = p 1 x + p 2 ) function was fitted to ln(p) for considering a 1 f baseline. Then, z ← e p 2 + f p 1 and s ← p − z were calculated to remove the 1 f component of the spectrum (Nikulin and Brismar, 2006). was fitted to the detrended power spectrum, s, to consider one peak. 95% prediction bounds, i.e., confidence interval, [cl l , cl u ] for a 1 and b 1 were calculated. If a 1 + y 0 (b 1 ) > cl u , then f α ← b 1 and p α ← a 1 . To determine IAF, center of gravity within the alpha band could be estimated. At first, the individual frequency interval, namely [f 1 , TF stands for transition frequency and defined as the EEG frequency lower than the alpha peak frequency showing the minimum power (Klimesch, 1999). Then, the center of gravity . Finally, f 2 was updated by f 2 = |5 − (IAF − 1)| + IAF and IAF was re-calculated based on same definition. Compared to root-MUSIC based alpha peak frequency (MF2813), the IAF is expected to be less sensitive to bandwidth around the observed frequency, yet both parameters are highly correlated. Long-Range Temporal Correlations of Alpha Activity Processes that do not have a characteristic scale (i.e., scale-free processes) cannot be described completely in terms of spectral concepts (e.g., peak frequency). There is convincing evidence that EEG time series exhibit scale-free dynamics (He, 2014). One of the successful methods to analyze these scale-free signals is longrange temporal correlations (LRTC). LRTC has been developed to quantify how much future dynamics of a signal are influenced by past temporal events (Linkenkaer-Hansen et al., 2007). In fractal geometry, LRTC could be interpreted by a selfsimilarity behavior, which suggests the signal dynamics are similar in different time scales. One of the most common techniques to quantify LRTC is detrended fluctuation analysis (DFA) (Peng et al., 1994). The presence of a trend in the signal can cause an overestimation of LRTC, hence DFA tries to eliminate the trend. Indeed, DFA is employed to quantify how slowly the autocorrelations of signals decay in power law, which is called the scaling exponent value, a. The power or scaling law states that a relative change in one quantity results in a proportional relative change in another, namely one quantity varies as a power of another. Distributions of the form p(x) = Cx −a are said to follow a power law. The constant a is called the exponent of the power or scaling exponent value (SEV) (Newman, 2005). If 0.5 < a < 1, the signal likely exhibits strong LRTC . We employed the DFA algorithm to quantify LRTC for each EEG channel signal in the alpha band using the NBT version 0.5.5 as suggested in Hardstone et al. (2012). First, the EEG signals were band-pass filtered from 8 to 13 Hz (alpha range used in Hardstone et al., 2012) using the Hamming windowed FIR filter of order 0.25 s (2 cycles of the lowest frequency, 8 Hz). Second, the amplitude envelope of the band-pass filtered signal was estimated based on the Hilbert transform. Third, the cumulative sum of the amplitude envelope was calculated as follows: where e(k) is the amplitude envelope at time instant k,ē is mean of the amplitude envelope, and c(t) is the cumulative sum of amplitude envelope at time instant t (a.k.a signal profile). We defined a set of window size, s = {s 1 , ...s N }, which are equally spaced on a logarithmic scale in a predefined calculation range. The cumulative sum of amplitude envelope (c(t)) was then split into a set of n separated time windows of length ∀ l ∈ s, which have 50% overlap. For each time window, the linear trend was removed using a least squares method and obtained the detrended version. After calculating the standard deviation of the detrended time windows, the fluctuation function as the mean standard deviation of all windows was computed as follows: where σŵl i is the standard deviation of ith time window of length l ∈ s, n is the number of time windows. Finally, we plotted the fluctuation function,f(l), along l on logarithmic axes. The slope of the trend line was computed in a predefined fitting interval using the linear regression as a measure for LRTC which is called scaling exponent value (SEV). Two different calculation ranges of 2.5-180 s and 0.1-180 s were evaluated (SEV1 and SEV2, respectively). A fitting interval of 5-18 s was considered such that the filter effect is negligible . The signal length in the LA task was about 360 s, whereas the signal length in the BA and BUA tasks was about 180 s. To minimize the effect of signal length, 180 s was selected as the upper bound of calculation range for the three listening tasks. Unsupervised Analysis Using Principal Component Analysis Let X ∈ R n×p contains n observations of p EEG features, where could be obtained by concatenating the EEG features per participant, channel, stimulus, and condition in rows. In order to emphasize variation and identify strong patterns in EEG features, a principal component analysis (PCA) was applied on X which is a broad dataset including explicit listening conditions and persons. All power-based EEG features (i.e., absolute and peak powers) were mapped to logarithmic scale (log-transforming) before applying PCA. Since the EEG features do not have the same scales, the data was normalized using zscore transformation such that each column of X re-centered to have zero mean and scaled to have a unit standard deviation. PCA seeks a linear combination of features such that the maximum variance is extracted from the feature. One of the methods of performing PCA is the singular value decomposition (SVD) method. The SVD decomposes X into three matrices, i.e., X = USV T . The PCA results are expressed by two matrices: (1) the PC loadings (coefficients), V ∈ R p×p , can be understood as the weights for each original variable when calculating the principal component; (2) the PC scores (PCSs), Us T ∈ R n×p referring to the representation of X in the PC space, where s ∈ R p×1 is the vector containing the main diagonal elements of S (i.e., the singular values). In other words, each observation in the original space may be projected onto a given PC in order to get a coordinate value along the PC-line. This new coordinate value is known as the PC score. The PC scores are the representation of X in the PC space. In fact, the PC scores can be calculated with X/V T . Grouping the Channels in Subregions The 64 EEG channels were labeled with six fixed subregions: frontal, central, left and right temporal, parietal, and occipital. This division, while allowing four main lobes of cerebrum (Graimann et al., 2010), also considers the central region and left & right hemispheres for the temporal lobe. A similar grouping of channels has been used in previous studies, e.g., for the short-term memory task (Schack et al., 2002). Although subsequent analyses are presented in section 2.8 was performed per channel, the subregion was used a categorical fixed factor in the mixed modeling of EEG-PC scores. However, EEG-PC scores averaged across subregions were used to model the exam result (section 2.9). Statistical Analysis of EEG-PC Scores Linear mixed-effect modeling (LMEM) was used to model the EEG-PC scores as a linear combination of the predictors using the LME4 package (Bates et al., 2015) of the statistical software R (R Core Team, 2019) to explain to origin of EEG-PC scores. The LMEM extends the general linear models (GLMs) to allow both fixed and random effects. A fixed effect is a constant variable across individuals while a random effect varies across individuals. Different LMEMs have been built separately for the nine response variables (EEG-PC scores) as a function of the fixed and mixed |
(random) effects of interest. Since the person-dependent effects may not be captured in the response variables, the participant variable has been considered as a random effect in all the LMEMs. On the one hand, the EEG-PC scores were modeled as a function of task type and channel subregion for each specific background noise type based on formula (3), which is hereafter referred to within-background modeling: In formula (3), PCS j i ∈ R n j ×1 is a vector including ith EEG-PC scores for jth background noise and all 64 EEG channels, where i = {1, ..., 9}, j = {1, ..., 4} and n j is the number of observations belonging to all listening tasks in jth background noise. The symbol "∼" implies that left term is modeled as a function of right terms. The fixed effects include task and subregion. The constant and random terms are expressed in 1 and (1|participant), respectively, where participant is a categorical variable that has 23 possible outcomes. The term task includes the listening task types and has three possible values: lecture attended (LA), background attended (BA), and background unattended (BUA). The last term, subregion, is another categorical variable and has six possible outcomes: frontal, parietal, occipital, central, left, and right temporal. Since for each type of background noise, one model is separately defined, no interaction between task and background noise type can be considered. On the other hand, the EEG-PC scores were modeled as a function of background noise type and channel subregions for each specific listening task based on formula (4), which hereafter referred to within-task modeling: In formula (4), PCS k i ∈ R n k ×1 is a vector including ith EEG-PC scores for kth listening task noise and all 64 EEG channels, where i = {1, ..., 9}, k = {1, 2, 3}, and n k is the number of observations belonging to all background noises tasks for kth listening task. The term background includes the background noise types and takes four possible values: pink (PK), multi-talker (MT), highway (HW), and fluctuating traffic (FT). After estimating the coefficients (intercept and slope) for each fitted model, general linear hypotheses and Tukey posthoc multiple comparisons were then performed to test for the significance of EEG-PC scores changes across the task and background types. For example, we may consider the six pairwise comparisons between the background noises for the fitted model of the first EEG-PC score in the LA task. The question is which specific background's means (compared with each other) are different. A pairwise Tukey's test examines more than one pair of means the same time and corrects for family-wise error rate. Statistical Analysis of Exam Results As mentioned in the section 2.2, we performed a written exam to check the participant's learning during the lecture attended (LA) task. The exam was carried out after all lectures and the attentive listening to background sounds (see Figure 1). Open and closed questions were asked per topic. Open questions were either factual or insight questions. Closed questions consisted of sentences that had to be completed with a specific word or number. The questions were carefully designed so that the answers could be found well-spread over the whole lecture. Over the different topics, the order of question types was randomized. For the open questions, the answers could always be found in three or four connected sentences. The total number of keywords vary per topic. This was deliberately done to capture as closely as possible anything the participants might have recalled, which is important for the EEG analyses (distinguishing between attention and no attention with remembered keywords as ground truth). The topics of the lectures were chosen to avoid prior knowledge by the participants, yet some topics may be more difficult to grasp than others for the average participant. Moreover, there could be small differences in difficulty between the questions. Prior knowledge and logical reasoning of listeners about the answers are not reflected in listening conditions (background sound) nor in the EEG during listening. Therefore, the number of correctly retained keywords was normalized per participant, background noise and topic and the exam z-scores were calculated as follows: where µ pink and σ pink are the mean and standard deviation of the number of correctly retained keywords across all subjects for each topic presented in pink noise (LA in silence), respectively. This a fair reference, as all topics are sufficiently represented in silence. To validate the predictability of exam results by a linear combination of the EEG-PC scores, we used the linear fixed and mixed-effect modeling as explained in the previous section. In fact, the response variable here is the exam z-scores and the EEG-PC scores are considered as the predictors. Moreover, to show that the EEG contains more information than the listening condition, the exam results was also modeled as a function of background noise type and performance was compared to the models based on EEG. Person-dependent differences in the exam results may include the following: mental state, traits, physiological differences, prior knowledge, etc. Some of these differences may reflect in EEG, others may not. Hence it is useful to use both linear fixed and mixed-effect modeling. Linear fixed-effect model regresses the exam results as a function of desired fixed factors without considering participant as a random factor, whereas linear mixed-effect model includes participant as a random effect to capture between-subject variability. The latter implies that a fixed offset in exam results per participant is included in the model. Linear fixed-effect models (LFEM) are expressed in the following formulas: LFEM background ← exam z-scores ∼ 1 + background type, where exam z-scores (as the response variable) were defined by a vector including all exam z-scores computed by Equation (5). PCS Avg ij includes ith EEG-PC scores for jth channel subregion for lecture attended task in all background noises, which were obtained by averaging the PC scores across the channels corresponding to the given subregion (see the section 2.7). The background type term is a categorical variable that has four possible outcomes: pink, multi-talker, highway, and fluctuating traffic. Since 54 EEG-PC scores (the 9 components for each of the 6 subregions) are available to regress exam z-scores, a stepwise regression method can be used to choose the most contributing predictive variables. The backward-elimination approach was applied on both full models (LFEM EEG and LMEM EEG ). To this end, we used "step" function in "STATS" v3.6.2 package of the statistical software R (R Core Team, 2019). This function starts from 54 candidate variables, tests the effect of the deletion of each variable using the Akaike information criterion (AIC) (Akaike, 1974), deletes the variable whose loss gives the least statistically insignificant deterioration of the model fit, and repeats this process until no further variables can be deleted without a statistically significant loss of fit. RESULTS The results consist of two parts: (1) unsupervised analysis of the EEG features observed under different listening conditions (sections 3.1-3.3) and (2) supervised analysis to predict the exam results in lecture attended task (section 3.4). Section 3.1 presents the loading of principal components (PC) on underlying features; section 3.2 demonstrates the scalp topographies of the PC scores; and section 3.3 explains the relationship between PC scores, listening conditions and backgrounds. In the last section, a supervised training of models was used to investigate the predictability of acquiring and retaining performance scores (exam results) by EEG-PC scores. Principal Component Analysis The explained variances by the ten most important principal components in percent are displayed in Figure 3A. Together these ten components explain about 94% of the variability in the dataset. The coordinates of individual EEG feature in principal component (PC) domain are visualized in Figure 3B. The correlation between a feature (variable) and a PC is used as the coordinates of the variable on the PC. The size and darkness of circles in Figure 3B is proportional to the correlation value between an EEG-feature and a given PC. The positive and negative correlation values are visualized by cool and warm colors, respectively. The contribution of EEG feature to the PCs in percentage, i.e., the squared coordinates were normalized to total sum of squared coordinates on the PCs. The larger and darker circles indicate the EEG features contributes more to the given component. The difference between (B,C) is that the (B) shows the correlation between features and PCs, while (C) shows the representation quality of features on the PCs (i.e., normalized squared correlation values in percentage). Figure 3C visualizes the contribution of EEG features to the PCs in percentage. The contribution of ith EEG feature to jth PC is expressed in (y ij ) 2 n j=1 (y ij ) 2 × 100, where y ij is the coordinate of ith EEG feature on jth PC and n = 10 is the number of PCs. In fact, in Figure 3C, the squared coordinates were normalized to total sum of squared coordinates on the PCs. The squared coordinates can be a quantity to measure the quality of representation of the features on PC domain. Each topographic map has been obtained by averaging EEG-PC scores across participants and fragments per listening condition and EEG channel. Each column belongs to one specific listening condition. The type of background noises is shown above the corresponding columns. The warm and cool color-coded areas represent the positively and negatively correlated cortical areas with the extracted components, respectively. Red frames show some spatial and activation differences suggesting the EEG-PC scores might contribute to statistical significance in the discrimination between the three listening tasks and the three background noises. Specifically, (i) PCS4 is higher in BA compared to BUA, (ii) PCS5 is the maximal and minimal in BUA and LA, respectively, (iii) PCS 6 is the maximal for highway in LA and multi-talker in BA, and (vi) PCS7 is the maximal for multi-talker in LA and BUA. As can be seen from Figures 3B,C, the different features contribute to each component. Accurate grouping of these PCs is not possible due to presence of different positively and negatively correlating features with the PC scores ( Figure 3B). It is worth noting that normalized version of squared coordinates ( Figure 3C) shows that the last five PCs have more specific loading (representation quality) than those of first five PCs. Specifically, the long-range temporal correlations of alpha band and frequency information of gamma and beta bands are most contributing features to represent PC domain. Scalp Topographic Maps For visualization across the scalp, 2D topographic maps of the component scores are shown in Figure 4. The topographies of the nine first PC scores (PCS i , i = 1, ..., 9) were obtained by averaging across all subjects and the specific fragments for each listening conditions. In fact, for cth EEG channel, jth listening task, and kth noise, the average value of ith PC scores was calculated using PCS Note that here we do not aim at reporting the statistical differences between the listening conditions in terms the PC scores. However, some spatial and activation differences can be observed between different listening conditions (shown by red frames in Figure 4) suggesting the EEG-PC scores might contribute to statistical significance (refer to section 3.3) in the discrimination between the three listening tasks and the three background noises. Each component is a linear combination of different positively and negatively correlated features with the components (refer to Figure 3). Therefore, in Figure 4, both of the warm and cool color coded areas are important, which represent the positively and negatively correlated cortical areas with the extracted components, respectively. The qualitative differences of some components between different conditions have been shown by red frames in Figure 4. Specifically, the PCS 5 is the lowest in the LA compared to other tasks, the PCS 6 is the highest in the highway during the LA, and the PCS 7 is the highest in the multi-taker for the three tasks. In addition, these topographies indicate that different PCs contribute to different cortical areas. For example, the third PC score is positively dominant over temporal and occipital regions. Explainable Origin of EEG-PC Scores The unsupervised extraction of PCs from our dataset implicitly attempts to discriminate between participant, listening task (LA, BA, BUA), and background (MT, HW, FT). One way to analyse |
the origin of a PC is to construct a regression model for its score based on the above-mentioned factors as explained in section 2.8. A constant mixed-effect model for predicting EEG-PC scores is expressed in [PCS i ∼ (1|participant)], where PCS i is ith PC score for all listening conditions and channel subregions. If channel subregion is added as a fixed factor to the constant model, the new model could better predict all PC scores (p < 10 −15 ) compared to the constant model. By adding listening task type to the current model, all PC scores except the sixth PC score are better predicted (p < 10 −8 ). Background noise type as an additional fixed effect could improve the current model for all PC scores except the ninth PC score (p < 0.05). By adding interaction between background noise and task types, the improvement of current model is significant for all PC scores (p < 10 −4 ) except the ninth PC score. Since the interaction between task and background noise type significantly improves modeling EEG-PC scores, its effect was separately investigated using two distinct models, within-background and within-task modeling based on formulas (3) and (4), respectively. Tukey post-hoc multiple comparison for within-background and within-task models were reported in Tables 1, 2, respectively (refer to section 2.8). Each test in the sub-matrices was run independently. For example, for a particular background type and PC score, the listening conditions are compared. In each 4 × 4 and 3 × 3 sub-matrices, upper triangular elements denote p-values of significant differences for corresponding comparisons, lower triangular elements denote which noise or task results in higher values of the given EEG-PC score and main diagonal elements denote which background noise or task results in the maximum/minimum values of the given EEG-PC score. For example, in Table 2, MT PCS 1 is significantly higher than that of PK during LA because e 1,4 and e 4,1 elements of matrix corresponding to LA and PCS 1 are < 0.001 and an arrow directed toward MT, respectively. Note e i,j represents the element at the ith row and jth column of the sub-matrix. In within-background modeling (see Table 1), the first PC score is (significantly) the highest and lowest in the BUA and BA tasks for all the background noises, respectively. Moreover, the BUA task has the highest PCS 3 values compared to other tasks for all background noises. For the MT background noise, the LA has the highest PCS 2 compared to other tasks. The fourth PC score exhibits significant contrast between background attended and other tasks for all background noises. The lecture attended can be discriminated from other tasks for all background noises by the fifth PC score. The sixth PC score has a significant contrast between LA and BA tasks in the MT background noise. The seventh PC score is the highest for the BUA compared to other tasks in the MT and HW noises. For all the background sounds PCS 8 is consistently minimal in the BA task. Finally, the ninth PC score is the maximal and minimal for the BA task in the MT and FT sounds, respectively. In within-task modeling (see Table 2), the MT has the highest PCS 1 compared to other background noises in the LA task, whereas in the BUA task, the MT has the lowest PCS 1 . The second PC score in the HW is significantly lowest compared to other noises in the LA task. The third PC score exhibits only significant differences in the BA and BUA tasks. The background sounds can be discriminated by the fourth PC score in the LA and BA tasks. The fifth PC score has the highest values in the HW noise during the LA task compared to other background noises. The sixth and seventh PC score are significantly able to distinguish the background sounds for all the listening tasks. The eighth PC score exhibits the highest value for the MT and HW in the LA and BA tasks, respectively. The ninth PC score is not very capable of distinguishing the background sounds. Remark 1: The statistical results reported in Tables 1, 2 have been obtained by eliminating the person-dependent effects, while in the previous section, the topographic maps (Figure 4) were obtained by averaging across all subjects without eliminating the person-dependent effects. As a result, the differences are seen in Figure 4 are not only due to differences between tasks and between noises (like Tables 1, 2) but also due to differences between participants. This means that some of the differences seen in the tables and the topographies are not comparable due to the presence of the effect of the changes between individuals. For example, in the highway noise, although the second PC scores of the BUA task are qualitatively lower than other tasks based on Figure 4, Table 1 shows only the dominance of the LA over the BUA. To explain this difference, we performed Tukey's post-hoc testing of linear fixed-effect modeling (without participant as a random factor). The post-hoc test revealed that BUA<BA (p < 10 −5 ) and BUA<LA (p < 10 −5 ) meaning that the second PC score can be affected by individual differences likely due to the wideband power (1 − 45 Hz) contributing to this component. Remark 2: Referring to section 2.8, in Tables 1, 2, the results were shown for a model also including the subregions. This implies that a statistically significant difference in one subregion is sufficient for obtaining significant differences. In the maps of Figure 4, the reader is expected to interpret the differences in this way. However, the effect of different subregions were separately investigated to model the exam results in the LA task (refer to section 2.9 and 3.4). Predictability of Exam Results in Lecture-Attended Task As noted in section 2.9, the exam z-score defined in Equation (5) ) to quantify the amount of information that participants have actually acquired and retained from the lectures. To normalize the exam results (the number of correctly retained keywords) and find the exam z-scores, the exam results of a lecture-attended task in pink noise (lecture in silence) were used. Figure 5 visualizes the number of correctly retained keywords for lecture attended task in pink noise across thirteen topics. Mean and standard deviation values (µ pink and σ pink in Equation 5) are shown by circles and triangles, respectively. The boxplots display the median marked as a bold line. The lower and upper whiskers represent another 50% data distributed outside the interquartile box. As can be seen from Figure 5, the number Tukey variable is the task type (LA, BA, and BUA) and all possible pairs of means in each subtable are compared. Significant p-values are reported in upper triangular. Main diagonal denotes which task is the maximum (Max) and the minimum (Min) compared to other tasks in terms of a given PC score. Lower triangular arrows are directed toward the tasks which have higher PC scores, when comparing two tasks. The non-significant (p >0.05) differences are shown by dash signs. The type(s) and frequency band(s) associated with each component (using the features have the strongest impacts; refer to Figure 3C) are reported in the first column. Each element of the sub-matrices is corresponding to the listening tasks labeled above and to the left side: LA, lecture attended, BA, background attended, BUA, background unattended; Each sub-matrix is corresponding to the background noise type: MT, multi-talker; HW, highway; FT, fluctuating traffic and EEG principal component scores (PCS i ) labeled above and to the left side, respectively. of retained keywords in silence for different topics are different, and hence, the difficulty of retaining information in each topic is different. In order to assess the effect of background noise type on predicting the exam z-scores, the exam z-scores were modeled using formula (10) and then, Tukey post-hoc multiple comparison testing was used to compare the background noise types. The statistical results are reported in Table 3. As can be seen, there are significant differences between pink and multitalker, between pink and highway, and between fluctuating traffic and multi-talker noises. This means that the exam z-scores are higher in the pink noise (LA in silence) than those of in the multi-talker and highway background noise (as we expected). In addition, the fluctuating traffic background noise leads to the higher exam z-scores compared to those of the multi-talker background noise. Therefore, compared to the fluctuating traffic noise, the multi-talker noise leads to more difficult condition for information retention. Note that pink noise refers to a very lowlevel pink noise (see section 2.2) and means that subjects have listened to the lectures in silence. Tukey variable is the background noise type and all possible pairs of means in each subtable are compared. Significant p-values are reported in upper triangular. Data can be decoded like Table 1. The type(s) and frequency band(s) associated with each component (using the features have the strongest impacts; refer to Figure 3C) are reported in the first column. Each element of the sub-matrices is corresponding to the background noise type labeled above and to the left side: MT, multi-talker; HW, highway; FT, fluctuating traffic; Each sub-matrix is corresponding to the listening tasks: LA, lecture attended, BA, background attended, BUA, background unattended and EEG principal component scores (PCS i ) labeled above and to the left side, respectively. TABLE 3 | Effect of background noise on exam z-score; Tukey post-hoc multiple comparisons between different types of background noise for modeling exam z-score in lecture attended task (using mixed-effect modeling). Background type Pink Highway Multi-talker Fluctuating traffic To identify the link between EEG-PC scores and the exam z-scores, both fixed and mixed-effect models were employed as presented in section 2.9. Note that the EEG-PC scores used in this section were obtained by averaging across the channels corresponding to the given subregions. The models were compared using two criteria. First, χ 2 test was used to compare between the two models using "anova" function in STATS v3.6.2 package of the statistical software R (R Core Team, 2019). A good model not only needs to fit data well-it also needs to be parsimonious. This criterion takes the model objects as arguments and returns an ANOVA testing whether or not the more complex model is significantly better at capturing the data than the simpler mode. If the resulting p-value is <0.05, we conclude that the more complex model is significantly better than the simpler model. If the p-value is >0.05, we should favor the simpler model. The second criterion used to compare the fitted models was the Akaike information criterion (AIC) (Akaike, 1974). When comparing models fitted by maximum likelihood to the same data, a lower AIC value indicates a better fit. We have used "extractAIC" function in STATS v3.6.2 package of the statistical software R (R Core Team, 2019). The following equation is used to estimate AIC: −2 log(L)+(k×edf ), where k = 2, L refers to the likelihood, and edf stands for the equivalent degrees of freedom (i.e., the number of free parameters for the models) of fit. Table 4A reports the predictability of exam z-scores based on linear fixed-effect modeling (without considering participant as a random factor). The following predictors (fixed factors) were used: (1) no fixed factor (constant), (2) background type, and (3) 54 EEG-PC scores as defined by formulas (6), (7), and (8), respectively. Furthermore, a stepwise fixed-model regression was performed to regress exam z-score using the most significant EEG-PC scores (refer to section 2.9). P-values shown on the Upper triangular elements are pairwise p-values (p) when two models are compared using χ 2 test (if p <0.05). Lower triangular arrows are directed toward the better models when comparing two models. If the resulting p-value is <0.05, the more complex model is significantly better at capturing the data than the simpler model. If the p-value is > 0.05, we favor the simpler model. Main diagonal elements indicate AIC values for given models. The lowest AIC value (corresponding to the best model) is shown in bold. The PC scores obtained by the stepwise method are reported below the tables and (•) denotes the regression coefficient (slope) of each factor. upper diagonal of Table 4A, suggest that there are pairwise |
significant differences between all models except between two models which use 54 EEG-PC scores and stepwise EEG-PC scores as the fixed factors. This means that the stepwise model (simpler model) is better than the full model (more complex model) in terms of χ 2 test criterion. AIC values shown on the main diagonal of Table 4A, suggest that stepwise EEG-PC scores can predict the exam z-scores better than other models (the lowest AIC). We found the 23 predictors that play more significant roles to predict the exam z-scores. The names of these predictors, their p-values (to predict the exam zscores), and their coefficient (slope in regression) are reported below Table 4A. They were ordered according to their statistical significance. As can be seen from Table 4. the parietal PC score 7 (related to alpha LRTC), which is negatively correlated with exam z-scores, is the most important predictor to model the exam z-scores using the linear fixed-effect modeling. The results of the mixed-effect models (formulas 9-11) to model exam z-scores are presented in Table 4B. By including the participant as a random factor, the models are less likely to be affected by individual differences. Therefore, those EEG features that contribute to differentiate between participants are expected to be less relevant in this modeling. In contrast to the fixed-effect model, in the mixed-effect model, background noise type can better predict the exam z-scores compared to 54 EEG-PC scores (in terms of AIC and not χ 2 test). However, the stepwise EEG-PC scores results in a significantly better model than knowing background noise type to predict exam z-scores (the lowest AIC). According to the tables, in the both fixed and mixed-effect models, the modes which use stepwise EEG-PC scores predict the exam z-scores better than all other models. It is worth noting that unlike the fixed-effect model which all the components in the certain subregions play the significant roles in predicting, in the mixed-effect model, the most contributing predictors are limited to the PCS 1, 2, 4, 5, 7, and 8 in the particular subregions (as can be seen from below Table 4B). These results are consistent with the results of section 3.3, where the importance of these components (especially PCS 7) to distinguish between the background noises in the lecture attended task was shown (refer to Table 2). The relationship between these components and hypotheses presented in the introduction section and their underlying mechanisms will be discussed in the next section. DISCUSSION The present study used a single-trial 64-channel EEG measurement and ecologically valid stimuli to investigate the neural correlates of acquiring and retaining vocally presented information. To identify significant EEG components, a broad set of three listening tasks were performed: (1) attentive listening to 5-min lectures in the environmental sound (LA), (2) attentive listening to environmental sounds (BA), and (3) inattentive listening to environmental sounds (BUA). The environmental sounds included multi-talker, highway, and fluctuating traffic sounds. During this unsupervised learning step, a wide range of features of sensor-space EEG signals were collected and their principal component scores (PCSs) were calculated. Unlike the attention decoding studies that aim to explicitly decode an attended from unattended speech stream based on the supervised approach (Horton et al., 2014;O'Sullivan et al., 2014), we aimed to distinguish between attentive and inattentive listening conditions. To this end, we used an unsupervised learning method that, as such, did not require knowledge of the attended sound signal. During the LA task, the mixture of verbal lectures and different types of background noise were presented. The lectures were related to topics for which prior knowledge is expected to be minimal. A written exam was taken after the experiment to quantify the amount of information that participants have acquired and retained from the lectures. Since the exam included the questions related to fact and insight, memory is expected to be more specifically involved. It is worth noting the following: (1) although the background sounds could distract the participants while listening to the speech, they did not mask the speech energetically, and (2) no visual distractor was presented during the experiment. Essential EEG-PC Scores to Predict the Exam Results The predictability of exam results of the LA task by the EEG-PC scores (EEG-PCSs) has been assessed by linear fixed and mixed-effect modeling of the exam z-scores. It is expected that differences in the exam z-scores can arise from the instantaneous listening state but also from the overall state, personal traits, physiology, and prior knowledge, hence both fixed and mixedeffect models were used to regress the exam z-scores. The fixedeffect model, not considering participant as a random factor, assumes that all relevant differences for predicting exam zscores are visible in EEG-PCS, whereas the mixed-effect model, considering participant as a random factor, assumes some personal differences are not visible in the EEG-PCS. We first consider the latter approach. Firstly, it could be confirmed that knowing the type of background sound improves the predictability of exam zscores (refer to Table 4). Exams on information presented in background noise always gave significantly lower scores, except for fluctuating traffic noise that did not seem to significantly affect exam z-scores (refer to Table 3). Note that in our experiment, noise may affect speech perception, listening comprehension, distraction, and memory encoding. Speech perception in noise was found to be consistently worse in babble than in traffic noise in previous research (Shukla et al., 2018). For episodic memory tasks, it was found that encoding under traffic noise and meaningful irrelevant speech were worse than under silent conditions, but scores were lower for traffic noise than for competing meaningful speech (Hygge et al., 2003). Thus, our results seem to confirm previous works. We can now turn to the question of whether EEG allows us to disentangle the multitude of interacting effects that play a role. A stepwise mixed-effect model identified that a few specific EEG-PCSs play a more significant role in modeling the exam z-scores (refer to Table 4B). These EEG-PCSs are the central, occipital, and left temporal PCS 1, the occipital and left temporal PCS 2, the parietal PCS 4, the central and parietal PCS 5, the parietal PCS 7, the central, and parietal PCS 8. The underlying mechanisms of these components and their links with our hypotheses are discussed based on the unsupervised learning phase and the previous studies as follows. • The first component: overall attentive state In general, the alpha-band activity has been assumed as an idling rhythm (Pfurtscheller et al., 1996) meaning the power of alpha activity increases during resting state and conditions of mental inactivity. During the cognitive effort, alpha activity usually diminishes, which is referred to as alpha desynchronization (Pfurtscheller and Da Silva, 1999;Sauseng et al., 2005). In addition, previous studies have argued increased occipital (task-irrelevant) and decreased frontal (taskrelevant) alpha activity can reflect the distracted auditory attention (Pfurtscheller and Da Silva, 1999;Sauseng et al., 2005;Clayton et al., 2015). Our results showed the occipital and PCS 1 is negatively correlated with the exam z-scores (p = 0.02, s = −0.17, where s is the slope of corresponding factors in the linear regression). Based on the results yielded by PCA (refer to Figure 3), the alpha peak power and alpha bandwidth are the most positively and negatively contributing feature to this component, respectively. Therefore, an increase in the exam z-scores can be associated with a decrease in this component score due to overall mind wandering and distracted attention. This statement is in accordance with the unsupervised analysis results where the multi-talker and pink (lecture in silence) PCS 1 is the maximal (the least attention) and minimal (the highest attention) compared to other background sounds during the lecture attended task (see Table 2). In addition, the ratio of theta to alpha power (RPTA in Figure 3) also positively contributes to this component which also confirms that an increase in PCS 1 indicates the deterioration in attention (in agreement with Holm et al., 2009;Borghini et al., 2014). • The fourth component: low-frequency speech envelope following The parietal PCS 4 is negatively correlated with the exam z-scores (p = 0.027, s = −0.12). The fourth PC is strongly determined by various characteristics of the delta frequency band, such as bandwidth, central frequency, and spectral edge frequency (refer to Figure 3). This frequency band is observed during speech envelope following (Kerlin et al., 2010;Vanthornhout et al., 2019). In addition, the gamma central frequency and the alpha-band LRTC negatively contribute to the fourth PCS and are visible in the occipital, temporal, and parietal regions (see Figures 3, 4). The unsupervised analysis revealed that the fourth PCS exhibits the highest and lowest values in background attended and unattended tasks, respectively (see Table 1). Therefore, the fourth PCS may reflect speech envelope following and listening attentively without necessarily linguistic processing or gating out (our third hypothesis). This interpretation could be consistent with the lower values (more negative values) of the parietal and occipital fourth PCS in fluctuating traffic noise compared to other background noises in the lecture attended and background unattended tasks (refer to Figure 4). • The fifth component: decreased focusing during listening The parietal fifth PCS exhibits a reverse relationship with the exam z-scores (p = 0.020, s = −0.38). The positively contributing EEG features to the fifth PCS include the beta central frequency and the gamma absolute power. Based on the unsupervised analysis, the fifth PCS is the lowest in the lecture attended (LA) task compared to other tasks for all background noises (refer to Table 1). Therefore, decreased fifth PCS is likely associated with more focus during listening, where the exam scores are expected to improve as well. • The sixth component: cognitive prediction error Although the sixth PCS is not obtained from the mixedeffect stepwise regression as a contributing component, the left temporal PCS 6 is the most significant component obtained from the full mixed model (p = 0.02, s = 0.55). The sixth PCS positively loads on the gamma spectral edge frequency, bandwidth, and central frequency. Moreover, the frontal and central sixth PCS is negatively correlated with the exam z-scores (s = −0.10 and s = −0.20). Based on the unsupervised analysis, the sixth PCS is more discriminating between the background noises. Its highest values are observed for attended speech in continuous highway sound (LA-HW) and for attended multitalker sound (BA-MT) (refer to Table 2). These two conditions have in common that one may rely on linguistic processing and prediction to complete the information. This factor is therefore likely associated with predictive coding. Higher values of the sixth PCS result in lower exam z-scores which may be explained by the fact that a need for prediction to complete the information may result in poor encoding. This finding is in line with Bastos et al. (2012), Sedley et al. (2016), and Alexandrou et al. (2017) where has been shown the prediction violations or errors (our fifth hypothesis) are encoded by gamma-band activity (especially over higher brain areas). It was also found that this component over the left temporal region is positively correlated with the exam z-scores reflecting task-relevant gamma-band activity role on speech processing in alignment with Giraud et al. (2007), Morillon et al. (2012), and Alexandrou et al. (2017). • The seventh component: alpha-as-inhibition and inhibitionexcitation balance The parietal seventh PCS, which positively loads on alphaband LRTC, is negatively correlated with the exam z-scores (p = 8 × 10 −6 , s = −0.37). Interestingly, this PC score in multi-talker noise and independent of task type is significantly dominant compared to other background noises. Increased alpha-band LRTC reflects that the autocorrelations of alpha activity slower decay in power-law behavior and as a result, the self-similarity of alpha activity increases. In fact, high levels of alpha-band LRTC reflect the enduring alpha waves. In agreement with Poil et al. (2012), this increased self-similarity or longlasting changes could reflect more balance between excitation and inhibition states of alpha-band activity during the auditory stimulus (our second hypothesis). Both excitation and inhibition sates are therefore involved during attentive listening to the lecture in multi-talker sound, which is required for more listening effort due to multi-talker distraction. In contrast to multitalker, attentive listening to lectures in pink noise (lecture in silence), the alpha-band LRTC is the lowest compared to other noises due to less need for inhibition during listening. In |
fact, during this listening condition, the excitation state is more dominant than the inhibition state. Increased PCS 7 could thus be associated with a higher inhibition-excitation balance. This component can be linked to the alpha-as-inhibition (Clark, 1996;Uusberg et al., 2013) hypothesis (our first hypothesis) where alpha synchronization reflects suppression of irrelevant information (inhibition). For the fixed-effect model, where all differences between people are assumed to be explainable through EEG, also adding second (over non-occipital regions), third, sixth, and ninth PCSs improves the predictability of exam results (refer to Table 4A). The second PCS loads strongly on the wide-band absolute power and absolute powers in the low-frequency bands (delta and theta). It is probably related to the observability of EEG for each specific person and may not indicate specific brain-related functions. The third PCS mainly loads on alpha peak frequency, alpha central frequency, and related factors. As PCS 1, the third PCS is significantly higher in the BUA task. Literature is not univocal on the expected trends in relation to tasks (Angelakis et al., 2004;Mierau et al., 2017) but points at a significant difference between persons (Klimesch et al., 1993;Haegens et al., 2014). The latter may explain why PCS 3 only occurs as a significant predictor in the fixed-effect model where it helps to differentiate between persons. EEG-PC Scores Related to Task Difficulty-Based Cognitive Load In this experiment, adding background sound to the lectures increases the effort needed to process the sound, but it may also affect cognitive load and task difficulty. The cognitive load of subjects has been assessed from different perspectives using EEG depending on the type of task. For instance, the task difficulty during the intelligence test (Friedman et al., 2019) and learning task (Mills et al., 2017) as the cognitive load has been linked to EEG features. Moreover, the cognitive load during a visual task has been associated with the attentional demand using an ERP analysis (Grassini et al., 2019). There is no unique EEG feature that is directly related to cognitive load. Theta power has been suggested as an indicator for the average cognitive load of subjects and the linguistic complexity of educational videos (Castro-Meneses et al., 2019). Mu rhythm oscillations (8 − 13 Hz over the sensorimotor cortex) could be affected by the cognitive load during speech perception due to attention and working memory processes (Jenson et al., 2019). In addition to the task difficulty, the listener's skill also may affect the cognitive load. In this paper, although the cognitive load of listeners has not been explicitly investigated, some PCSs may reflect the task difficulty-based cognitive load, such as the sixth and seventh PCSs (reflecting the prediction error and the inhibition during listening, respectively). However, caution is needed to link neural results to these behavioral outcomes as this study is based on a sample of young adults only. Aging populations might react differently. Since there are more noiseless gaps during fluctuating traffic sound compared to the highway sound (refer to Figure 2), it is expected that less mental resources are needed to predict the missing part (less PCS 6) during LA in fluctuating traffic sound. Therefore, LA in the highway sound (LA-HW) is likely more difficult task compared to LA in the fluctuating traffic sound (LA-FT). However, the task difficulty can be reflected either in the continuous inhibition by increased PCS7 (highway sound) or in the fluctuating inhibition by decreased PCS7 (fluctuating traffic sound). Moreover, in the BUA task, the fluctuating traffic sound is the most difficult sound to predict (the highest PCS6) compared to the other sounds. Although the BUA in the multi-talker sound exhibits more inhibition compared to the fluctuating traffic (higher PCS7), the multi-talker sound in the BUA can be easier predicted (lower PCS 6) compared to the fluctuating traffic sound. These findings may explain the impacts of different types of environmental sound during daily activities. CONCLUSION The current study showed that it is possible to predict beyond the chance level the amount of vocal information that participants acquire and retain from the lectures presented in different environmental sounds using 64-channel EEG. Five principal component scores of the EEG features obtained under different listening conditions and for different persons were essential for this prediction. Based on their loading on the spectral range and their ability to distinguish between listening tasks, we associate them with overall attentive state, speech envelope following (listening attentively without necessarily linguistic processing), focusing during listening, cognitive prediction error, and specific inhibition. Part of the variance between persons could further be explained by principal component scores that tend to relate to overall signal strength, an indication of observability of EEG signals, and person identification through inter-individual differences between typical alpha peak frequencies. Inhibition-excitation balance (reflected by alpha-band representation) and predictive mechanisms (reflected by gammaband representation) play a more important role than might have been expected and could be observed via EEG. Furthermore, the results of comparing the principal components scores of three different auditory tasks (attentive listening to the lecture in environmental noise, attentive listening to the environmental sound, and inattentive listening to the environmental sound) showed the extracted principal components scores are able to discriminate the different listening tasks and background noises. Specifically, (i) the sixth and seventh principal component scores, which reflect prediction error and inhibition-excitation balance, respectively, allow us to distinguish different types of background sound. Moreover, (ii) the type of listening tasks could be completely distinguished by the first and fifth principal component scores, which reflect the overall attentive state and decreased focusing, respectively. In terms of methodology, by combining different listening conditions to train in an unsupervised way the definition of orthogonal features based on EEG, a more efficient supervised model for the prediction of the memorization of information could be obtained. This methodology could be relevant for assessing the impact of environmental sounds on daily activities, such as communicating, learning, and relaxing as some of the principal components identified could be related to increased cognitive load. They could also be relevant for future artificial intelligence communicating optimally with humans based on observed brain activity. The methodology also allows us to assess individual differences in the ability to process speech in noise. DATA AVAILABILITY STATEMENT The datasets presented in this article are not readily available because further analysis is ongoing. Requests to access the datasets should be directed to the first author. The Matlab R and R R codes implementing the algorithms and statistical analyses are publicly accessible on GitHub (https://github. com/EhsanEqlimi/EEG-Correlates-of-Learning-From-Speech-Presented-in-Environmental-Noise). ETHICS STATEMENT The studies involving human participants were reviewed and approved by International Laboratory for Brain, Music and Sound Research (BRAMS), Montreal, Canada. The participants provided their written informed consent to participate in this study. Written informed consent was obtained from the individual(s) for the publication of any potentially identifiable images or data included in this article. AUTHOR CONTRIBUTIONS EE carried out the data analysis and interpretation, signal processing, statistical analysis, and writing of the manuscript. AB carried out the data acquisition, the experiment design, study idea, statistical analysis, data interpretation, and editing of the manuscript. BD carried out the data interpretation and the editing the manuscript. MS carried out the experiment design, data acquisition and interpretation, and editing of the manuscript. DT carried out the data interpretation and editing of the manuscript. DB carried out the original idea for study, data interpretation, experiment design, and editing of the manuscript. All authors contributed to the article and approved the submitted version. FUNDING This study was part of DUCK project (Distraction from learning by Unrelated auditory events assessed by Computational modeling and Knowledge extraction from single-trial electroencephalography). This research received funding from the Flemish Government under the Onderzoeksprogramma Artificiële Intelligentie (AI) Vlaanderen programme and the Belgium Special Research Fund (Bijzonder Onderzoeksfonds, BOF). KinetochoreDB: a comprehensive online resource for the kinetochore and its related proteins KinetochoreDB is an online resource for the kinetochore and its related proteins. It provides comprehensive annotations on 1554 related protein entries in terms of their amino acid sequence, protein domain context, protein 3D structure, predicted intrinsically disordered region, protein–protein interaction, post-translational modification site, functional domain and key metabolic/signaling pathways, integrating several public databases, computational annotations and experimental results. KinetochoreDB provides interactive and customizable search and data display functions that allow users to interrogate the database in an efficient and user-friendly manner. It uses PSI-BLAST searches to retrieve the homologs of all entries and generate multiple sequence alignments that contain important evolutionary information. This knowledgebase also provides annotations of single point mutations for entries with respect to their pathogenicity, which may be useful for generation of new hypotheses on their functions, as well as follow-up studies of human diseases. Database URL: http://lightning.med.monash.edu/kinetochoreDB2/ Introduction During cell mitosis and meiosis, the kinetochore plays a critical role of locating the attachments on chromosomes and pulling sister chromatids apart. Assembled on centromeric chromatin, kinetochore functions during the cell cycle (1)(2)(3)(4)(5)(6). During the last few decades, numerous studies of the kinetochore and its related proteins have characterized its function, architecture and the repertoire of its related proteins using biochemistry, structural biology and cell biology techniques (4,(7)(8)(9)(10)(11). Both the stability of the kinetochore-microtubule interface and mutations occurring in the kinetochore and its related proteins are associated with a number of human diseases (12)(13)(14)(15). Dynamics studies of the kinetochore have also shown that deregulation of the kinetochore-microtubule dynamics frequently results in chromosome instability, leading to the development of cancer (10,11,16). Other experimental studies reveal that mutations of the kinetochore and its related proteins are closely linked to human diseases. For example, the adenomatous polyposis coli protein, localized in both centrosome and kinetochore, contains 15 disease-associated mutations that cause familial adenomatous polyposis (14,15) and Medulloblastoma (12). Despite its biological significance and our increasing awareness of its potential roles in human diseases, there is currently a paucity of publically available databases or resources that focus on providing comprehensive functional annotations of the kinetochore and its related proteins. The only available database is MiCroKiTS (17), an integrated online resource for kinetochore, midbody, telomere, centrosome and spindle proteins. However, important annotations of entries in MiCroKiTS are not available in terms of protein 3D structure, protein interaction partners, metabolic/signaling pathways etc., all of which are crucial aspects for follow-up functional studies of these proteins. In an effort to address this knowledge gap, we created KinetochoreDB, which integrates several public databases, computational annotations and experimental results for currently 1554 related entries. KinetochoreDB contains several important features, the majority of which are not available in MiCroKiTS (Table 1): i. It provides annotations of protein 3D structure when structural information is available. For protein entries with available structural information, the corresponding Protein Data Bank (PDB) IDs and their related information are provided. In addition, information on predicted intrinsic disorder is provided, which is particularly important for providing structural insights into those entries in KinetochoreDB whose 3D structures have not been solved. ii. It provides comprehensive annotations of single point mutations and their pathogenic effects. These mutations are classified as either pathogenic or nonsense mutations in KinetochoreDB. For disease-associated pathogenic mutations, KinetochoreDB provides details of the disease caused by the mutation, allows users to search the entire database with the disease name of interest, and provides user-friendly options to browse the related kinetochore proteins that harbor such disease-associated mutations. iii. It provides metabolic/signaling pathway information for each entry by cross-referencing the KEGG database. Such information is important for understanding the functions of kinetochore proteins from a biochemical network perspective. In particular, the pathway information and the link to Kyoto Encyclopedia of Genes and Genomes (KEGG) will be provided if an entry has pathway information available in KEGG. iv. It provides multiple sequence alignments (MSAs) for all included entries, thereby allowing users to readily identify evolutionarily conserved regions within the family of a kinetochore protein. In addition, the visualization of MSAs implemented by Jalview is userfriendly and customizable. v. It provides convenient user enquiry and new entry submission options by enabling users to automatically upload their newly discovered sequences into the online database. Materials and Methods for Database Construction We define 'kinetochore and its related proteins' with respect to protein subcellular location and Gene Ontology (GO). The entries of KinetochoreDB originate from three major resources; QuickGo database (18), UniProt database (19) and MiCroKiTS, and the database was populated as follows. From MiCroKiTS, we obtained data |
entries that have been experimentally verified to be located in kinetochore. By searching GO terms from QuickGo with the keyword 'kinetochore', we obtained 64 GO terms related to kinetochore. For each GO term, we searched and filtered the reviewed entries from the UniProt database to ensure that all the downloaded entries contain the GO annotation. Applying this procedure resulted in 53 GO terms including 25 cellular component terms, 2 molecular function terms and 26 biological process terms (Supplementary Table S1). In addition, we queried 'subcellular location' with the keyword 'kinetochore' in UniProt and downloaded the entries with published experimental evidence from the search results. After the removal of redundant entries, the resulting dataset contained a total of 1554 carefully reviewed entries. The detailed procedures of database construction and data collection are illustrated in Figure 1 and a statistical summary can be found in Figure 2 and Table 2, respectively. For each entry, KinetochoreDB integrates several public resources, including the UniProt database, NCBI Gene database (http://www.ncbi.nlm.nih.gov/gene/), 1000 Genome Project database (20) (for human proteins), BioModels database (21), Research Collaboratory for Structural Bioinformatics (RCSB) PDB (22), Online Mendelian Inheritance in Man (23), BioGRID 3.2 (24), Pfam database (25) and KEGG (26), in order to provide a comprehensive description with respect to basic protein information, protein structure, function, mutation and evolutionary conservation. An important feature of KinetochoreDB is the provision of 3D structure. To achieve this, we manually searched all the entries against the PDB database using their corresponding UniProt identifiers and protein names. For protein complex structures, we identified the PDB chain for each entry and annotated the entry with that chain. In addition, we also generated MSAs using all homologous sequences for each protein entry. Homologous sequences were retrieved by PSI-BLAST (27) search against the Swiss-Prot dataset obtained from UniProt. The alignments were generated using Clustal Omega (28). We also predicted natively disordered regions for all protein entries using one of the most widely used disorder predictors, namely VSL2B (29). A residue is annotated as disordered by VSL2B if its prediction score was >0.5. We used Jmol (http://jmol.sourceforge.net/) and pViz (30) for visualization of protein structures, and Jalview (31) for customizable editing and display of MSAs for each protein entry. In order to provide the annotations of the domain context including functional domains and sites for each protein entry in KinetochoreDB, another Javascript based plug-in, The Protein Feature View (http://andreaspr lic.github.io/proteinfeatureview/), was used to display protein structural, functional domains and sites in a better and more interactive way. The display of protein domain context can be found in 'Protein Domain Context' section of the webpage. The information stored in KinetochoreDB resides in a MySQL relational database. A highly interactive web front-end to the data was implemented using the Javascript framework, JQuery. Apache Tomcat handles serving of data to users on the web, utilizing a set of Java Servlets and JavaServer Pages for data searching and viewing. Database Utility The 'Search' page (http://lightning.med.monash.edu/ kinetochoreDB2/Search.jsp) ( Figure 3) allows users to search the database in several different ways. These search options can be generally classified into two groups: search with ID or search with keywords. Examples are provided below to assist users to understand how to perform the search. When searching the database with IDs, KinetochoreDB provides two different kinds of IDs to facilitate the search: UniProt ID and KinetochoreDB ID. The latter is composed of 'KD' and five digits, e.g. KD00095. Considering that the database includes 1554 entries in total, each entry is accordingly numbered as KD00001-KD01554. In addition KinetochoreDB offers alternative search options with keywords. These include protein name, kinase name, post-translational modification (PTM) type, name of protein interaction partner and name of diseases caused by single point mutations ( Figure 3B). After selecting the 'Submit' button, the corresponding search results will be shown on the webpage. For each entry, there are generally 10 sections of structural and functional categories, including general information, protein domain context, protein structure, disordered regions prediction, To provide an illustration of the annotations for each entry in KinetochoreDB, we use 'UniProt ID ¼ O14965' (KinetochoreDB ID ¼ 'KD01531') as an example query. The resulting page with ten sections is shown in Figure 4. 3D structures. A single structure for the current protein entry can be examined by clicking the 'View' button to launch Jmol, a Java applet for displaying 3D structures (Supplementary Figure S1C). Multiple structures can also be viewed together as an ensemble using pViz (Supplementary Figure S1B). With respect to protein-protein interaction, the interaction partner is highlighted if this protein is also an entry of KinetochoreDB. As a result of our search strategy (see 'Materials and Methods for Database Construction' for details), certain proteins in MSAs might not be included in the current KinetochoreDB. To facilitate the comparison between entries in KinetochoreDB and their homologs, we archived the homologs by extracting protein UniProt IDs. Detailed information for these files is available in the 'Protein alignment' section of the webpage. KinetochoreDB will be updated on a regular basis (typically every 2-3 months) to include newly available entries from various resources, in order to allow an up-to-date archive of recent results of the kinetochore and its related proteins. All the updates will be carefully reviewed prior to their release. In addition, we also allow users to submit their research findings related to kinetochore to our KinetochoreDB. New submission should include protein/ gene detailed information (such as sequence, molecular weight and protein function description), protein structural information, protein functional annotation, single point mutation and its pathogenicity and metabolic/signaling pathway (Supplementary Figure S2). After careful review and validation, new data will be included in KinetochoreDB and made publically available. Discussion Certain protein entries in KinetochoreDB harbor mutations. We thus provide a brief statistical analysis of these mutations. A total of 1424 mutations were found to occur in the 206 entries in KinetochoreDB. Among these, 689 (48.4%) mutations were found to cause diseases, while 735 (51.6%) were nonsense mutations ( Figure 5A). PTMs, on the other hand, extend the chemical repertoire of amino acids by attaching new chemical groups and small molecules to the side chains of amino acids. Using the available PTM annotations in KinetochoreDB, we analysed the distribution of different types of PTMs for all the entries. We note that the kinetochore and its related proteins possess many PTM sites, the top three of which are phosphorylation, acetylation and methylation, respectively ( Figure 5B). The distribution of different subtypes of acetylation and phosphorylation is also shown in Figure 5B. We further plotted the distributions of different types of mutations in Figure 6. We can see that there is no apprarent difference between the two types of mutation patterns ( Figure 6). In addition, with the comprehensive dataset from KinetochoreDB, we conducted a statistical analysis of the number of proteins involved in different GO terms including cellular component, molecular function and biological process. The results are shown in Figure 7. (GO:0003777) and kinetochore binding (GO:0043515). For biological process, the 10 top ranked GO terms are protein localization to kinetochore (GO:0034501), attachment of spindle microtubules to kinetochore (GO: 0008608), kinetochore assembly (GO:0051382), attachment of mitotic spindle microtubules to kinetochore (GO:0051315), attachment of spindle microtubules to kinetochore involved in homologous chromsome segregation (GO:0051455), centromere complexe assembly (GO: 0034508), positive regulation of attachment of spindle microtubules to kinetochore (GO:0051987), sister chromatid biorientation (GO:0031134), regulation of attachment of spindle microtubules to kinetochore (GO: 0051988) and kinetochore organization (GO:0051383). More specifically, 414, 285 and 72 proteins contain the annotation of condensed chromosome kinetochore (GO: 0000777), microtubule motor activity (GO:0003777) and protein localization to kinetochore (GO:0034501). It should be noted that the statistical results regarding the PTM sites, mutations and GO terms are merely observations based on the entries in KinetochoreDB, rather than being interpreted as a biologically siginificant finding, as the entries included in KinetochoreDB only represent a subset of the entire proteome and are from different species. Conclusions The kinetochore and its related proteins play extremely important roles during cell division and mitosis. In the past few decades, research on this topic has attracted a great deal of interest, not only because they are important for the cell cycle, mitosis and meiosis (1-6), but also because they harbor mutations that can cause human diseases (12)(13)(14)(15). In this context, databases such as KinetochoreDB that provide comprehensive annotations on the repertoire of kinetochore-related proteins will greatly facilitate in-depth functional investigation of these proteins and their relationships with human diseases. Through effective data integration from multiple public resources, KinetochoreDB has collected large amounts of information for related protein entries with respect to their amino acid sequence, protein 3D structure, biological function and evolutionary conservation. By providing comprehensive functional annotations of all available kinetochore-related proteins, we believe that this online resource will be used as a powerful tool to bridge functional characterization and disease-associated mutation studies of this important class of proteins. In the future, we will endeavor to improve and update the annotations and analysis of data entries in KinetochoreDB by the following means: (i) We will keep the database updated and provide up-to-date information to synchronize with the research progress in the kinetochore and its related proteins; (ii) We will integrate genomic information into our database and source these data from publicly available information or bioinformatics programs. These include coding sequence, transcription factor binding site, enhancer, promoter and other upstream or downstream regulatory information; (iii) We will combine other state-of-the-art predictors to annotate the natively disordered regions of all entries in the database, while highlighting the consensus prediction. Meanwhile, we will also collect experimentally verified disordered regions from DisProt (32), the most comprehensive resource dedicated to annotating disordered region of proteins; (iv) we will encourage experimental biologists to contribute to the development of KinetochoreDB by submitting their recent findings by making available newly added entries in the database after careful review. In addition we will continue to improve and update the annotations and analysis of all entries in KinetochoreDB by implementing secondary analysis functions of the database and by integrating high-throughput experimental data. In particular, we will explore gene expression microarray data, transcriptomics and proteomics and detailed functional pathway data, so as to provide a comprehensive useful resource for the wider research community. Supplementary data Supplementary data are available at Database Online. Real-world outcomes of postmastectomy radiotherapy in breast cancer patients with 1–3 positive lymph nodes: a retrospective study Objective: To assess the treatment outcomes and to explore the determinants of clinical outcome in breast cancer patients with 1–3 positive nodes who did or did not receive postmastectomy radiotherapy (PMRT) in a tertiary care referral cancer center in Northern Thailand. Methods: We investigated a retrospective cohort of registered breast cancer patients at the Faculty of Medicine, Chiang Mai University, Thailand from 2001–2007. Analysis was performed using Cox regression models to identify factors affecting the overall survival (OS) and relapse-free survival (RFS) rates. Comparisons were made between two cohorts: women who received adjuvant PMRT (74 patients) and women who did not receive adjuvant PMRT (81 patients). Results: A total of 155 patients were included with a median follow-up period of 4.45 years. There was a statistically significant 4-year OS difference between the two groups of patients: 100% for the PMRT group and 93.1% for the non-PMRT group (P = 0.044). The 4-year RFS was 85.9% for patients receiving PMRT and 78.3% for patients who did not receive PMRT (P = 0.291). On multivariate analysis of OS, using hormonal treatment was the only significant independent factor associated with improved OS. On multivariate analysis of RFS, none of the variables were significantly associated with improved RFS. PMRT was notfound to be a prognostic variable related to the outcome of patients using a logistic regression model. Conclusion: Our retrospective, hospital-based analysis demonstrated that PMRT improved the treatment outcome in terms of OS for women with 1–3 node positive early-stage breast cancer. INTRODUCTION Breast cancer is the most common cancer found in Thailand with 21 967 patients diagnosed in the years 2001-2003 [1]. In early breast cancer, the complete cure of disease and prevention of recurrence are the primary goals of therapy, and these can be achieved by a multimodality of approaches. Surgery is accepted as a standard treatment for patients with early-stage breast cancer and is usually followed by adjuvant chemotherapy or radiotherapy to control local |
recurrence arising from residual disease. Adjuvant chemotherapy has been proven to decrease the incidence of recurrence by 23% with a corresponding 15% decrease in mortality [2]. Adjuvant radiotherapy is another approach demonstrated in many randomized control studies to be effective in reducing local recurrence by 60-90% [3][4]. In early-stage breast cancer, the recurrence rate was reduced from 24% to 8.5% with the addition of radiation after breast-conserving surgery (BCS). The same benefit was also seen among those with high risk, as the local recurrence rate was reduced from 35% to 10% [5][6][7]. A study by the Early Breast Cancer Trialist's Collaborative Group (EBCTCG) in 2005 investigated the effects of adjuvant radiotherapy in early breast cancer patients. The 15-year mortality was seen to be significantly lower among patients receiving either breast-conserving therapy or mastectomy followed by irradiation [8]. Overgaard demonstrated that the benefit of postmastectomy irradiation (PMRT) was present regardless of the extent of disease [9]. In this study, PMRT resulted in a reduction in the 15-year locoregional recurrence rate from 51% to 10% among patients with ≥ 4 positive lymph nodes, and from 27% to 4% in patients with 1-3 positive lymph nodes, corresponding to a relative risk of 0.17 and 0.10, respectively [9]. However, immediate-and long-term adverse effects of PMRT are a major concern as significant numbers of deaths due to contralateral breast cancer, non-breast cancer, lung cancer and cardiac death have been observed among irradiated women [2,8,10]. Due to these concerns, debate on whether PMRT should be administered to those with < 4 nodes still exists. It remains controversial whether the risk of cardiac disease outweighs the disease-free survival benefit, buoyed by the inconsistency of results among different studies. Despite the controversy, clinicians for the most part have adopted the concept of adjuvant radiotherapy as a standard approach. However, there remains a paucity of evidence showing the real-world clinical outcomes of this practice in clinical settings. This retrospective cohort study was undertaken to assess clinical outcomes for those who have and those who have not received radiation therapy after surgery and/or chemotherapy/hormone therapy. The objectives of the study were to describe and to determine the differences in clinical outcomes in patients with 1-3 node positive early-stage breast cancer treated with and without PMRT. MATERIALS AND METHODS The records of 1883 breast cancer patients who were treated at the Faculty of Medicine, Chiang Mai University between 2001 and 2007 were reviewed using a nonrandomized retrospective cohort study design. As this study's objective was to evaluate the real-world clinical outcomes of adjuvant radiotherapy in patients with early-stage breast cancer, we retrospectively collected such outcomes from the existing medical records. In our hospital, the indications for administering PMRT were locoregionally advanced stage at diagnosis, such as clinical stage T3-T4 and/or N2 or more, pathological ≥ 4 positive lymph nodes, and a close/positive surgical margin. The use of PMRT for 1-3 positive lymph nodes is dependent on a decision by the multidisciplinary team according to adverse pathological factors (e.g. < 50 years of age, negative hormonal status, or positive HER-2 status). The inclusion criteria for this study were patients who had: undergone modified radical mastectomy with or without adjuvant chemotherapy or adjuvant radiotherapy, Stage I-IIB breast cancer, pathological 1-3 axillary nodes positive, and negative resection margins. The exclusion criteria included: previous radiotherapy (breast or chest wall), incomplete treatment (either chemotherapy or radiotherapy), clinical N2 disease, pathologically revealed perinodal/extracapsular extension, metastatic breast cancer, or incomplete follow-up during the study period. The primary endpoints were 4-year relapse-free survival (RFS) and 4-year overall survival (OS). RFS was defined as the time from the date of primary surgery to the date of documented recurrence. OS was defined as the time from the date of primary surgery to the date of expiration. Locoregional recurrence was also assessed in the study group, although not as a primary endpoint. Locoregional recurrence was defined as recurrence at the skin or soft tissue over the ipsilateral chest wall or a recurrence at the ipsilateral regional lymphatic sites. Statistical analysis Summary statistics for continuous variables were presented as categories. All outcomes for patients who had undergone PMRT were compared with those for patients who had not received PMRT. RFS and OS were estimated using the Kaplan-Meier method. The difference in 4-year RFS and 4-year OS between the two groups was determined using the Log-Rank test. In order to investigate the association between 4-year OS and baseline patient and treatment characteristics, univariate and multivariate Cox regression models were used. Using a stepwise backward approach to variable selection, we fit a Cox regression model to variables associated with the outcome (P < 0.20) in the univariate analysis. All P-values were two-sided, and P-values < 0.05 were considered to be statistically significant. All analyses were performed using STATA, version 10.1. (Stata Corp LP, Texas, USA). RESULTS Among the 1883 primary breast cancer patients, 316 patients were identified as having 1-3 positive lymph nodes from surgical pathology specimen. Of those, 155 patients were eligible according to the inclusion and exclusion criteria (see above). Most of the patients (137/155, 88.4%) had received adjuvant chemotherapy. Among these, 63.2% had received a cyclophosphamide, methotrexate and fluorouracil (CMF) regimen, 35.5% had received an anthracycline regimen with or without a taxane. Adjuvant hormonal treatment was administered to 105 (67.7 %) patients. It was found that 74 (47.7%) patients had undergone PMRT while 81 (52.3%) patients had not received PMRT. The median number of axillary nodes removed was 13 and 15 in patients who had received PMRT and not received PMRT, respectively. The patient and treatment characteristics are shown in Table 1. With the median follow-up of 4.45 years, 116 (74.8%) patients were alive without any evidence of disease, 31 (20.0%) patients were alive with disease, 7 (4.5%) patients were deceased and 1 (0.7%) patient was lost to follow-up. The 4-year RFS was 81.6% (95% CI, 73.8%-87.2%) ( Fig. 1) (Figs 3 and 4). The 4-year RFS rates were not significantly different between those who had received and had not received PMRT (85.9 vs 78.3%; P = 0.291). However, the 4-year OS rates were found to be significantly different (100 vs 93.1%; P = 0.044) between the two groups. On univariate analysis of RFS, having a positive progesterone receptor (PR) test, and receiving adjuvant endocrine therapy were significantly associated with improved RFS. The results of univariate analysis of RFS are summarized in Table 2. A multivariate analysis was performed to determine the contributing factors for RFS and no variable was found to be an independently significant factor for RFS (Table 2). Univariate analysis was then performed to identify factors affecting OS (Table 3). In brief, not using adjuvant hormonal treatment was a poor prognostic factor for OS. Multivariate analysis revealed that using hormonal treatment was the only significant independent factor for OS (Table 3). PMRT was not found to be prognostic with respect to RFS and OS using the logistic regression model. DISCUSSION It has been shown in multiple studies that radiotherapy reduces locoregional recurrence. The locoregional failure rate from the landmark trials ranged from 13-33% in this group of patients [11][12][13][14]. One of the possible reasons for the wide range in locoregional failure rate was the number of axillary nodes removed, which was low to moderate in some studies: a median of 7 in the Danish trial [9], and 11 in the British Columbia study [15], compared with higher numbers in the others. Our study had a relatively high number of nodes removed (a median of 13 nodes in patients who had received PMRT, and 15 nodes in patients who had not received PMRT). Overgaard et al. reports a 114-month rate of locoregional recurrence of 26% for patients who did not receive radiotherapy and only 5% for those received radiotherapy [16]. The question is whether or not this locoregional recurrence decrease translates into a further benefit of OS. The benefit of radiotherapy on OS has been demonstratively shown in women of all ages with positive lymph nodes. However, it remains unclear whether this benefit is simply due to benefit in the ≥ 4 positive node group, in which there is already no controversy for the effectiveness of radiotherapy. This is an important question as a significant percentage of women in today's patient population present with 1-3 positive axillary nodes. They could be spared treatment sequelae (eg. lymphedema) if radiotherapy was found not to be effective. In a Danish trial by Overgaard et al., in the 1-3 node positive patients, adjuvant RT improved locoregional control and also increased survival at 10 years by 17% [17]. At the median follow-up of 4.45 years, our retrospective study concurred with the Overgaard et al. result for OS, demonstrating a significant increase in OS with adjuvant radiotherapy compared with no radiotherapy (P = 0.044, Fig. 4). However, there was no significant difference seen for RFS (P = 0.291, Fig. 3). In a retrospective study from Taiwan [18], Cheng et al. analyzed the incidence of locoregional failure in 125 postmastectomy patients with 1-3 positive axillary lymph nodes Cosar et al. [19] conducted a retrospective study of 90 patients with a similar design to and the same endpoints as our study, and demonstrated that PMRT in T1-2 and 1-3 axillary lymph node positive patients made a statistically significant improvement in RFS (P = 0.034), but no improvement in OS (P = 0.087). Tendulkar et al. [20] reported the results from a retrospective review of 369 breast cancer patients with 1-3 positive lymph nodes, of whom 271 did not receive PMRT and 98 received PMRT. Their 5-year rate of locoregional recurrence (LRR) was only 8.9% without PMRT Most studies have concluded that locoregional treatment with PMRT improved survival by reducing locoregional failure rate [21][22][23][24]. Although we did find a statistically significant improvement in OS, our study found similar rates of locoregional failure between the two cohorts (20.8% in the PMRT group, and 21.4% in the no-PMRT group). This finding could be explained by the retrospective nature of our study, and also the smaller number of patients in our study than in others. In our univariate analysis, not receiving chemotherapy and hormonal therapy were statistically significant factors associated with a lower OS rate. Lin et al. [25] reported tumor size, age and estrogen receptor (ER) status to be independent prognostic factors for OS in breast cancer patients with 1-3 axillary lymph node metastases in multivariate analysis. Adjuvant hormonal therapy turned out to be the only indicator with an independent impact on OS by multivariate analysis. Most likely due to the sample size limitation, our study could not demonstrate that PMRT was an independent prognostic factor for OS as determined by univariate analysis. For the 4-year RFS, although we did not find a statistically significant difference with and without PMRT, there was a trend to higher RFS in the PMRT group, especially for the first 20 months of the follow-up time (Fig. 3). CONCLUSION In conclusion, our study reported a significant improvement in the 4-year OS rate with PMRT (100 vs 96.1%; P =0.04). It also showed a 7.6% improvement in the 4-year RFS rate with PMRT, although this result was not statistically significant. Our study is one of a number investigating treatment of breast cancer with 1-3 positive lymph nodes that supports the use of PMRT, especially in Asian women. The SUPREMO trial, a prospective evaluation of PMRT in this 1-3 node patient population subset, will hopefully provide clearer answers to this controversy. FUNDING This study was supported in part by grant from F. Hoffmann-La Roche Ltd. (Thailand). Inaccuracies in electronic health records smoking data and a potential approach to address resulting underestimation in determining lung cancer screening eligibility Abstract Objective The US Preventive Services Task Force (USPSTF) requires the estimation of lifetime pack-years to determine lung cancer screening eligibility. Leading electronic health record (EHR) vendors calculate pack-years using only the most recently recorded smoking data. The objective was to characterize EHR smoking data issues and to propose an approach to addressing these issues using longitudinal smoking data. Materials and Methods In this cross-sectional study, we evaluated 16 874 current or former smokers who met USPSTF age criteria for screening (50–80 years old), had no prior lung cancer diagnosis, and were seen in 2020 at an academic health system using the Epic® EHR. We described and quantified issues |
in the smoking data. We then estimated how many additional potentially eligible patients could be identified using longitudinal data. The approach was verified through manual review of records from 100 subjects. Results Over 80% of evaluated records had inaccuracies, including missing packs-per-day or years-smoked (42.7%), outdated data (25.1%), missing years-quit (17.4%), and a recent change in packs-per-day resulting in inaccurate lifetime pack-years estimation (16.9%). Addressing these issues by using longitudinal data enabled the identification of 49.4% more patients potentially eligible for lung cancer screening (P < .001). Discussion Missing, outdated, and inaccurate smoking data in the EHR are important barriers to effective lung cancer screening. Data collection and analysis strategies that reflect changes in smoking habits over time could improve the identification of patients eligible for screening. Conclusion The use of longitudinal EHR smoking data could improve lung cancer screening. INTRODUCTION Lung cancer is the leading cause of cancer deaths in the United States. 1 Lung cancer screening with low-dose computed tomography could result in a 20% relative reduction in lung cancer mortality. 2 In March 2021, the US Preventive Services Task Force (USPSTF) recommended that annual low-dose lung cancer screening be offered to patients 50-80 years old with a 20þ pack-year smoking history who are current smokers or quit in the last 15 years. 3 Despite potential benefits, lung cancer screening rates remain at about 5% in the United States among individuals meeting USPSTF screening eligibility criteria. 4 Electronic health record (EHR) systems could help improve lung cancer screening rates by enabling the collection and use of detailed, discrete data on smoking history. Such data can be used to automate determination of USPSTF eligibility and inform high-quality shared decision-making based on individual-specific benefits and risks from screening. 3,5 In addition, these data can be used to help identify patients who are particularly good candidates for screening due to their individual benefit and risk profile. 6,7 Key smoking data collected in EHRs include the number of packs smoked per day, years smoked, and smoking quit date. [8][9][10][11] For example, the vendors Epic and Cerner, which held a combined 58% ambulatory EHR market share in 2019, 12 use the most recently recorded packs-per-day and years-smoked to calculate lifetime smoking exposure. In the default configuration for Epic, lifetime smoking exposure is calculated using the most recently recorded packs-per-day and years-smoked as "pack-years" (Figure 1). Similarly, Cerner calculates "total pack-years" from the most recent record. 13 Thus, while some EHRs hold over 25 years of longitudinal smoking data, these data are oftentimes not being used to identify patients eligible for lung cancer screening. Previous studies have reported on the low quality of the most recent smoking data in EHRs. 8,[14][15][16][17][18] One retrospective study found 96.2% discordance in pack-year smoking history between the EHR and data obtained through a shared decision-making conversa-tion. 14 In a qualitative study of primary care physicians, some providers expressed lack of trust in EHR smoking data and perceived smoking documentation in the EHR as inaccurate or insufficient for determining whether to order low-dose computed tomographies. 19 Another study evaluated the impact of random error in pack-year estimation. 17 However, these studies did not quantify the potential impact of such data issues nor propose an approach to overcoming such issues using existing longitudinal data. OBJECTIVE Our objectives were to address both limitations by (1) characterize EHR smoking data issues, and (2) propose a potential approach to addressing these issues using longitudinal EHR smoking data. METHODS This was a cross-sectional study of University of Utah (UU) Health patients 50-80 years old with a history of smoking. This study was approved by the UU Institutional Review Board. UU Health is comprised of 5 hospitals and 11 community health care centers providing inpatient and ambulatory care across various specialties including primary care, cancer care, and pulmonary care. UU Health has 12 primary care clinics, and all primary care clinics have used the EpicV C EHR system since 1999. Smoking data for some patients were electronically collected since 1995. Patients were included based on the following criteria: (1) inperson or virtual provider visit in a study clinic in 2020; and (2) age 50-80 years on December 31, 2020. Patients were excluded based on the following criteria: (1) had a lung cancer diagnosis before January 1, 2020, (2) had no smoking status recorded, or (3) never smoked according to the EHR data. Lung cancer diagnosis was determined based on Epic diagnosis codes in the problem list, medical history, and visit diagnoses (Supplementary Appendix Box 1). "Never smoker" status was assigned to patients who reported that they never smoked at every visit when the smoking history was recorded. An individual who had previously smoked, but who was not documented as such in the EHR, would have been missed and represent a false negative. Age, gender, race, ethnicity, and smoking history data were extracted from the EHR on September 2, 2021. Due to the elevated risk of lung cancer among Black patients, the study race was set to Black if at least one race recorded in the EHR was Black. 20 We used smoking data entered in the EHR in designated structured fields. In the default configuration for Epic used by UU Health, smoking history is usually recorded by the medical assistant or provider in 4 fields: smoking status, packs-per-day, yearssmoked, and smoking quit date ( Figure 1). Smoking status is a drop down. Years-smoked and packs-per-day fields allow entering free text. Smoking data required some data cleaning before use. For example, we replaced '20þ' with 20, and '0.5' with '0.5'. An additional field for smoking start date is available for patients to complete through the patient portal, but the field is usually not populated. Every time smoking history is documented in the EHR, a new smoking history record is created. The most recent smoking data are displayed on the screen ( Figure 1). Longitudinal smoking data can be accessed by clicking a link to the "Audit Trail" in the History Tab for "Substance and Sexual Activity." For screening-eligible patients, we estimated their 10-year risk of developing lung cancer and estimated life-expectancy using equations developed by Bach et al. 21,22 Then, based on the work of Caverly et al 6 and Mazzone et al, 7 we classified patients as being particularly 'high-benefit' patients with regard to lung cancer screening if they had a 10-year risk of developing lung cancer !5.2% and an estimated life-expectancy >10 years. Baseline, Longitudinal, and Combined Approaches Both EHR vendors and external clinical decision support vendors usually use the most recently documented years-smoked, packs-perday, and years-quit to assess for screening eligibility (Box 1). 19 For example, this Baseline Approach was used by UU Health to generate care reminders and to support shared decision-making for lung cancer screening. 19,23 To address the issues identified with the smoking data, as well as the presumably inaccurate assumptions used by the Baseline Approach, we developed a Longitudinal Approach using longitudinal smoking data. The main purpose of the Longitudinal Approach was to leverage longitudinal EHR data to identify individuals who may be eligible for lung cancer screening but were deemed to be ineligible using the Baseline Approach. Figure 2 illustrates the Baseline and Longitudinal Approaches. The logic of the Longitudinal Approach is described in Supplementary Box 2 of the Appendix. The main assumptions of the Longitudinal Approach are that (1) current smokers continue to accumulate smoking exposure and that (2) the packs-per-day recorded at a given point in time reflects the packsper-day the individual was smoking at that time. To identify as many patients as possible who may be eligible for lung cancer screening based on available EHR data, we combined the Baseline and Longitudinal Approaches into a Combined Approach that identifies patients eligible for screening through either approach. Quantification of smoking data issues, iterative development of Longitudinal Approach, and manual verification of approach To characterize smoking data issues and to develop a Longitudinal Approach to address these issues, the following steps were taken. Box 1 Logic of the Baseline Approach The Baseline Approach is as follows: 1. Pack-years are calculated by multiplying the years-smoked and packs-per-day most recently recorded in the EHR. 2. For former smokers, years-quit are calculated as the years that have passed since the quit date most recently recorded in the EHR. 3. The following assumptions are used: a. The most recently recorded years-smoked is accurate. b. The most recently recorded packs-per-day reflects the average packs-per-day smoked over the individual's entire smoking history as opposed to the packs-per-day currently smoked. c. If the most recent record lacks a needed data point, there is no relevant data available. For example, even if it is documented that a patient transitioned from a current smoker to a former smoker 3 years ago, the years-quit is deemed unknown if the field is not populated in the most recent record. First, a sample EHR smoking records were reviewed to identify patterns of data issues and an initial Longitudinal Approach was developed to address these issues. Then, the eligibility of study patients based on USPSTF criteria was determined using both the Baseline and Longitudinal Approaches, and patients with discrepancies in eligibility determinations were selected for manual review so as to focus on cases of potential clinical significance. During this review, the patient's longitudinal smoking history was reviewed to identify additional data issues not yet accounted for in the Longitudinal Approach. To facilitate the review, data records that were identical across visits were merged into a single record (eg, records from 10 visits spanning March 1, 2016 to July 15, 2019 all noting that the patient was a 20 years, 1 pack-per-day smoker were collapsed into a single record spanning these dates). Finally, additional identified data issues were addressed by updating the Longitudinal Approach. These last 2 steps of data issue identification and approach enhancement were iteratively repeated until no further data issues were identified following a review of 51 patient record. Following the development of the Longitudinal Approach addressing all identified issues, the appropriateness of the Longitudinal Approach was manually verified. In this verification process, PVK and KK conducted independent manual review of the smokingrelated EHR records of 100 randomly selected patients with available smoking data. The reviewers manually assessed the patient's lifetime tobacco exposure using the Longitudinal Approach and independently determined whether the patient met USPSTF eligibility criteria. Discrepancies between the 2 reviewers were adjudicated through discussion, and all discrepancies between reviewers were identified as resulting from an error in manual application of the approach by one of the reviewers. No new data issues were identified during this process. Finally, the prevalence of each data issue was quantified. Statistical analysis All statistical analysis was performed using R version 4.1.0. We used n (%) and mean (standard deviation) to describe nominal and continuous variables. We used bootstrapping to calculate the 95% confidence intervals (CIs) of the relative changes and the absolute differences in average pack-years and years-quit. McNemar's Chisquared test was used for logical variables and Wilcoxon signed rank test was used for numeric variables to compare algorithm accuracy and patient eligibility for lung cancer screening according to the Baseline and Longitudinal Approaches. We performed a sensitivity analysis in which we removed from the Longitudinal Approach the assumption that the packs-per-day recorded at a given point in time reflects the packs-per-day the individual was smoking at the time. Given that 2020 data could have been affected by the disturbances from coronavirus disease-19 (COVID-19), we repeated the analysis using 2019 data. To visualize the relationship between pack-years estimated by the approaches, we used a local polynomial regression (R ggplot2 package, geom_smooth function, "loess" method). An analysis stratified by race/ethnicity and gender was conducted to ensure fairness. Patient characteristics Patient inclusion and exclusion criteria are shown in Figure 3. In total, 16 874 patients met study criteria. About 35% of patients smoked at some point in their life. Patient characteristics for patients meeting inclusion criteria are summarized in Table 1. Among included patients, about 25% of patients were current smokers and 75% were former smokers. Smoking data issues Among 16 874 patients meeting eligibility criteria, the EHR data contained 12 types of issues that could cause errors in calculating pack-years and years-quit using the Baseline Approach (Table 2). Over 80% of patient records had at least one such data issue. |
The 3 most common issues were missing data, stale data, and changes in packs-per-day that affected all the previous years. A given patient could have more than one data issue, including issues that are specific subsets of more general issues. Lung cancer screening eligibility determination by Baseline and Longitudinal Approaches The Longitudinal Approach was verified as being implemented as intended, with the computationally implemented Longitudinal Approach having a sensitivity of 0.97 (95% CI: 0.92, 1) and specificity of 1 (95% CI: 1, 1) with regard to USPSTF eligibility classification in comparison to the manual application of the approach for the 100 randomly selected patients. Interrater reliability for this determination, as calculated using Cohen's kappa, was high (0.91). Figure 4 presents patient eligibility diagram based on USPSTF 2021 eligibility criteria using Baseline and Longitudinal Approaches. Baseline Approach identified 2228 patients and the Longitudinal Approach identified 3167 patients as eligible for lung cancer screening. Figure 5 depicts a Venn diagram demonstrating how many patients were identified by just one approach. A total of 2066 individuals were identified as eligible by both approaches. The Longitudinal Approach enabled identification of 1101 screening-eligible individuals in addition to the 2066 individuals identified by both algorithms. These additional records included 271 patients who were missing required data (packs-per-day, years-smoked, and/or years quit) in the most recent observation. The rest were missed by the Baseline Approach due to the other inaccuracies described in Table 2. Conversely, the Baseline Approach identified 162 patients as being eligible whom the Longitudinal Approach did not consider to be eligible. One hundred forty-one (87%) of these 162 patients had a recent increase in packs-per-day. For example, after reporting smoking 0.5 packs-per-day for 30 years, they reported smoking 1 pack-perday, which would make them eligible according to the Baseline Approach, but not the Longitudinal Approach. Since there are conceivable situations were both algorithms could be correct, it could be reasonable to use both algorithms in a Combined Approach. Figure 6 depicts the pack-year discrepancies between the 2 algorithms in a scatter plot. The dark vertical lines indicate data patterns related to EHR-recorded pack-years remaining static over multiple years at increments such as 10, 20, and 30 pack-years rather than incrementing upwards with continued tobacco exposure. The regression line shows that the Longitudinal Approach provides higher pack-year values compared to the Baseline Approach in the 0-50 pack-years range, but lower values for the higher pack-year range. The sensitivity analysis evaluated the impact of removing the Longitudinal Approach's assumption that the packs-per-day recorded at a given point in time reflected the packs-per-day the indi-vidual was smoking at the time. Even with this assumption removed, the Longitudinal Approach identified a significantly greater number of patients as being eligible for screening (2758 vs. 2228, P < .001). In addition to 2115 individuals identified by both algorithms, the Longitudinal Approach identified 643 additional individuals and the Baseline Approach identified 113 additional individuals. Additional patients identified using the Combined Approach Using the Combined Approach, 3329 patients were identified as potentially meeting USPSTF eligibility criteria for lung cancer screening, which was 1101 (49.4% [95% CI: 46%, 53%]) more patients than the 2228 patients identified using the Baseline Approach alone (P < .001) ( Given that 2020 data could have been affected by the disturbances from COVID-19, we repeated the analysis using 2019 data (Supplementary Appendix Table 1). All the conclusions of the study held for 2019 data. Fairness analysis Supplementary Appendix Table 1 shows results of the fairness analysis stratified by race/ethnicity and gender. Using longitudinal data identified significantly more patients across race/ethnicity and gender as potentially eligible for lung cancer screening compared to using the Baseline Approach. Using the longitudinal data might be especially beneficial for Hispanic patients and women. The relative increase in the identification of potentially eligible patients using the Combined Approach was 75.4% (95% CI: 57.5%, 96.1%) for Hispanic patients, 47.1% (95% CI: 43.6%, 50.8%) for White patients, 56.8% (95% CI: 50.8%, 63.4%) for female patients, and 43.7% (95% CI: 39.3%, 48%) for male patients. DISCUSSION We quantified issues in EHR smoking data at a single site. Over 80% of patient records had at least one issue, including missing, outdated and inaccurate data. To partially address these issues, we developed an approach that uses longitudinal smoking data. This Longitudinal Approach can help identify patients eligible for lung cancer screening who would be missed by the Baseline Approach, which uses only the most recent EHR data and is the predominant algorithm used by market-leading EHR systems. Among 16 874 current and former smokers with no prior history of lung cancer, the Combined Approach leveraging both most recent and longitudinal data was able to identify 49.4% (95% CI: 46%, 53%) more patients potentially eligible for lung cancer screening than the Baseline Approach (P < .001). This included identifying 40.4% (95% CI: 36%, 45.2%) more high-risk, high-benefit patients (P < .001). Screening is particularly important for this high-benefit population and misclassifying such patients as ineligible is particularly disconcerting. If implemented in clinical practice, the Combined Approach could substantially increase the number of individuals who are evaluated for lung cancer screening. This study has several implications. First, this study underscores the critical need to improve the collection of smoking history in the EHR in order to improve lung cancer screening. In our patient popu- Other race/ethnicity includes non-Hispanic participants with race other than White or Black or those who chose not to disclose race. lation, only about 60% of ever-smokers in the USPSTF-eligible age range had the requisite smoking data recorded in the EHR to determine screening eligibility. Moreover, we found that over 80% of patients with the requisite smoking data had issues in that data. This is consistent with others' findings, such as a study by Modin et al 14 which found that pack-year history calculated using the most recent EHR data were almost always different from pack-year history obtained through clinical interview at a centralized lung cancer screening program. As EHR data are known to be incomplete when it comes to cohort identification, 24 approaches improving data col-lection should be considered, such as leveraging health information exchanges and directly engaging patients in the collection and review of the needed data. As a second implication, smoking history should always be verified prior to making screening or treatment decisions dependent on smoking history. Clinicians and health IT implementers need to be mindful of the potential impact of missing and inaccurate data when making decisions about lung cancer screening based on EHR smoking data. As a third implication, the Longitudinal Approach described in this paper add clinical value in several ways. As one potential use, the Longitudinal Approach could be used to alert clinicians about patients with significant discrepancy between the results of the Baseline and Longitudinal Approaches, so that the clinicians can verify the smoking history with the patient. For example, when the Baseline Approach estimated higher pack-years than the Longitudinal Approach in our dataset (Figure 6), this often appeared to be a result of cigarettes-per-day being mistakenly entered as pack-per-day by EHR users, which were taken at face value in the Baseline Approach but assumed to be cigarettes-per-day in the Longitudinal Approach (eg, assumed to be 10 cigarettes-per-day as opposed to 10 packs-perday, which would likely not be possible to achieve). Second, patients eligible for lung cancer screening through either the Baseline or Longitudinal Approaches (ie, through the Combined Approach) could be flagged for evaluation for lung cancer screening. This could significantly increase the number of eligible patients who are identified for screening, which is a critical need given that <5% of eligible patients are currently screened in the United States. 4 Third, more individuals with high lifetime smoking exposure could be targeted for smoking cessation outreach, such as through electronic reminders as described by Bar et al. 25 Fourth, the Combined Approach could be used to identify household members with concerning levels of secondhand tobacco exposure. Fifth, the Longitudinal Approach could be used to pre-populate an EHR's smoking history if and when EHR vendors move to a period-based reporting approach (eg, smoked 1 pack-per-day from age 18 to 35, quit from 35 to 40, and then smoked 0.5 packs-per-day since). An important assumption of the Longitudinal Approach was that patients tend to report how much they are currently smoking instead of an average of how much they smoked over their lifetime. A period-based reporting approach would obviate the need for such assumptions, and we are aware of at least one major EHR vendor planning to convert to such an approach. As a final implication, this study suggests that historical smoking data should ideally be made accessible to clinical decision support systems in addition to the most recent smoking data. In particular, it would be ideal if longitudinal smoking data were available from EHRs through their Health Level 7 (HL7) Fast Healthcare Interoperability Resources (FHIR) application programming interfaces, so that the data can be used by external apps and Web Services connected to the EHR to identify a patient's eligibility and appropriateness for screening. 19,26 One of the strengths of this study is that it used a computational approach to quantify issues in EHR smoking data among a large group of patients (ie, all patients with a history of smoking seen at an academic medical center). In comparison, prior studies have only characterized the issue qualitatively or in a much smaller sample of patients, 14,16 or focused on issues arising from random noise rather than bias in the data. 17 As a second strength, this study proposes a computable algorithm to address these issues. Thus, we not only quantify the problem but also provide an actionable approach to addressing the problem. Finally, this study evaluated the accuracy of the Longitudinal Approach in identifying patients considered to be particularly high benefit. 6,7 This study has several limitations. First, this study did not include a true gold standard for lifetime tobacco exposure. However, due to the evolving nature of smoking habits over time, creating a true gold standard would have required decades-long prospective data collection with frequent questionnaires. When estimating lifetime smoking exposure through a single patient interaction, there are methods to reduce recall bias such as the Timeline Followback. 27 However, it is questionable whether an individual's recollection many decades later is more accurate than what they reported at the time. For example, for a 50-year-old patient who began reporting their smoking habits in the EHR 20 years ago, it is questionable whether their recollection at age 50 of their smoking habits in their 30s or 40s is more accurate than what they reported back when they were that age. Indeed, a 2020 study found that when individuals were prospectively asked to provide a self-reported lifetime smoking history 1 month apart, differences in their recollection just over this 1 month was such that 12% of participants were eligible for lung cancer screening at one but not both assessments. 28 Given the potential inaccuracy of self-recollection for lifetime tobacco exposure, even when done prospectively, we did not seek to establish a gold standard. Instead, we sought to characterize data issues and to develop an approach to help identify as many patients as possible who may be eligible for lung cancer screening based on available EHR data. While use of the Combined Approach may result in a false positive identification of screening-eligible patients as compared to the ground truth, we believe this is a reasonable approach to using EHR data to identify patients for lung cancer screening, as potentially eligible patients can be evaluated by their clinical providers to verify eligibility, whereas patients not flagged as being potentially eligible may easily be overlooked for screening purposes. As the second limitation, the data analysis was conducted using data from one healthcare system; as such, replication is needed. However, we have no particular reason to believe that our findings are significantly divergent from what would be found in other clinical settings. Of note, our findings are consistent with earlier studies in this area, and the EHR product used at the study site is one of the most widely used EHRs in the United States. 12 As a third limitation, the study was performed in Utah, where the population is mostly White (78%). However, we have included a fairness analysis, which showed that the |
Combined Approach would allow identification of more patients potentially eligible for lung cancer screening in White, Black, Hispanic and other populations. The fairness analysis showed that using the longitudinal data would disproportionately benefit women and Hispanics. As a fourth limitation, we did not explicitly account for the random error inherent in patient-reported measures, including smoking history. 17 As a fifth limitation, this study did not analyze free text data. However, including free text data in the Longitudinal Approach might have further limited the portability of the approach to other healthcare systems by requiring more adaptation and validation prior to clinical use. CONCLUSION This study contributes to the body of research indicating that smoking data recorded in the EHR at a point in time are often inaccurate for assessing longitudinal smoking history. This makes relying on only the most recent smoking history suboptimal. This study showed that a Longitudinal Approach, which leverages longitudinal smoking records in the EHR, can substantially improve upon the Baseline Approach, which uses only the most recent record. Healthcare organizations, EHR vendors, and researchers should consider adopting the Longitudinal Approach. FUNDING The work reported in this paper was supported in part by the Agency for Healthcare Research and Quality under Award Number R18HS026198. TJR was supported by the U.S. National Library of Medicine of the National Institutes of Health through grant T15LM007124. The funding organizations had no role in the conceptualization, design, data collection, analysis, decision to publish, or paper preparation for this case study. The content is solely the responsibility of the authors and does not necessarily represent the official views of the organizations involved. AUTHOR CONTRIBUTIONS PVK takes responsibility for the content of the manuscript, including the data and analysis. Each author made substantial contributions to the drafting or substantial revision of the paper. All authors contributed substantially to the study design, data interpretation, and the writing of the manuscript. All authors also approved the paper for submission and agreed both to be personally accountable for the author's own contributions and to ensure that questions related to the accuracy or integrity of any part of the work are appropriately investigated, resolved, and documented in the literature. PVK designed and implemented the Longitudinal Approach. PVK, HL, and YZ conducted the statistical analyses. Newborn Size and Body Composition as Predictors of Insulin Resistance and Diabetes in the Parents OBJECTIVE We aimed to examine detailed neonatal measurements as predictors of later diabetes in both parents. RESEARCH DESIGN AND METHODS Babies (n = 617) born to nondiabetic parents in Holdsworth Memorial Hospital, Mysore, India, were measured at birth for weight; crown-to-heel length (CHL), crown-to-rump length (CRL), and leg length; skinfolds (triceps and subscapular); and circumferences (head, abdomen, and mid–upper-arm circumference [MUAC]). Nine and a half years later, glucose tolerance and fasting insulin were measured in their parents (469 mothers and 398 fathers). RESULTS Sixty-two (15.6%) fathers and 22 (4.7%) mothers had developed diabetes. There were linear inverse associations of the children’s birth weight, CHL, CRL, MUAC, and skinfolds with paternal diabetes and insulin resistance (P < 0.05 for all). Offspring birth weight and adiposity (MUAC, abdominal circumference, and skinfolds) showed U-shaped associations with maternal diabetes (P for quadratic association <0.05 for all). These associations persisted after adjusting for the parents’ current adiposity and maternal glucose concentrations and adiposity during pregnancy. Newborn adiposity was positively related to maternal insulin resistance; this association was nonsignificant after adjusting for maternal current adiposity. CONCLUSIONS Newborn size is a window into the future health of the parents. Small newborn size (especially soft-tissue body components) predicts an increased risk of later diabetes in both parents, suggesting a genetic or epigenetic link between parents’ diabetes risk and reduced fetal growth in their children. The association of higher birth weight and newborn adiposity with later maternal diabetes suggests effects on fetal adiposity of the intrauterine environment in prediabetic mothers. A number of studies have reported associations between low birth weight and risk of type 2 diabetes and cardiovascular disease in later life (1,2). Possible mechanisms include a programming effect of intrauterine undernutrition (the fetal origins hypothesis) (3) or common genetic factors that either increase insulin resistance or reduce insulin secretion, leading to both low birth weight and disease in later life (the fetal insulin hypothesis) (4). Several studies have reported that low offspring birth weight is related to an increased risk of cardiovascular disease and diabetes in either or both parents (5)(6)(7)(8)(9)(10). These data are consistent with the fetal insulin hypothesis. It is well known that babies born to mothers with gestational diabetes mellitus (GDM) tend to have higher birth weight because of an intrauterine overnutrition effect (11). Because mothers who develop GDM are at increased risk of developing diabetes in later life (12), this would create a positive relationship between offspring birth weight and risk of maternal diabetes. A combination of the above low-birth weight effect and the fetal overnutrition effect of maternal GDM could therefore result in a U-shaped relationship between offspring birth weight and risk of diabetes in mothers. Indeed such U-shaped associations have been reported in some studies (9,13). Of interest, these studies showed a U-shaped association even when mothers with GDM were excluded. Other studies, however, have found no association between offspring birth weight and maternal diabetes (14,15). Studies examining the risk of diabetes in parents in relation to offspring birth size are mainly restricted to birth weight. Birth weight reflects a crude composite measure of bone, fat, muscle, and visceral mass. These components may have different relationships with long-term health outcomes in the parents. This may, in turn, shed light on the mechanisms involved. The Mysore Parthenon Study (16) has collected detailed neonatal anthropometric data as well as parental glucose tolerance data 9.5 years following the index pregnancy. Using this data, we aimed to examine detailed neonatal measurements as predictors of diabetes in the parents 9.5 years later in an Indian population. Study population during pregnancy Details of the Mysore Parthenon Study, a prospective birth cohort study initiated during 1997-1998, have been described earlier (16). In brief, 830 eligible women (no known history of diabetes, singleton pregnancy of ,32 weeks' gestation) booking consecutively into the antenatal clinic at the Holdsworth Memorial Hospital (HMH), Mysore, India, took part in a study to investigate the incidence and determinants of GDM. GDM was diagnosed in 49 (6.2%) women. Of 830 women, 663 who chose to deliver at HMH gave birth to live babies without major anomalies and were included for additional follow-up. Neonatal anthropometry Detailed newborn anthropometry was performed according to a standard protocol, within 72 h of birth. Weight (birth weight) was measured using a digital weighing scale (Seca, Hamburg, Germany) and crown-to-heel length (CHL) and crown-to-rump length (CRL) using a Harpenden neonatal stadiometer (CMS Instruments, London, U.K.). Head circumference, abdominal circumference (at the level of the umbilicus), and mid-upper-arm circumference (MUAC) were measured with a blank tape, marked and measured against a fixed ruler. Skinfold thicknesses (triceps and subscapular) were measured using Harpenden skin-fold calipers (CMS Instruments). Leg length was derived by subtracting the CRL from the CHL. Arm muscle area (AMA) was calculated using the formula [(MUAC-p triceps) 2 /4p] (17). Follow-up of parents Additional examination of these women and their husbands was based on the follow-up of their children. Children were seen annually until the age of 5 years and every 6 months thereafter. Of 663 births, 25 children died between birth and 5 years of age, 8 children with major medical problems were excluded from the study, and 91 families either declined to participate or moved away from Mysore, leaving 539 families (Fig. 1). Among 539 mothers, 2 had died, 12 were pregnant, and 6 declined to participate in the study, leaving 519 mothers; 17 fathers had died and 88 declined to participate in the study, leaving 434 fathers. Mothers' and fathers' weight (Salter, Kent, U.K.) and height (Microtoise; CMS Instruments) and triceps, biceps, and subscapular and suprailiac skinfold thicknesses (Harpenden calipers; CMS Instruments) were measured using standardized methods. After an overnight fast, mothers and fathers with no known history of diabetes underwent a 2-h 75-g oral glucose tolerance test (World Health Organization protocol). Blood samples (fasting and 120-min post-glucose load) were collected for plasma glucose and insulin. A fasting blood sample only was collected for 21 mothers and 51 fathers who already were diagnosed as having diabetes. Samples were stored at 2808C and analyzed at the end of data collection at the Diabetes Research Centre, KEM Hospital, Pune, after transfer from Mysore in dry ice. Plasma glucose concentrations were measured by the glucose oxidase-peroxidase method (Alcyon 3000 Autoanlyzer; Abbott Laboratories); Interassay coefficients of variation were ,5% for all. Insulin was measured using a time-resolved fluoroimmunoassay (Delfia; Wallac QY, Turku, Finland). Intra-assay and interassay coefficients of variation were 12.5% at ,45 pmol/L and ,10% at .45 pmol/L. Diabetes was defined as a fasting glucose concentration $7.0 mmol/L and/or a 120min glucose $11.1 mmol/L (World Health Organization criteria) (18) and/or having been diagnosed with diabetes by a doctor since the index pregnancy. Insulin resistance was estimated using the homeostasis model assessment (HOMA-IR) equation (19). The HMH Research Ethics Committee approved the study, and informed verbal consent was obtained from the parents. Analysis sample We excluded 33 families in which the mother was diagnosed as having diabetes during the pregnancy with the index child, following an oral glucose tolerance test at 30 6 2 weeks' gestation (none of the mothers were known to be diabetic before pregnancy). We also excluded four families in which the father was known to have diabetes (diagnosed by a doctor as having diabetes and was on antidiabetes medication) before the child's conception. An additional 13 mothers care.diabetesjournals.org DIABETES CARE, VOLUME 35, SEPTEMBER 2012 1885 and 12 fathers had incomplete glucose tolerance test data. This left 469 motheroffspring pairs and 398 father-offspring pairs in the analysis (Fig. 1). Statistical methods Variables with skewed distributions were log transformed (maternal and paternal biceps skinfold thickness, fasting and 120-min glucose, and fasting insulin and HOMA-IR). Associations of neonatal measurements with diabetes and HOMA-IR in their parents were analyzed by multiple logistic and linear regression, respectively adjusting for child's sex, gestational age, and the parent's current age (model 1) and further for the parent's current BMI and sum of skinfolds (model 2). Quadratic terms were used (birth measurement 2 ) to examine nonlinear associations between birth measurements and parental diabetes. Differences in associations between the sexes were examined using interaction terms (sex 3 birth measurement) in these regression models. Data were analyzed using Stata version 10 (Stata Corporation, College Station, TX). RESULTSdAnthropometric characteristics of the offspring at birth and anthropometric and biochemical characteristics of the parents 9.5 years later are described in Table 1. At birth, boys were heavier, were longer, and had larger head circumferences and AMA than girls, whereas girls had larger triceps than boys (Table 1). There were no significant differences in neonatal measurements between babies whose parents did and did not take part in the study (data not shown). A total of 9% of fathers and 15% of mothers were obese (BMI .30 kg/m 2 ). Sixty-two (15.6%; 45 known plus 17 new) fathers and 22 (4.7%; 7 known plus 15 new) mothers were found to have developed diabetes. In three families (,1%), both parents had developed diabetes; these were included in both mother-offspring and father-offspring analyses. Maternal and paternal age was positively correlated with their 120-min glucose concentrations (r = 0.1; P = 0.01 for both). Both parents' adiposity (BMI and sum of skinfolds) was positively correlated with their 120-min glucose and fasting insulin concentrations and HOMA-IR (r = 0.2-0.6; P , 0.0001 for all). The prevalence of diabetes among obese fathers was 22% compared with 15% among nonobese fathers and 8% among obese mothers compared with 4% in nonobese mothers. Associations of neonatal measurements with maternal and paternal diabetes Among fathers, there were linear inverse associations of newborn birth weight, CHL, CRL, AMA, MUAC, and triceps, subscapular, and sum of skinfolds with prevalence of diabetes (Table 2). Among mothers, offspring birth weight and adiposity measures (MUAC and abdominal circumference and triceps, subscapular, and sum of skinfolds) showed U-shaped associations with diabetes prevalence. As in fathers, there was a linear inverse association of neonatal CHL with prevalence of maternal diabetes ( Table 2). All |
these associations remained similar after adjustment for the parents' current BMI and sum of skinfolds (Table 2) and additionally for maternal gestational glucose concentrations and adiposity (sum of skinfolds) during pregnancy (data not shown). There were no associations of neonatal leg length and head circumference with either paternal or maternal diabetes. There were no sex interactions in the associations between offspring birth measurements and prevalence of parental diabetes. Associations of newborn measurements with maternal and paternal insulin resistance There were inverse associations of all the newborn measurements except leg length Odds ratios (ORs) and P values are derived using logistic regression. *P adjusted for child's sex, gestational age, and parents' current age. †P adjusted for child's sex, gestational age, parents' current age, and current adiposity (BMI and sum of skinfolds). ‡P value for quadratic association between neonatal measurements and maternal diabetes, adjusted for the child's sex, gestational age, and mother's current age. xP value for quadratic association between neonatal measurements and maternal diabetes after additional adjustment for mother's current adiposity (BMI and sum of skinfolds). care.diabetesjournals.org DIABETES CARE, VOLUME 35, SEPTEMBER 2012 1887 with paternal HOMA-IR after adjusting for current paternal adiposity (Table 3). These associations persisted after additional adjustment for maternal adiposity and glucose concentrations during pregnancy (data not shown). Among mothers there were positive associations of neonatal birth weight, head circumference, AMA, and adiposity measures (MUAC and subscapular and sum of skinfolds) with maternal HOMA-IR (Table 3). All these associations were attenuated and no longer significant after adjustment for maternal current adiposity (Table 3). There were no associations of CHL, CRL, leg length, abdominal circumference, and triceps with maternal HOMA-IR. Although there was some evidence of U-shaped associations (the highest values in the highest and lowest birth-size groups for birth weight, MUAC, and AMA) ( Table 3), there were no significant nonlinear relationships or sex interactions in the associations of offspring birth measurements with either paternal or maternal HOMA-IR (Table 3). CONCLUSIONSdIn this prospective study of both parents, who were nondiabetic at the time of the pregnancy, we found an inverse relationship between measurements of their newborn babies and their diabetes risk 9.5 years later. Smaller newborn size, especially smaller soft-tissue measurements (weight, MUAC/ AMA, and skinfolds), was associated with an increased risk of diabetes in both parents and higher insulin resistance in fathers. In addition, higher birth weight and greater newborn adiposity predicted an increased risk of maternal diabetes, resulting in clear U-shaped associations between these newborn measurements and maternal diabetes risk. These findings persisted after adjusting for potential confounding variables. There were positive associations of neonatal birth weight, head circumference, AMA, and adiposity measures with maternal HOMA-IR, which seemed to be mediated by maternal current adiposity. Strengths of the study were its prospective design with continuous followup of the children, which enabled us to examine the glucose/insulin metabolism of their parents. Apart from birth weight, detailed neonatal anthropometric data were available that were based on direct measurements using standardized methods and not on recall. A limitation was loss to follow-up, which could have introduced selection bias. However, a high proportion of the original cohort (76% of mother-infant pairs and 64% of fatherinfant pairs) was followed-up, and children's birth measurements were similar among those whose parents did or did not participate in the study. Our finding of an inverse association between neonatal birth weight and paternal diabetes and insulin resistance is not new and is consistent with earlier studies in the U.K. (4,5), the U.S. (6), and Sweden (7,8). As described in the introduction, this is consistent with the fetal insulin hypothesis, whereby genes associated with impaired insulin secretion or sensitivity, and shared by both father and fetus, could result in impaired fetal growth and increased diabetes risk in the father. A number of genetic polymorphisms have been related to both newborn size and diabetes risk. Many studies, including a genomewide association meta-analysis and review, identified specific gene markers that are associated with reduced birth weight (20,21) or increased susceptibility to diabetes (22,23) or both (21,24,25). Evidence from white populations suggests that the genetic markers ADCY5, glucokinase, KCNJ11, CDKAL1, and HHEX-IDE are associated with both diabetes and lower birth weight by reducing insulin secretion (21,24,25). An alternative possibility to the fetal insulin hypothesis is that epigenetic changes in the sperm of men at risk for diabetes also can lead to reduced fetal growth. Although there is no data to support this in humans, animal studies suggest that epigenetic changes can be transmitted by fathers as well as mothers and may mediate effects of preconceptional paternal diet on metabolic parameters in the offspring (26). Our finding of a U-shaped relationship between offspring birth weight and maternal diabetes is similar to a study among 60-to 79-year-old women in the U.K. (14) and another among Swedish women (7). The lower part of the "U" could reflect the same genetic or epigenetic phenomenon described above for fathers, whereas the upper part of the "U" could reflect effects on fetal development of the intrauterine environment in a prediabetic mother. Offspring born to mothers with gestational glucose intolerance are macrosomic (12); following the delivery, these mothers develop insulin resistance and are at increased risk of developing diabetes later (13). Alternatively (reverse causality), variations in the fetal genome that alter fetal growth also could alter maternal appetite and metabolism (for example a larger fetus and placenta may stimulate greater maternal food intake), and this could lead to insulin resistance and glucose intolerance during pregnancy and a higher diabetes risk later (27,28). Previously reported associations between offspring birth weight and diabetes risk among mothers have been quite variable; therefore, although two studies have reported a U-shaped association (7,14), others have reported a linear inverse association (4), two have reported a linear positive association (6,8), and two found no association (10,11). Whether a U-shape is present or whether one or other arms of the "U" predominate would depend upon the prevalence of relevant genetic/epigenetic markers and the prevalence of gestational diabetes within particular populations. One of two studies reporting a positive association between birth weight and maternal diabetes was in the Pima Indians, who have exceptionally high rates of GDM (6). The same would be true of associations between maternal insulin resistance and offspring birth weight. In our study, this was a significant positive association, although there was evidence of a weak U-shape, consistent with the diabetes results. Striking exceptions to the neonatal measurements that were inversely related to diabetes risk in both parents were leg length and head circumference, which are both direct measurements of skeletal size. The newborn body components most consistently related to parental diabetes were the soft-tissue measurements (apart from birth weight, MUAC, AMA, skinfolds, and abdominal circumference). CHL and CRL also were inversely associated with diabetes in the parents, but this could reflect newborn fat or muscle, which influence these measurements on a stadiometer, especially CRL, because of increased buttock size. The main growth-promoting hormones during intrauterine life are insulin and IGF-I and IGF-II (29,30). Evidence from infusion experiments in animals suggests that insulin directly promotes the growth of fetal adipose and skeletal tissue, whereas IGF-I stimulates skeletal, but not adipose tissue, growth (30). The IGFs also have been linked to a lower diabetes risk (31). Our findings are therefore in keeping with an effect mediated by reduced insulin action or secretion, rather than IGF action/secretion. To conclude, in this Indian population, smallness in all body components at birth, except leg length and head circumference, and P values are derived using linear regression. †P adjusted for child's sex, gestational age, and parents' current age. ‡P adjusted for child's sex, gestational age, parents' current age, and current adiposity (BMI and sum of skinfolds). xP value for quadratic association between neonatal measurements and maternal HOMA-IR adjusted for the child's sex, gestational age, and mother's current age. {P value for quadratic association between neonatal measurements and maternal HOMA-IR after additional adjustment for mother's current adiposity (BMI and sum of skinfolds). care.diabetesjournals.org DIABETES CARE, VOLUME 35, SEPTEMBER 2012 predicts an increased risk of later diabetes in both parents; this suggests a genetic or epigenetic link between diabetes risk in either parent and reduced fetal growth in their children. In addition, higher birth weight and greater newborn adiposity predict an increased risk of maternal diabetes; this suggests either that prediabetic metabolic changes in the mother during pregnancy (other than her glucose concentrations) increase fetal adiposity or that fetal adiposity induces maternal diabetes. This study adds to very few other studies that have shown these two effects so clearly. COVID-19: lipid disruption is pushing the envelope A plethora of articles have been published on severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), and science has delivered, given the rapid develop-ment of vaccines and of novel antiviral therapeutics evaluated for their efficacy efficiently in platform tri-als. An unfolding story of interferon genetics and au-toantibodies has begun to help us parse the reasons for varied susceptibility to severe disease and sequencing, and tracking at a global level has allowed for the rapid detection of new variants as they emerge. For all that progress and the political allocation of substantial resource to the World Health Organization in support of future pandemic preparedness, we failed to deploy the social science that would have anticipated what is politely called “ vaccine hesitancy ” or the psychological strategies to cope with what remains a continuing problem. This phenomenon and the failure equitably to distribute the opportunity for vaccination and novel therapeutics at a global level represent the biggest challenges to concluding the pandemic of coronavirus disease 2019 (COVID-19). , mouthwash to disrupt the integrity lipid envelope that facilitates viral A plethora of articles have been published on severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), and science has delivered, given the rapid development of vaccines and of novel antiviral therapeutics evaluated for their efficacy efficiently in platform trials. An unfolding story of interferon genetics and autoantibodies has begun to help us parse the reasons for varied susceptibility to severe disease and sequencing, and tracking at a global level has allowed for the rapid detection of new variants as they emerge. For all that progress and the political allocation of substantial resource to the World Health Organization in support of future pandemic preparedness, we failed to deploy the social science that would have anticipated what is politely called "vaccine hesitancy" or the psychological strategies to cope with what remains a continuing problem. This phenomenon and the failure equitably to distribute the opportunity for vaccination and novel therapeutics at a global level represent the biggest challenges to concluding the pandemic of coronavirus disease 2019 (COVID-19). In the Journal of Lipid Research, Saud et al. (1) presented a therapeutic opportunity that involves neither vaccines nor drugs. It may translate into a cheap and effective approach of relevance, not just to COVID-19, but generally to viral pandemics of the future; swilling and gargling mouthwash to disrupt the integrity of the lipid envelope that facilitates viral infection and proliferation. Attention has been payed to the modulation of cellular entry afforded by linoleic acid binding to SARS-CoV-2 (2), to how the lipid composition of cell membranes can influence viral entry by mediating fusion or affecting receptor conformation, and how, upon infection, viruses can reprogram cellular metabolism to remodel lipid membranes and fuel the production of new virions (3). In addition, eicosanoids and sphingolipids have been implicated in the immune response evoked by the virus, which, when unrestrained, contributes substantially to the clinical phenotype of severe disease (4). While there have been articles describing the lipidomic response to SARS-CoV-2 infection (5, 6), these have been relatively few when compared with interrogation of the immunophenotype (7,8). How the high-and lowabundance lipids that are formed in response to infection integrate in a multiomic context to drive the immune response remains to be described. The novel contribution from Saud et al. (1) is a focus on the viral envelope. Because coronaviruses bud from the endoplasmic reticulum-Golgi intermediate complex and exit via lysosomal secretion, the lipid composition of their envelope might be expected to differ from that of cell membranes. Indeed, that is the case, perhaps in a way that affords therapeutic opportunity. Viral lipids were extracted following infection of both VeroE6 and A459 cells and analyzed using both targeted and untargeted LC/MS/MS approaches. |
VeroE6 cells were originally isolated from kidney epithelial cells extracted from an African green monkey, and A549 cells were isolated from the lung tissue of a 58-year-old Caucasian male with lung cancer. Across the two cell types, roughly 260 lipids were reproducibly detected. The most abundant were phosphatidylcholine, phosphatidylethanolamine (PE), and phosphatidylinositol, along with several respective lyso, and ether/plasmalogen forms, and the relative proportions differed somewhat across the two preparations. The most abundant fatty acids detected were 16:0, 18:0, and 18:1, whereas others again differed, depending on the cell type infected. The authors focus on the high content of aminoPhospholipids (aPLs) in the viral envelope and specifically on the very high percentage of phosphatidylserine and PE that is externalized (although this again differed somewhat between the cells infected) compared with mammalian cell membranes. The authors highlight the comparatively small degree of phosphatidylserine/PE externalized on the membranes of activated platelets, where they are sufficient to catalyze and support the assembly of the prothrombinase complex (9). Hence, they speculate a role for externalized aPLs in the viral envelope driving the thrombotic complications of severe COVID-19. Consistent with this, they show that virions dose dependently shorten the activated partial thromboplastin time, a measure of induced coagulation in vitro. Infection of endothelial cells with SARS-CoV-2 as such would afford a potential site of such a thrombogenic mechanism (10). *For correspondence: Garret A. FitzGerald, [email protected]. Having implicated the viral envelope as a potential contributor to one aspect of disease severity, they report surfactant (such as cetylpyridinium chloride)dependent differential toxicity and efficacy amongst seven oral rinses to suppress viral infectivity in VeroE6 cells in the presence of a soil load to mimic the components of the nasal/oral cavity. Finally, they attempt to assess the ability and persistence of several rinses to suppress oral viral load in patients infected with SARS-CoV-2 in a randomized trial. Unfortunately, their trial design was undermined by a rapid decline in cases during their study such that they could only gather preliminary information. However, this was encouraging and provided data consistent with what had been seen in vitro, suggestive of a differential efficacy of this approach, at the least sufficient to permit safe oral examination. The authors are to be applauded for bringing delayed attention to the lipid composition of the viral envelope. Their suggestion of a prothrombotic contribution from its disproportionate content of exposed aPLs is provocative although many factors of relevance-blood flow, relative abundance compared with activated platelets in a developing thrombus, and perhaps most importantly, the modulatory effect of the concomitant inflammatory response on envelope lipids as viruses infect native epithelial cells of the respiratory tract-remain to be addressed. Thrombotic events complicate severe COVID-19, but no more so than in other severe viral illnesses (11), and low-dose aspirin does not influence the clinical course when begun early after likelihood of hospitalization (12). Rinsing and gargling may reduce viral load in the oropharynx, but infection in the respiratory tree already established before this intervention will be unaltered. If these rinses can be shown safely to reduce oropharyngeal viral load in a suitably powered clinical trial, this may prove to be a simple and perhaps broadly acceptable approach to attenuating the consequences of infection, not only of SARS-CoV-2 but also of other viruses. Most importantly, we would then have a therapeutic intervention, perhaps acceptable to the "vaccine hesitant" but certainly one that is economically practical to make available across the world. Conflict of interest The author declares no conflicts of interest with the contents of this article. Microbial Stimulation and Succession following a Test Well Injection Simulating CO₂ Leakage into a Shallow Newark Basin Aquifer In addition to efforts aimed at reducing anthropogenic production of greenhouse gases, geological storage of CO2 is being explored as a strategy to reduce atmospheric greenhouse gas emission and mitigate climate change. Previous studies of the deep subsurface in North America have not fully considered the potential negative effects of CO2 leakage into shallow drinking water aquifers, especially from a microbiological perspective. A test well in the Newark Rift Basin was utilized in two field experiments to investigate patterns of microbial succession following injection of CO2-saturated water into an isolated aquifer interval, simulating a CO2 leakage scenario. A decrease in pH following injection of CO2 saturated aquifer water was accompanied by mobilization of trace elements (e.g. Fe and Mn), and increased bacterial cell concentrations in the recovered water. 16S ribosomal RNA gene sequence libraries from samples collected before and after the test well injection were compared to link variability in geochemistry to changes in aquifer microbiology. Significant changes in microbial composition, compared to background conditions, were found following the test well injections, including a decrease in Proteobacteria, and an increased presence of Firmicutes, Verrucomicrobia and microbial taxa often noted to be associated with iron and sulfate reduction. The concurrence of increased microbial cell concentrations and rapid microbial community succession indicate significant changes in aquifer microbial communities immediately following the experimental CO2 leakage event. Samples collected one year post-injection were similar in cell number to the original background condition and community composition, although not identical, began to revert toward the pre-injection condition, indicating microbial resilience following a leakage disturbance. This study provides a first glimpse into the in situ successional response of microbial communities to CO2 leakage after subsurface injection in the Newark Basin and the potential microbiological impact of CO2 leakage on drinking water resources. Introduction Global water scarcity and the reduction of industrial CO 2 emissions are considered to be two of society's major environmental management challenges for this century [1]- [7]. Among currently proposed CO 2 mitigation techniques, geological storage (GS), accomplished by injection of CO 2 into deep geological formations, is one of the most promising alternatives [8][9][10][11][12][13]. Progress with implementation of this technology has been slowed, in part, by concerns over poorly understood alterations to subsurface water resources. In addition to the potential alteration of groundwater geochemistry, CO 2 leakage from deep formations to shallow drinking water aquifers has the potential to alter microbial communities by reducing pH and introducing, directly or indirectly, alternative substrates for microbial growth. Microbial life in the subsurface is thought to represent a globally significant reservoir of biodiversity that remains largely unexplored [14][15]. Initial characterizations of subsurface microbial communities suggest that geochemistry has a controlling influence on microbial community structure (e.g. [16]) but it is also known that microbial communities have an influence on the rate of geochemical reactions (e.g. [17]). In a survey of the bacterial communities in drinking water wells, temperature and iron concentration were determined to be controlling factors of community structure and composition [18]. Even slight increases in Fe concentrations resulted in modified bacterial communities and promoted the growth of iron oxidizing bacteria, which are known to foul well pumps and degrade water quality [18]. A similar survey across the pristine Mahomet aquifer of east-central Illinois found that sulfate concentration was an important indicator for the balance between sulfate reducers and iron reducers as dominant components of the microbial community [19]. Interactions between microbes and geochemistry remain poorly constrained in most natural systems targeted for geological sequestration of CO 2 and the in-situ successional response of microbes in shallow drinking water aquifers to a CO 2 leakage event is still poorly constrained. The experiment reported in this paper was designed to simulate an unintended leakage or migration of dissolved CO 2 into a shallow drinking water aquifer after CO 2 injection for the purpose of geological storage. Once in the subsurface, the injected CO 2 may be transformed by geochemical reactions controlled by the chemical composition of the formation water, pressure, temperature, rock mineralogy [20][21][22], and microbial biogeochemistry [23]. As a result of these reactions and physical mixing the formation fluid would be initially acidified, followed by a potential neutralization. Subsequent leakage of the acidified formation fluid into shallower drinking water aquifers would be expected to shift pH and, presumably, lead to succession of microbial community structure and metabolism. Onstott [23] predicted rapid successional patterns linked to altered availability of electron donors and acceptors, including the potential for increased microbial metal reduction and methanogenesis. Previous laboratory-based investigations suggest that exposure to CO 2 results in initial declines in bacterial abundance, but survival and adaptation of a stress resistant component of the community [24]. The CO2SINK project, an in situ CO 2 injection designed to systematically monitor the microbial response to elevated concentrations of CO 2 , observe a shift from sulfate reducers to methanogens following a high pressure CO 2 injection into a saline aquifer [25][26]. Bacterial taxa can respond quite differently to changes in pH, with methanogens displaying broad tolerance to pH variability [27], while other bacterial genera like Acidothiobacillus have growth optima at low pH, as seen in acid mine drainage [28]. Strong selective pressures favor broad tolerance to acidic conditions for some pathogens, such as E. coli, which are known to have diverse physiological adaptations that allow resistance to intestinal pH ranging from 2.0 to 7.0 [29]. Many of these pathogens prefer to grow at near neutral pH, such as those generally found within aquifers, but can persist and remain infective under variable pH [30], potentially allowing them to rebound from a subsurface pH disturbance more quickly than many other bacterial taxa. Despite the potential for large changes in microbial composition and metal mobilization in aquifer drinking water, the microbially mediated biogeochemistry in response to elevated CO 2 remains understudied compared to engineering and geochemical constraints [22] and to date there have been no in situ assessments of the microbial response to CO 2 leakage into drinking water aquifers. To begin bridging this gap, we conducted the first microbial characterization of Newark Basin subsurface bacteria from a shallow drinking water aquifer subjected to repeated CO 2 leakage scenario experiments. Using large-scale 16S rRNA gene sequencing from aquifer water samples collected during the two field injections, we attempted to address the following questions: 1) Are there observable microbiological responses to geochemical changes in a shallow drinking water aquifer following GS and potential CO 2 leakage?; 2) What are the successional responses of the aquifer microbial communities?; and 3) Is it possible to identify microbial groups, over-represented in the early-and mid-phases of the leakage experiment, whose changing abundance may act as useful indicators for acidification from GS and potential CO 2 leakage in the Newark Basin? Study site and injection experiment Description of study site and injection experiment. Located in Palisades New York (41.0039°N , 73.9126°W), the Lamont-Doherty Earth Observatory test well (TW-3) penetrates the Palisades Sill and extends into the underlying sedimentary rock formations of the Newark Basin. The Newark Basin interbedded sedimentary layers are abundant in iron oxide minerals and include sandstone, siltstone, and mudstone [31]. The Palisades Sill, a diabase intrusion, is approximately 230 m thick at this location, capping metamorphosed sedimentary formations [32][33]. The injections for this study were targeted to a low transmissivity (0.023 m 2 day -1 ) permeable zone at 364 m with no detectable ambient flow, allowing for an appropriate duration for experimental incubation of CO 2 saturated waters within the formation. As described by Yang et al [34], two push-pull experiments were conducted into an isolated interval at 362-366 m depth in the test well during summers of 2011 and 2012. The interval was isolated by an inflatable packer system, and for several days prior to each injection experiment, formation water was pumped out at a flow rate of~3.5 L min -1 to estimate interval transmissivity and characterize background conditions of the interval. During these pumping tests, continuous sensor measurements were made for pH, specific conductance (SC), dissolved oxygen (DO), and oxidation-reduction potential (ORP). While characterizing background conditions the pumped formation water was also used to extensively rinse three previously cleaned polyethylene tanks at the test well site, and then 3,400 liters of the formation water was captured in the tanks for acidification with CO 2 and mixing with chemical tracers for injection. Once in the tanks, the formation water was slowly bubbled with CO 2 for 16 hours to achieve saturation at one atmosphere partial pressure and Potassium Bromide (KBr) was added to a final concentration of 45.53 mg L -1 as a tracer. The KBr tracer allowed hydrogeological flow and mixing patterns of injection water with background formation fluid |
to be constrained, improving interpretation of geochemical and microbiological dynamics during the in situ experiment. Approximately 3,050 liters of the CO 2 acidified aquifer water was then injected from the holding tanks back into the isolated interval, at a flow rate of 4.6 L min -1 over an 11 hour period to simulate leakage, or unintended migration, of fluids from a GS site into a shallow drinking water aquifer. The CO 2 bubbling of fluid in holding tanks continued during the injection to maintain CO 2 saturation. Immediately following injection of the acidified fluids, 70 liters of unaltered aquifer water, that had been temporarily held in a fourth polyethylene tank without CO 2 bubbling, was injected as a "chaser" to clear the injection tubing and to ensure full injection of the 3,050 liters of acidified aquifer fluid into the targeted subsurface interval. The 70 liter chaser was approximately twice the volume of the injection tubing. The injection interval, between 362 and 366 m depth, is more permeable radially than the intervals above and below and remained sealed by the packer system, eliminating vertical exchange within the test well, during the entirety of the background aquifer water pumping, injection, in situ incubation, and extraction "pump-back" sampling. The pressure above and below the packed off interval within the well was monitored during injection, incubation, and pumping to ensure that flow did not re-enter the well from another interval, but no pressure anomaly was detected, suggesting that the injected fluids remained within the formation injection interval. To simplify description and analysis of the data, both experiments have been divided into experimental phases labeled "Background", "Early", "Mid", and "Late", which corresponded to the ratio of the volume of water pumped (Vp) out versus the volume of water injected (Vi) into the aquifer (as per Table 1). "Background" samples were collected pre-injection while water was being pumped out of the test interval prior to acidification via CO 2 bubbling, but only after environmental parameters including pH, dissolved oxygen (DO), and oxidation-reduction potential (ORP) stabilized for the pumped aquifer fluid. "Early" samples were taken postacidification and post-injection, during the initial extraction pump-back and while the ratio of volume pumped/volume injected was less than 1. "Mid" samples were taken during the later stages of pump-back when the ratio was greater than 1 but less than 10. "Late" samples were taken after the ratio was greater than 10. During the 2011 experiment, an initial sample was collected seven days after the injection to measure the tracer concentration, pH, and specific conductance, by pumping to extract approximately 90 liters water. This sample was considered part of the "Early" phase, even though continuous pump-back had not begun. Before any discrete sample was collected and evaluated by sensor or laboratory processing, a volume greater than the internal volume of the injection/ recovery tubing was purged, so that only fluids incubated in the test interval were evaluated. This small volume discrete sample at 7 days was intended to create minimal disturbance to the in situ incubation, but it allowed an initial evaluation of the mixing and transport of injected fluids in the test interval so that a decision could be made about the appropriate duration of in situ incubation before continuous pump-back. It was decided that the in situ incubation, prior to continuous pump back of the interval fluids, would last for 20 days after the injection. This time interval was viewed as a balance between providing adequate time for microbial and geochemical succession to occur and also permitting adequate recovery of injected fluids and avoiding complications from extensive mixing and dilution with waters surrounding the test zone. Twenty days after incubation, recovery of injected fluid began with continuous pumping at 2.6 L min -1 for 33 days, equivalent to 40 times the injected volume, consisting of the injected fluids mixed with background interval water. Measurement of the conservative KBr tracer allowed the fraction of initial injected fluid versus background water to be evaluated in the recovered fluids. During this pump-back 20 sets of samples were collected to characterize geochemical alteration and microbial succession. It should be noted that dilution of the injected fluid, geochemical alteration (e.g. pH), and length of the incubation are correlated factors in this experiment. A repeat of the experiment occurred in the summer of 2012, following the same procedure, however, the in situ incubation lasted for 40 days to allow additional time for microbial succession following the injection disturbance, and the continuous pump-back lasted for 30 days. Ethics Statement. Authorization for the injection experiment was obtained from the United States Environmental Protection Agency Underground Injection Program (UIC ID: 04NY08707045). No animals or endangered species were involved in this research. Geochemical and microbiological sample collection Geochemical sampling was conducted as described by Yang et al [34]. Briefly, groundwater was filtered through 0.45 μm membranes for major anions, filtered and acidified to 1% HNO 3 (Fisher Optima) for major cations and trace elements, and unfiltered water was used for Brtracer analysis. A YSI multi-parameter meter was used with a flow through cell to monitor temperature, pH, SC, ORP and DO concentrations throughout the injection experiments. Samples for microbiological analyses were collected simultaneously with the geochemical samples using sterilized collection containers. At each sampling point, 50 ml aliquots were formaldehyde-preserved (3% final concentration) in sterile tubes and stored at 4°C until microbial concentrations were microscopically ascertained. Approximately 10 L of aquifer water were passed through duplicate 0.2 μm "sterivex" cartridge filters (Millipore, Darmstadt, Germany) that were immediately flash-frozen in liquid nitrogen and then stored at-80°C until molecular analyses were performed. One L of unfiltered water was transported to the laboratory, in the dark and on ice, to allow for measurement of Fecal Indicator Bacteria (FIB) within 6 hours of collection. Microscopic cell count, DNA concentration, and fecal indicator bacteria To characterize the microbial response to CO 2 leakage after GS, microbial cell concentrations were determined for each sample through the course of the injection experiments. Cell abundances were microscopically determined from formaldehyde fixed aquifer samples using SYBR Green staining, according to Noble et al [35]. The concentration of DNA extracted from samples (see below) was used as a second, independent, method of estimating microbial abundance in the samples to evaluate the microbial dynamics and to confirm patterns observed with microscopic cells counts. DNA was extracted from "sterivex" filters using the PowerWater DNA isolation kit (MoBio Laboratories, Carlsbad, CA), according to manufacturer's instructions. DNA concentration of extractions and subsequent amplifications were determined using a Qubit fluorometer (Invitrogen, Grand Island, NY, USA). Enumeration of the FIB, E. coli and Enterococcus, were performed from 100 ml of unfiltered water using IDEXX enterolert and colilert selective media (www.idexx.com). FIB were enumerated to test for contamination with surface water and also because, as a commonly used indicator of drinking water quality, the response of FIB to CO 2 leakage events was of interest. Microbial community successional dynamics DNA amplification and sequencing. Bacterial primers 8F (5 0 -AGRGTTTGATCCTGGCT-CAG-3 0 ) and 1492R (5 0 -CGGCTACCTTGTTACGACTT-3 0 ) [36] were used with 35 PCR cycles of 45 seconds of denaturation at 94°C, 45 seconds of annealing at 55°C, and 1 minute elongation at 72°C, followed by gel electrophoresis, to initially confirm that extracted DNA was PCR-amplifiable and to ensure that DNA-free controls did not yield amplification product. Subsequently, bacterial and archaeal community composition were determined using amplicon pyrosequencing performed at Molecular Research DNA labs (www.mrdnalab.com, MRDNA, Shallowater, TX, USA), using a protocol described by Dowd et al [37] which has been utilized in describing a wide range of environmental and health related microbiomes [37][38][39]. Briefly, parallel sequencing reactions were prepared from each DNA extraction using both the eubacterial primer 27F and archaeal primer 349F. Amplification was performed through a single-step 30-cycle PCR using HotStarTaq Plus Master Mix Kit (Qiagen, Valencia, CA). PCR was performed under the following conditions: 94°C for 3 minutes, followed by 28 cycles of 94°C for 30 seconds; 53°C for 40 seconds and 72°C for 1 minute; after which a final elongation step at 72°C for 5 minutes was performed. All amplicon products were then mixed in equal concentrations and purified using Agencourt Ampure beads (Agencourt Bioscience Corporation, MA, USA.) and sequenced using Roche 454 FLX titanium instruments and reagents, following manufacturer's guidelines. Sequence read files and associated sample data are available from the National Center for Biotechnology Information database (www.ncbi.nlm.nih.gov/sra) under Bioproject PRJNA258542, including Sequence Read Archive accession numbers SRX729863 and SRX732154 to SRX732204. Analysis of DNA Sequence Data. To generate a high-quality dataset for analysis, DNA sequences were processed with the Mothur platform [40][41]. PyroNoise was used allowing for 1 mismatch to the barcode and 2 mismatches to the forward primer. Sequences were then trimmed to remove barcodes and primer sequences, and sequences with less than 200 bp and/ or homopolymers greater than 8 bp were removed from downstream analyses. Remaining sequences were aligned using SILVA reference alignments and chimeras were detected and removed using UCHIME [42]. Sequences were then taxonomically classified using the Mothur-formatted version of the Ribosomal Database Project training set [43], at 80% cutoff, then binned by genus to create phylotype OTU's. The PC-ORD software package (version 4.01; MJM Software) [44] was utilized for multivariate analyses of microbial community and environmental data. Before analysis, OTU data were relativized for sampling effort, as is commonly required in microbial ecology [45], by dividing the number of sequences for an OTU by the total number of sequences from the sample, creating a relative frequency for each OTU. Environmental data were normalized by dividing each variable by the maximum observed value across samples, placing all environmental values on a relative scale of zero to one. Environmental factors that initially contained negative values, such as ORP, had a number equal to the most negative value observed from any sample added before normalization to avoid inclusion of negative values that can be incompatible with some downstream analyses. Hierarchical cluster analysis [46] was conducted for both environmental factors and community composition using Sorenson (Bray-Curtis) distance and farthest neighbor group linkage to examine the similarity of samples. Environmental factors used in the clustering analysis included: pH, specific conductance, ORP, pCO 2 , sulfate, manganese, and iron. A Mantel test [47] was used to compare the structure in the observed data to 1,000 randomizations of the same dataset, allowing distance matrices from environmental factors and community composition to be examined for non-random structure. A Multi-Response Permutation Procedure (MRPP) was used to test for differences among a priori selected groups, corresponding in this case to samples grouped by phase of the experiment (i.e. "Background", "Early", "Mid" and "Late"; Table 1). Non-metric Multidimensional Scaling (NMS) was used to visualize the similarity of samples based on DNA sequence composition analyzed at the level of genera and to examine patterns in the groups of samples corresponding to phase of the injection experiment. A NMS Scree plot was used to determine that the ordination was optimized using three dimensions, with significantly lower NMS stress (p = 0.0196) for the observed data compared to 1,000 randomizations of the data set, with the majority (0.887) of variation described by these three axes. Indicator Species Analysis, also with samples grouped based on phase of injection, was used to detect genera of specific interest for future environmental monitoring for GS [48]. The value of an OTU as an indicator for a particular phase of the experiment was determined by the proportional abundance of an OTU relative to the abundance of the OTU in all other groups examined. The statistical significance of these indicator values is determined by comparison of observed indicator values to indicator values resulting after 1,000 randomizations of the data set. Results Injection experiment and geochemical context CO 2 injection into the aquifer produced clear geochemical responses during both field injection experiments. Background aquifer water showed a Na-SO 4 water type. Pre-injection bubbling with CO 2 in storage tanks did not change major ion concentrations or alkalinity, but resulted in an increase in the total dissolved CO 2 (= [H 2 CO 3 ] + [HCO 3 -] + [CO 3 = ]). During both experiments, Background (pre-CO 2 bubbling) samples were slightly basic (pH 8.2-8.5) (Fig. 1a,b), and microaerobic (DO = 1.0-1.5mg L -1 ). Injection of the CO 2 -saturated water |
(pH of approximately 4) resulted in an initial sharp decrease in aquifer pH, with an eventual return to pre-injection conditions (Fig. 1a,b). The KBr tracer allowed estimation of the mixing ratio of the injected high CO 2 water with aquifer water as well as the total percent recovery of the injected water during pump-back extraction phases (Early, Mid, and Late). Recovery of injected water was estimated to be 73% in the Injection 1 experiment and 79% in the Injection 2 experiment, with the KBr concentration approaching zero (Fig. 1c,d), and injected fluid estimated to contribute less than 1% of the concentration in the final Late phase fluid samples. Water samples collected after injection, during the Early extraction phase of the experiment (pumped/injected volume ratio < 1), demonstrated major geochemical alterations, with notable increases of HCO 3 -, alkalinity, Ca, Mg and Si concentrations, increased by 2-12 times over Background phase concentrations. Sulfate initially decreased by 20-30% (Fig. 1e,f) while both Fe (Fig. 1g,h) and Mn [34] immediately increased in concentration during both experiments along with an initial doubling of conductivity, which also returned slowly toward pre-injection conditions (Fig. 1i,j). Microbiological response to geochemical change As geochemical conditions, such as pH, changed in the early phase of the experiment, it was expected that microbial communities would respond to these alterations, providing potential biogeochemical interactions with metabolically relevant factors such as ORP (Fig. 2a,b), Fe (Fig. 1g,h), and sulfate (Fig. 1e,f). Background microbial cell concentrations were low in both injection experiments with a mean of 4.0 x 10 4 cells ml -1 , and decreased slightly during CO 2 . Dashed and shaded lines on x-axis separate: "Background" aquifer fluid prior to bubbling with CO 2 ; "Injection" fluid sampled during the injection, following CO 2 bubbling but prior to in situ incubation; and "Early", "Mid", and "Late" phases of the in situ experiment, with longer incubation represented by a higher ratio of volume pumped to volume injected (see Table 1). doi:10.1371/journal.pone.0117812.g001 bubbling and injection, but spiked significantly after injection (to 2.0 x 10 5 in Injection 1 and 3.5 x 10 5 in Injection 2). These concentrations then returned to background concentrations (Fig. 2e,f) in the later stages of recovery. The concentration of cells and DNA extracted from samples (Fig. 2c,d) were positively associated. Assuming 2 fg of DNA per cell [49], the pattern of microbial abundance estimated by microscopy and DNA concentration were in close agreement for both injections, displaying very similar dynamics that confirm an initial peak in microbial abundance in the early phases of recovery that approached background concentrations in later phases of the experiment (Fig. 2c,d). After a quality check (see methods) of 454 sequencing results, 150,417 bacterial sequences and 31,109 archaeal sequences were maintained for downstream taxonomic analyses (Table 1). These sequences revealed a diverse and dynamic microbial community in aquifer water over the course of both injection experiments, including 633 OTUs that were identified at the genus level for most downstream analyses. Hierarchical cluster analyses were performed using both environmental factors and microbial community composition (S1 Fig. and S2 Fig.). Cluster analyses of environmental factors demonstrated distinct clusters for Early phases of the experiment (S1 Fig.), while cluster analyses of microbial communities demonstrated distinct clusters for both Background and Early phases of the experiment, indicating substantial microbial succession following the initial injection disturbance. A mantel test detected significant (r = 0.373; p = 0.001), non-random, structure between distance matrices from microbial community composition and environmental conditions, supporting an interaction between environmental conditions and microbial composition in this in situ aquifer disturbance experiment. Microbial successional dynamics Phylum-level patterns of aquifer bacterial community alteration were similar in the two injection experiments (Fig. 3). Before both injections, background aquifer samples were dominated by Proteobacteria, but after CO 2 injection Verrucomicrobia, Crenarchaeota and Bacteroidetes had increased representation in the sequence libraries. It is worth noting that in both injections, microbial communities did not return to pre-injection conditions on the phylum level even though the tracer concentration indicates that injected fluids constitute 1% or less of the final Late phase fluid. One year after the first injection disturbance, the Background condition in 2012 (Injection 2) was similar, although not identical, to the 2011 (Injection 1) Background community (Figs. 3 and 4), indicating some degree of microbial community resilience on longer time scales. MRPP analyses, conducted at the level of genera, indicate that significant differences in community composition occurred among the four phases of the experiment (p < 0.01). Dominant genera (< 25% in any single sample) in aquifer waters changed dramatically after CO 2 injection in both experiments (Fig. 4), with Sphingobium decreasing in relative abundance immediately after both injections and Verrucomicrobia increasing after both injections. Dechloromonas peaked, after a lag, in the Mid phases of both Injections 1 and 2, with a gradual decline in the Late phases. Rhizobacter was more dominant after Injection 1, while Thermosphaera had much higher relative abundances after Injection 2. Therefore, changes in the microbial community following injection were evident at both the levels of phyla and genera and similarity of samples to initial Background condition (prior to Injection 1), evaluated using the Bray-Curtis similarity index (Fig. 4), decreased post-injection in both experiments, demonstrating rapid community succession in response to the experimental disturbance. NMS (Non-metric Multidimensional Scaling) ordination of injection samples in OTU space (Fig. 5) displays patterns generally consistent with hierarchical clustering and MRPP analyses, and demonstrates the rapid change in microbial community composition that occurred in the Early and Mid phases of the injection experiment. Background samples form groups distinct from all other phases, most notably the Early phase samples following the injection disturbance. The community composition immediately following the 2011 and 2012 injection disturbances are similar (Figs. 4 and 5), with the ordination of Early and Mid phases demonstrating a shift across axis I, correlated with the changing pH for these samples. In contrast, samples collected from the Late phases of the two injections are more distinct from one another as compared to the Background samples, demonstrating a change in the microbial community that remained weeks after injection, even after >98% of the injected fluid had been recovered and environmental conditions such as pH and pCO 2 had returned to near Background levels. The two experiments were performed in the same isolated aquifer interval approximately one year apart, and therefore, the Background sample from the 2012 injection represents a longerterm (one year) indication of succession from the prior year's disturbance. While the background samples from the two injections are not identical, they do share the same dominant phylum (Proteobacteria, Fig. 3) and genus (Spingobium, Fig. 4) and group together in the NMS analysis and in hierarchical cluster analyses, indicating that the OTU composition of the microbial communities converge, reverting back toward the original Background assemblage, within one year of disturbance. The patterns of succession summarized by NMS were also evident in the patterns of individual genera, with dominant components of the bacterial and archaeal communities shifting significantly post-injection. For instance, Hydrogenophaga and Sphingobium, dominant in both Background samples (15% and 30% of background sample sequence libraries, respectively), Fig. 4). Samples are grouped by "phase" of the injection experiment (see Table 1) with biplot overlay vectors (e.g pH, Sulfate) indicating the direction and relative magnitude of association between the environmental variables and the two axes used to ordinate community composition. Fig. 1, the x-axis is a ratio of the water volume pumped (Vp) out of the aquifer during the experiment and the CO 2 saturated water volume injected (Vi) at the start of the experiment. Dashed and shaded lines on the x-axis separate: "Background"; "Injection"; "Early"; "Mid"; and "Late" phases of the in situ incubation. 2 (B, D)). Vp/Vi = Volume Pumped/Volume Injected. As in Fig. 1, the x-axis is a ratio of the water volume pumped (Vp) out of the aquifer during the experiment and the CO 2 saturated water volume injected (Vi) at the start of the experiment. Dashed and shaded lines on the x-axis separate: "Background"; "Injection"; "Early"; "Mid"; and "Late" phases of the in situ incubation. significantly decrease in representation in the sequence library post-injection (both < 5%, Figs. 4 and 6). Dechloromonas, Verrucomicrobia, Thermosphaera, and Sulfophobococcus, (all < 5% in background samples), become relatively more dominant (< 10%) in sequence libraries post-injection (Figs. 4 and 7). CO 2 injection also increased representation of biogeochemically-relevant genera, including genera that have been associated with metal-reduction like Geothrix, Geobacter, Desulfosporosinus, and Sulfophobococcus (Fig. 7), which responded to CO 2 injection, or the associated environmental changes, by sharp initial relative increases in library representation, then gradual return to near background levels in both field experiments. Genera associated with acid tolerance, and iron and sulfur oxidation, including Acidovorax, Ferroplasma, Acidothiobacillus and Ferrimicrobium, increased representation in sequence libraries immediately post-injection (Fig. 8), whereas Leptospirillium did not increase. Methanogens including Methanomicrobia and Methanobacteria also increased in library representation post-injection (Fig. 8). An Indicator Species Analysis was conducted to evaluate the successional patterns from the perspective of future environmental monitoring for acidification and CO 2 leakage into shallow drinking water aquifers due to GS. Twenty-nine genera were found to have significant (p-value < 0.05) use as indicators for the Early and Mid phases of the experiments, evident from their increased relative frequency in these phases ( Table 2). Many of the possible indicators included genera previously associated iron and sulfate reduction, such as Geobacter, Geothrix, Desulfosporosinus, Dechloromonas providing additional evidence that microbial transitions following 2 (B, D)). Vp/Vi = Volume Pumped/Volume Injected. As in Fig. 1, the x-axis is a ratio of the water volume pumped (Vp) out of the aquifer during the experiment and the CO 2 saturated water volume injected (Vi) at the start of the experiment. Dashed and shaded lines on the x-axis separate: "Background"; "Injection"; "Early"; "Mid"; and "Late" phases of the in situ incubation. doi:10.1371/journal.pone.0117812.g008 CO 2 leakage events respond to changes in chemistry and would be expected to influence biogeochemical activity and potentially water quality. Assays for viable FIB, Enterococcus and E. coli, were all negative indicating that the subsurface interval tested lacked any recent surface-associated sewage contamination. Similarly, enteric bacterial genera, such as Escherichia, Enterococcus, and Enterobacter, were either absent or extremely rare in all sequenced samples. We cannot adequately evaluate the successional response of enteric or sewage associated microbes to in situ acidification from this experiment and this remains an open question for future studies in contaminated drinking water aquifers. Observed disturbance from injection Conditions in the Early and Mid phases of this field experiment, including pH, conductivity, sulfate, and Fe concentration (Fig. 1) [34], indicate highly altered geochemistry near the borehole injection interval. Based on prior laboratory [24], [50], [51] and field injection experiments [52][53][54][55][56][57] it was expected that water-rock interactions would quickly (on a timescale of hours to days) alter the chemistry of the injected fluids during incubation, including acid neutralization, mineral dissolution, and precipitation of carbonate minerals. As the CO 2 saturated fluids were injected, the chemical changes induced by injected fluids presumably decreased with increasing distance from the well due to mixing with the aquifer water [52]. The spatial variability in the extent of disturbance was observed as temporal variability in our experimental data (comparing Early vs Late phases of pump-back). Similar geochemical gradients would be expected as CO 2 impacted fluids migrated from the source of a hypothetical GS reservoir leakage event into a shallow potable water aquifer. The first recovered fluids ("Early" phase) had pH values as low as 5.7, while the injected fluids had a pH of approximately 4, indicating mixing and neutralization near the borehole. In the late phase of the experiment pH increased to near background levels, consistent with expectations and inversely related to the bromide tracer concentration (Fig. 1a-d). Major and trace elements including Ca, Mg, Si, Fe, Mn, Cr, Co, Ni, Cu, Zn, Rb, Sr, Ba, and U displayed rapidly increasing concentrations during the early phase of the experiment, as compared to background conditions [34], with release rates from the experiment similar to those observed in batch reactor experiments [58]. Microbiological response to CO 2 injection and altered geochemical conditions Cell concentrations in Background water samples, estimated both microscopically and with DNA extractions (Fig. 2c-f), were low but comparable to measurements made in many |
other aquifers [19], [59][60][61]. Cell concentrations rose significantly in the Early phase of injection, potentially driven by altered metabolic substrates produced by the acidification disturbance, and then declined in the Mid and Late phases to approximately Background levels. ORP also decreased rapidly in the Early phase (Fig. 2a,b), likely as a response to the microbial stimulation indicated by rising cell counts, creating additional environmental variability that would cause microbial succession. The dropping ORP created conditions more favorable for reductive microbial metabolism and may have acted as a positive biogeochemical feedback on the observed metal mobilization (e.g. Fe) and reduction of sulfate. The CO2SINK experiment [25][26] in northeast Germany examined the microbial response to a CO 2 manipulation of a saline aquifer in siltstone, sandstone and mudstone of the Triassic Stuttgart Formation, which has a similar geological age to the Newark Basin formations, but is located at much greater depth (700-850 meters) below the surface. In Germany, the background cell concentrations were found to be approximately 10 6 cells ml -1 , noted to be at least an order of magnitude higher than most deep anaerobic aquifers previously measured [26]. Two days prior to CO 2 injection, the CO2SINK test wells were flushed with N 2 gas (referred to as "N 2 lift") to prepare them for CO 2 injection. Cell counts were found to decrease by three orders of magnitude following the N 2 lift and cell counts in the observation wells remained one order of magnitude below background levels for approximately 2 months following the subsequent CO 2 injection [26]. This is in contrast to the results presented in this paper, where cell counts were found to initially increase (Early phase) compared to Background conditions. The difference may be due to the fact that the CO2SINK injection was simulating a high pressure, super-critical CO 2 injection into a saline aquifer, while our experiment simulated leakage of diluted fluids from a CO 2 storage reservoir into a shallower potable aquifer. Interestingly, the pH and metal mobilization response of both experiments were similar, despite the different microbial response. Morozova et al [26] explain the decreased microbial abundances from CO2SINK as a cellular stress response expected due to acidification and the subsequent return to background levels as adaptation to these conditions [62][63]. Interpretation of the microbial abundance changes following the CO2SINK experiment is problematic because of the major cellular reduction following the N 2 lift disturbance. Although Morozova et al [26] report the pattern as a decrease in cell abundances after CO 2 injection, when compared to the abundances following the N 2 lift disturbance, cell abundances in the Stuttgart formation in fact increased by more than an order of magnitude in the weeks following the injection of CO 2 . An increase in the percentage of active cells was reported in the months following injection. Abundances did not surpass the background, pre-injection levels, however. It is difficult to determine if the CO 2 injection acted to stimulate the microbial community and increased the speed of recovery following the N 2 lift, or if the CO 2 injection acted in concert with the initial N 2 lift disturbance to repress cell abundances below background levels. Based on the comparison of these studies it would be reasonable to expect that a high pressure, super-critical CO 2 injection may initially cause repression of the microbial community, while a GS leakage scenario into a shallow potable aquifer may result in short term microbial stimulation. Both types of disturbances were found to cause microbial succession and point to the resiliency of the sub-surface microbial community. In addition to changes in pH, CO 2 chemistry, and cell abundance, the geochemical conditions in the Newark Basin injection shifted in other important ways including increased trace element concentrations (e.g. Fe) and lowered ORP, oxygen, and sulfate. These factors could be both a cause of microbial change and an effect of microbial change. For example, increased microbial abundance and activity would be expected to contribute to lowered oxygen and ORP. Similarly, increased microaerobic and anaerobic microbial activity, driven by decreased oxygen and ORP, would be expected to reinforce iron mobilization and sulfate depletion via microbial iron and sulfate reduction. While it is impossible to attribute these geochemical changes to biological activity alone, biogeochemical activity from stimulation of subsurface microbes is consistent with these geochemical patterns. Microbial successional dynamics and microbial indicators of CO 2 leakage events The Background bacterial community from the Newark Basin test interval, before CO 2 injection, was dominated by Proteobacteria as has been observed in many shallow drinking water aquifers [18][19], [61], while the Archaea were dominated by Crenarchaeaota, similar to the pattern found by Lavalleur and Colwell [61]. The two dominant genera detected in background samples, Sphingobium and Hydrogenophaga, have also been previously detected in drinking water and pristine aquifers [18], [24], [61], [64][65][66]. Sphingobium is an obligately aerobic chemoheterotroph genus [64] and Hydrogenophaga is a facultative anaerobic chemoorganotrophic or chemolithoautotrophic genus capable of oxidizing hydrogen as an energy source using oxygen or nitrate as a terminal electron acceptor [67][68]. Proteobacteria remained the dominant phyla through the experiment despite the increase in relative abundance of Verrucomicrobia and Crenarchaeota following injection (Fig. 3). These data show that the subsurface microbial community can be rapidly altered in response to injection or leakage of CO 2 . The relative abundance of Sphingobium and Hydrogenophaga, decreased substantially following injection (Figs. 4 and 6), while Dechloromonas became the most abundant genus in the mid stages of recovery (Fig. 4). In addition to acidification stress, the decrease in Sphingobium may be due to their obligate aerobic metabolism, which would not be well suited to the decrease in oxygen and ORP following injection. It is less clear why Hydrogenophaga may be poorly suited to the conditions in the Early and Mid-phases of the experiment. Laboratory incubations simulating acidification of in situ fluids as part of the CO2SINK project also found Hydogenophaga to be abundant pre-CO 2 acidification, but undetectable afterward [24]. Although the community composition from the deep saline fully-anoxic aquifer from the CO2SINK experiments was, not unexpectedly, quite different from the shallower Newark Basin potable aquifer, the importance of genera that have been associated with sulfate reduction was evident in both systems. The Stuttgart formation was dominated by fermentative halophilic bacteria such as Halanaerobium, and sulfate reducing bacteria such as Desulfohalobium and Desulfotomaculum [26], [69]. None of these genera were detected in the Newark Basin potable aquifer samples, perhaps due to their preference for higher salt environments. Genera associated with sulfur reduction, including Dechloromonas, Desulfosporosinus, and Thermosphaera did become dominant groups in the Early and Mid-phases (Figs. 4 and 7) following the drop in pH, oxygen, and ORP. Although the specific taxa vary between systems, as would be expected based on salinity differences, the functional group dynamics inferred from the 16S rRNA gene sequence data were similar. Iron and metal reduction associated genera, such as Geothrix, Geobacter, and Desulfosporosinus (Fig. 7), also increased in relative abundance following acidification of in situ fluids in the Newark Basin experiment, indicating a transition expected under lower ORP and coinciding with the observed decreases in sulfate and increases in Fe and other trace metals. The peaks in iron and sulfate reducing genera tend to occur immediately after injection, while peaks in genera associated with iron and sulfur oxidation were often slightly delayed following injection (e.g. Ferroplasma, Acidothiobacillus; Fig. 7a,b and Fig. 8). Their abundance may be enhanced by the prior reductive activity, one of many possible syntrophic interactions. It is important to note that functionality of taxa was not directly measured and energetics alone should not be used to explain the microbial transitions, as the susceptibility to pH may be just as important in constraining successional patterns [26], [70], [71] and many of the genera peaking in the early and mid phases of injection are known to include acidotolerant or acidophilic taxa. Finally, methanogenic genera, including Methanomicrobia and Methanobacteria, also increased in their relative abundance in the early-and mid-phases of the injection, coinciding with decreases in sulfate that have also been observed in other subsurface systems (e.g. [19], [59]). The peaks of methanogens were higher in 2011, when ORP was much lower than observed in the 2012 injection. Increases in both iron reducers and methanogens were predicted by prior modeling efforts of induced alterations due to GS [23] and were observed in the CO2SINK injections [26]. Indicator Species Analysis detected 29 genera whose relative abundance in sequence libraries increased significantly during the Early or Mid phases of injection (Table 2). These genera provide insights into the potential for metabolic transitions linked to geochemistry (e.g. increased iron and sulfur cycling) and provide possible targets for indicator organisms that could be used in GS leakage monitoring programs. It is interesting that the majority of these genera (26 of 29) were identified from the Early phase of the experiment, indicating that the largest changes in microbial composition would occur quickly after a leakage disturbance. In contrast, Dechloromonas displays a large increase post-injection, but peaks in the Mid phase, and becomes the most abundant genera. Other indicators of note include genera associated with iron and sulfur reduction such as Geobacter, Geothrix, and Desulfosporosinus. The patterns of dominant genera (Fig. 4) and NMS analysis (Fig. 5) allow visualization of the changing microbial community, with similar patterns in the Background, Early and Mid phase communities across both experimental injections. In combination, these data demonstrate substantial and predictable succession in the microbial community, with increases in taxa matching the observed geochemical transitions and prior modeling efforts (e.g. [23]). The phylum and genus-level composition are consistent with a rapid transition from a background aquifer microbial community similar to those observed in low nutrient, pristine environments [60][61] to a microbial assemblage with increased representation of taxa previously found in acidified, metal rich environments (e.g. [72]). Rare organisms gained prominence postinjection, including some taxa associated with iron reduction, sulfate reduction, anaerobic iron oxidation and methanogenesis, demonstrating the potential for microbes to contribute to altered biogeochemical processes involved in trace gas production and metal mobilization. Potential significance for metal mobilization Oxidative and reductive microbial metabolism can influence the mobility of metals, such as uranium or arsenic, and are therefore important for understanding potable water quality. For example, Senko et al [73] found that an acid-tolerant strain of Desulfosporosinus isolated from acid mine drainage reduced U (VI) in groundwater more rapidly at pH 4.4 than 7.1. Bioremediation strategies have relied upon microbially mediated reduction of U (VI) to insoluble U (IV) end products to reduce their mobility in groundwater [74], while increased ORP and dissolved oxygen have been associated with U oxidation and mobilization [75]. In the Early phase of the injection experiment, U concentrations were found to increase 100 fold as compared to Background levels, despite the lowering pH, DO, and ORP [34]. This pattern was attributed primarily to enhanced desorption caused by formation of uranyl carbonate complexes [34]. It is also possible that genera such as Geobacter, and Geothrix, which have been linked to the anaerobic oxidation of U [76] and were observed to increase in the Early and Mid phases, may also have influenced the observed U increase in this leakage scenario. Although As concentrations were not observed to increase compared to Background conditions [34], microbes isolated from Newark Basin black shale have previously been shown to mobilize arsenic under sulfide oxidizing conditions [77]. Newark Basin black shales incubated in groundwater were quickly colonized with biofilms, and scanning electron microscopy detected bacteria shaped pits on pyrite crystals indicating bio-weathering [78]. Cultures from these surfaces were dominated by iron and pyritic sulfur oxidizers including Geobacter and Dechloromonas [78]. Both of these genera were abundant in our field experiment and peaked in Early or Mid phases (Table 2; Figs. 4 and 7) of the experiment, with Dechloromonas accounting for more than 50% of the detected sequences in some post-injection samples. This is not surprising as Dechloromonas has previously been identified in association with black shales in low pH environments [79]. The association of microbes with black shales from the Newark Basin may be of large biogeochemical significance, potentially impacting water quality through metal mobilization. Zhu et al [80] found that biologically active incubations of Newark basin black shale and its associated microbes, under sulfate |
reducing conditions, resulted in seven times more mobilization of arsenic than sterile controls. This supports the possibility that the microbial dynamics observed during the injection experiment could play a role in trace metal geochemistry resulting from a CO 2 leakage event and could have important consequences for water quality in potable aquifers. Many prior laboratory and modeling experiments studying CO 2 injection or leakage associated metal mobilization have purposefully excluded or limited microbial activity (e.g. drying of samples in [49]), while other studies have stressed the importance of research that considers the microbial response and potential biogeochemical contribution [19], [23], [34], [61]. Although the duration of our field experiments is relatively short (weeks), disturbance conditions in a leakage scenario may also be short-lived, and the timescale of most metal mobilization is expected to be fast (days/weeks) [34], [50], [58]. Therefore, the rapid successional dynamics observed in this experiment may be well synchronized to the most important geochemical/biogeochemical alterations linked to metal mobilization. Even in experiments seeking to minimize biotic interactions, redox state was observed to be important in controlling metal mobilization [50], highlighting the potential for both direct (e.g. metal reduction) and indirect (e.g. lowered redox state and potential iron-sulfur interactions) influences of microbial activity on metal mobilization. Importance of free versus attached bacteria It is worth noting that our methods did not allow for determination of the relative contribution of microbes attached to substrate vs. microbes suspended in aquifer water. In fact, our microbial methods only assessed the suspended microbial community, while the measured geochemistry is expected to integrate the activity of both the attached and suspended components of the community. In addition, our experiment did not characterize the microbial transitions that would occur from a non-CO 2 saturated control injection, and the physical disturbance associated with injection may be important to consider. Previous research has determined that the percentage of microbial cells attached in aquifers is highly variable and poorly constrained, but that the majority of cells are often attached [15]. There is evidence that the abundance of unattached microbes may increase after aquifer disturbance [81] and that higher rates of well pumping can cause shearing of biofilms and increases in suspended cell counts [82][83]. Shearing of biofilms has also been speculated to account for some apparent shifts in microbial diversity or composition during groundwater experiments [84]. For example, Purkamo et al [66] speculate that increased concentrations of methanogens during a borehole experiment may be due to biofilm sheering. Apparent shifts in microbial composition due to biofilm sheering would be consistent with work by Flynn et al [19] demonstrating that the attached microbial communities contained a higher relative percentage of iron reducing bacteria, in comparison to the suspended community, which is also consistent with studies of iron oxidizers (e.g. [85]). Prior to the injection experiment, thousands of liters were pumped from the well at a constant rate until stable background parameters were observed. In addition to stabilizing the background geochemistry, this pump-down procedure is likely to have removed, or decreased, the influence of variable biofilm sheering from our successional results. Rodan and Zachara [86] found that the rate and extent of iron reduction was correlated to oxide surface area, suggesting that changes in surface associations, surface area, or surface morphology could be important to both microbial composition and biogeochemistry. It is possible that injection disturbed biofilms, through either physical or geochemical interactions, resulting in a functional shift in microbial communities and surface detachment. Tischer et al [81] showed that Desulfosporosinus, common in our post injection sequence libraries (Table 2; Fig. 7), were abundant in minicore samples acquired in the wall of a borehole. Flynn et al [19] found a distinct difference between attached and suspended microbial communities in pristine aquifer waters, with iron-reducing bacteria more abundant in attached communities than in suspended communities. The increase in genera associated with iron reduction in our post-injection samples may represent increased representation of attached communities in the samples and could be due to the shearing of biofilms from the borehole walls with pumping, the release of attached cells into aquifer water due to changes in conductivity or the increase of suspended iron reducing communities. It is possible that the observed microbial stimulation and succession instead demonstrate a shift from attached to suspended microbial communities. We are unable to differentiate between these responses in our data, although both are relevant considerations in the event of CO 2 migration or leakage from GS reservoir. Conclusion The microbial community from a Newark Basin test well was observed to display a large and significant shift in both microbial abundance and composition following an experimental injection of CO 2 -enriched aquifer water into a shallow potable water aquifer. This study represents the first experimental effort, using an in situ injection, to examine the microbial response to CO 2 migration or leakage from a GS reservoir. The observed increase in cell counts, by nearly an order of magnitude, following the injection disturbance indicate either microbial growth and/or mobilization of microbes from surface-attachment to suspension in aquifer water, with either form of microbial stimulation relevant for influencing water quality in a leakage scenario. The microbial successional patterns in the Early and Mid phases, including apparent decreases in obligate aerobes (e.g. Sphingobium) and increase in genera often associate with both acid tolerance (e.g. Acidothiobacilus) and metal reduction (e.g Dechloromonas, Geobacter), demonstrate the potential for microbes to influence geochemistry and metal mobilization in aquifer waters following CO 2 leakage. The early stages of succession were very similar in repeated injection experiments conducted one year apart in the same Newark Basin test interval and involved stimulation of a rare biosphere from the background community. In the one year period between experimental disturbances/injections, the microbial community was observed to shift back toward an abundance and composition more similar to the pre-injection background community. This suggests that the subsurface environment in Newark Basin is resilient and that low levels of CO 2 leakage from a potential GS reservoir are unlikely to create a permanent, large-scale, alteration of the subsurface microbial community. Supporting Information S1 Fig. Hierarchical cluster analysis of environmental parameters measured during the two injection experiments. Environmental parameters used in the analysis were: pH, conductance, ORP, pCO 2 , Sulfate, Manganese, and Iron. Samples are labeled with a number corresponding to injection 1 or injection 2, followed by a letter that corresponds to the order of samples with "A" representing background phase samples and consecutively collected samples listed in alphabetical order (See Fig. 4). (TIF) S2 Fig. Hierarchical cluster analysis of microbial community composition based upon DNA sequences identified to the level of genera during the two injection experiments. Samples are labeled with a number corresponding to injection 1 or injection 2, followed by a letter that corresponds to the order of samples with "A" representing background phase samples and consecutively collected samples listed in alphabetical order (See Fig. 4). (TIF) Congenital erythropoietic porphyria with two mutations of the uroporphyrinogen III synthase gene (Cys73Arg, Thr228Met) Congenital erythropoietic porphyria (CEP) is an autosomal recessive inborn error of metabolism that results from the markedly deficient activity of uroporphyrinogen III synthase (UROS). We describe a 14-year-old girl with red urine since infancy, progressive blistering and scarring of the skin, and moderate hemolytic anemia. After years of skin damage, her face is mutilated; she has a bald patch on the scalp, hypertrichosis of the neck, areas of skin darkening, and limited joint movements of the hands. Total urine excretion and fecal total porphyrin were both markedly raised above normal levels. Sequencing of the UROS gene identified two mutations causing CEP (Cys73Arg, Thr228Met). The patient lesions are progressing. Bone marrow transplantation and/or gene therapy are proposed as the next steps in her treatment. In brief, we describe a CEP with confirmed two pathogenic mutations, severe phenotype and discuss the various treatment options available. Introduction The inherited porphyrias are disorders of heme biosynthesis resulting from the deficient activity of a specific enzyme of the heme biosynthetic pathway. [1] Depending on the site of predominant porphyrin accumulation, the porphyrias are grouped into the following two types: erythropoietic and hepatic. [2] The following three different erythropoietic porphyrias (EP) have been reported: erythropoietic protoporphyria (EPP, MIM 177000), congenital erythropoietic porphyria (CEP, MIM 263700), and hepatoerythropoietic porphyria (MIM 176100). CEP or Günther's disease is an autosomal recessive disease resulting from deficient uroporphyrinogen III synthase (UROS) activity. [1] The clinical manifestations include chronic hemolysis, anemia, erythrodontia, and disfiguring cutaneous lesions. [3] Here, we describe a 14-year-old Macedonian girl with CEP, severe cutaneous disfiguration, and a known set of mutations (Cys73Arg, Thr228Met). Go to: Case Report At the age of 6 months, the patient was referred to our hospital for investigation of episodes of red urine. She was born at term after an uneventful pregnancy and delivery. Her birth weight and length were 2 900 g and 51 cm, respectively. As a 3-week-old neonate, she received a blood transfusion for, it was believed, hemolytic anemia of the newborn. In all three of her hospital admissions, a mild hemolytic anemia was diagnosed. At the age of 6 years, she was again admitted for investigation, on this occasion with a main complaint of easy skin blistering and scarring. It was noted that the intensity of the red coloration of the urine varied from day to day. Physical examination revealed multiple blistering and scarring on areas of the skin exposed to the sun [ Figures [Figures11 and and2]2] and hypertrichosis on the back [ Figure 3]. The spleen was of normal size and physical and intellectual development was normal. Again, moderate hemolytic anemia was also found (red blood cells: 3.68 × 10 6 , hemoglobin: 10.6 g/dl, hematocrit: 33.0%, MCV 89.7, MCH 28.8 pg, MCHC 32.1 G/DL, platelets 124 × 10 3 ml, white blood cells: 3.6 × 10 3 , reticulocytes: 58 (5-15)), but did not require a blood transfusion. Further laboratory investigations were as follows: urea, creatinine, uric acid, electrolytes, and alkaline phosphatase were normal. Ultrasound examination of the kidneys and heart were normal and bone densitometry was unremarkable. Using all available measures to protect herself from the sun (hat, eyeglasses, cosmetic camouflage, and long sleeves) has not succeeded in protecting her from the mutilating effects of sun. This has caused scarring of her skin at sun-exposed sites, including the backs of the hands, her face, and ears, and has resulted in bald patches on the scalp. In addition, she has restricted hand function due to scarring of the skin and has lost some of her eyelashes, which has made her eyes prone to irritation from small particles of dust. Moderate hypertrichosis on the back of the neck was also noted and her teeth have progressively stained brownish-red. The results of the biochemical investigations are summarized in Table 1. Total urine excretion [4] and fecal total porphyrin [5] were both markedly raised above normal levels. High-pressure liquid chromatography [6] of the urine demonstrated 85% of the uroporphyrin to be of isomer I type and fecal fractionation showed mainly coproporphyrin isomer I [ Figure 6]. Urinary aminolevulinic acid and porphobilinogen levels were within normal limits. The erythrocyte protoporphyrin level was greatly increased in both zinc and free forms. Plasma fluorescence spectroscopy revealed a prominent emission peak at 617. [7] These results confirmed a diagnosis of CEP. Table 1 High-pressure liquid chromatography of the urine demonstrated 85% of the uroporphyrin to be of isomer I type and fecal fractionation showed mainly coproporphyrin isomer I Figure 6 HPLC fractionation pattern of urine (blank trace) and fecal (red trace) porphyrin methyl esters. Uro I: uroporphyrin octamethyl ester isomer I, Copro I: coproporphyrin tetramethyl ester isomer I and Copro III: coproporphyrin tetramethyl ester isomer III. ... Molecular genetic analysis of the UROS gene revealed the sequence variant c.217T>C (p.Cys73Arg) in exon 4 and c.683c>T (p.Thr228Met) in exon 10. These missense mutations have previously been described, [8] where the patient had a moderate to severe phenotype. Go to: Discussion CEP has an estimated frequency of 1 in every 2 to 3 million people. As of 1997, about 130 cases had been reported. [1] In Switzerland, only four cases of CEP from a total of 217 porphyrias of different types have been described. [9] CEP affects males and females equally, and has no known predilection for any ethnic group. Mutation |
analysis has shown a large variety of molecular lesions. [10] Genotype/phenotype correlations have been demonstrated [10] showing that the clinical severity of the anemia and cutaneous lesions is highly variable, ranging from mild to severe. Some patients die in early adult life and others in the neonatal period. The mutations found in our patient are relatively frequent in CEP. Cys73Arg occurs in approximately one-third of cases and Thr228Met in around 6% (Desnick and Astrin, 2002). When expressed in an E. coli system, these mutations have <1% of normal UROS activity. [1] In case reports, a similar phenotype to that found in our patient has been reported, i.e., with severe photosensitivity and mild hemolytic anemia, but not transfusion-dependent. [8] The profound UROS deficiency [1] leads to a lifelong overproduction of isomer I porphyrins which are deposited in many tissues causing light-sensitization and severe damage to skin. Blistering and scarring of exposed areas may lead to mutilating deformity. Hypertrichosis is sometimes severe. Splenectomy, hypertransfusion, and orally administered drugs such as charcoal and cholestyramine, which binds porphyrins have also been used to treat CEP. Unfortunately, these classical treatments are unsatisfactory and do not effectively control the disease. In murine models, oncoretroviral and lentiviral vectors were used to successfully transduce HSCs, allowing full metabolic and phenotypic correction of both EPP and CEP mice. [11,12] These results form the basis for gene therapy clinical trials in severe forms of EP. In conclusion, we have described a 14-year-old girl with CEP, with defined UROS mutations and severe phenotype. BMT and gene treatment have been discussed with the parents as further treatment options. Go to: Footnotes Source of Support: Nil Short Term Outcome of Acute ST Elevation Myocardial Infarction in a Tertiary Care Cardiac Center Introduction Acute ST Elevation Myocardial infarction (STEMI) is a cardiovascular emergency and is associated with significant adverse short and long-term outcome. The objective of this study was to determine the shortterm outcome in terms of heart failure, myocardial reinfarction, stroke, hospital readmission and mortality in patients admitted with STEMI Methods It was a prospective observational study conducted at Manmohan Cardiothoracic Vascular and Transplant Center, Kathmandu from May 2014 to April 2015. All patients admitted with diagnosis of STEMI during study period were enrolled. Patients were treated on the basis of existing guidelines. Mode of management, adverse outcomes and mortality of patients during the study period were evaluated. The discharged patients were followed up for 30 days. Statistical analysis was performed with SPSS version 20. Ethical approval was taken from the Institutional Review Committee of Institute of Medicine. Results The median duration of presentation was 20 hours, and only 40% of the patients presented within 12 hours of symptom onset. Primary PCI was performed in 50 (33%), thrombolysis was performed in 29(19%) and conservative medical management was done in 72 (48%) patients. Overall outcome occurred in 52 (37.7%) patients. In hospital and 30 day mortality was 14 (9.2%) and 17 (11%) respectively. Heart failure was present in 28(18.5%), myocardial reinfarction 8 (5%), stroke 4 (2.6%), and hospital readmission was 18 (12%). Conservatively treated patients had significantly more adverse outcomes (p=0.02). More patients in conservatively managed group had hospital readmission. (p=0.04) Conclusion There were more overall adverse outcomes in conservatitley managed group which is mainly due to more hospital readmission. S T elevation myocardial infarction (STEMI) is a clinical syndrome defined by characteristic symptoms of myocardial ischemia accompanied by persistent elevation the ST segment on electrocardiograph (ECG) and subsequent release of biomarkers of myocardial necrosis. 1 Acute MI occurs when intra arterial thrombus propagates and completely occlude blood flow within the artery, resulting in ischemia and necrosis of cardiomyocytes distal to the obstruction. 2 Acute MI causes substantial morbidity and mortality worldwide. 3 Management of STEMI is directed towards reperfusion therapy, either by thrombolysis or primary percutaneous coronary artery intervention (PPCI). PPCI is the best available option and various studies have shown that PPCI is better than thrombolysis in reducing cardiovascular mortality and morbidity. 4,5 If the patient presents after 12-24 hours of symptoms onset without ongoing pain or complications, they are usually kept in conservative medical management and percutaneous coronary intervention (PCI) can be planned later. Adverse outcomes are usually based on the mortality, myocardial reinfarction, heart failure, stroke and recurrent hospital admission. Among patients with STEMI enrolled in trials, approximate 30day mortality rates were 13 percent with medical therapy alone, 6 to 7 percent with optimal fibrinolytic therapy, 6 and as low as 3 to 5 percent with primary percutaneous coronary intervention (PCI) when timely performed. 7 In our country, there is limited data regarding the outcomes of acute STEMI. Few studies have been published which compare the early outcome of PPCI in acute STEMI. 8 Outcomes related to thrombolysis and conservative treatment is not well studied. So, the general objective of the study was to know the short-term adverse outcome in hospital and after hospital discharge up to 30 days in acute STEMI in all three PPCI, thrombolysis and conservatively managed groups. The specific objective was to determine the heart failure, myocardial reinfarction, stroke, recurrent hospital admission and mortality in these different groups. METHODS It was a prospective observational study conducted in Department of Cardiology, Manmohan Cardiothoracic Vascular and Transplant Centre (MCVTC), Maharajgunj, Kathmandu. Acute STEMI was diagnosed on the basis of third universal definition of myocardial infarction. 9 Study period was one year from May 2014 to April 2015. Adult patients above 18 years, both male and female, diagnosed and admitted as acute STEMI were enrolled in the study. Excluded patients were those who died in ER before initiation of proper treatment, and those with medical surgical illness that precluded treatment (acute CVA, severe sepsis, ongoing UGI bleed). Informed consent was taken from all patients. The Institutional Ethical Review Committee of Institute of Medicine approved the study. All eligible acute STEMI patients were enrolled in the study. Presenting symptoms, general physical examination, and baseline investigations along with cardiac enzymes, ECG and echocardiography was performed at admission. Risk factors were determined on the basis of history and investigations. ↓ ↓ Patients were managed according to the guidelines given by ACC/AHA in 2013. 10 Depending upon the time duration at which patients presented, there were three groups of patients-Primary PCI, Thrombolysis and Conservatively managed group. (Fig 1) Non fatal outcome included heart failure, stroke, myocardial reinfarction and hospital readmission. Myocardial reinfarction was considered when ST elevation ≥0.1 mV recurs, or new pathognomonic Q waves appear, in at least two contiguous leads, when associated with ischemic symptoms for 20 min or longer. 9 Heart failure was defined as from crepts in chest, with S3 and raised JVP, to frank pulmonary edema. 11 Stroke was defined as a clinical syndrome consisting of rapidly developing clinical signs of focal (or global in case of coma) disturbance of cerebral function lasting more than 24 hours or leading to death with no apparent cause other than vascular origin. 12 Fatal outcome included in hospital and 30 day mortality. Overall adverse outcome was said to be present, if anyone of the fatal or nonfatal outcome occurred. Depending upon the clinical status and requirement, some patients were followed up after 1-2 weeks, and all patients were called upon at 30 days after discharge. Those patients who didn't come for follow up, telephone call was made to know regarding their status. If mortality had occurred, it was noted with likely cause of death. Statistical analysis was performed with SPSS version 20. For demographic profile, frequency and percentage distribution were obtained for each variable. Data were expressed as mean+ SD for continuous variables and as percentage for categorical variables. For continuous variables, differences between groups were compared with independent t-test. Chi-square test was used to detect linear association between type of treatment and adverse outcome. A two sided p value of <0.05 was considered statistically significant. RESULTS There were total of 166 patients presented to MCVTC Emergency room (ER) with diagnosis of acute STEMI, and 151 patients were included in the study who fit in inclusion criteria. (Fig 1) The mean age was 59.0 ±13 years. ( Heart failure was present in 28(18.5%) patients, myocardial reinfarction in 8(5%) patients, stroke in 4(2.6%) patients, hospital readmission in 18(12%) patients and death in 17(11%) patients. (Table 2) There were 7(14%) patients in primary PCI group, 5(18%) in thrombolysis group and 16(22%) in conservatively managed group that had heart failure. One (2%) patient in primary PCI group, 1(3.5%) in thrombolysis group and 6(8.2%) in conservatively managed group had myocardial reinfarction. No stroke occurred in primary PCI group. One (3.6%) in thrombolysis group and 3(4.1%) in conservatively managed group had stroke. Two (4%) patients in primary PCI group, 3(11%) in thrombolysis group and 13(18%) in conservatively managed group had hospital readmission for various causes during 30 days of MI. Overall adverse outcome occurred in 52(34.4%) patients. Eleven (22%) patients in primary PCI group, 8(28.5%) in thrombolysis group and 33(45%) in conservative management group had major adverse outcome. As compared to reperfusion therapy, there were significantly more overall adverse outcome in conservatively managed group(p=0.02). This was largely driven by increased hospital readmission in conservatively managed group, whereas there was no significant difference of other outcomes in both reperfused and conservatively managed group. There were total of 17(11%) thirty-day deaths. Seven (14%) in primary PCI group, 1(3.5%) in thrombolysis group and 9(12%) in conservatively managed groups had death. Three patients expired after hospital discharge, while in hospital mortality was 14(9.2%). There were more deaths in woman 10(20.8%) as compared to man 7(6.8%). About 53% of patients who presented in shock expired, Short Term Outcome of Acute STEMI JIOM Nepal 3. Causes of in-hospital death which was the most common cause of death ( Figure 3). While excluding patients in shock, the mortality was 6.6%. DISCUSSION The mean age of the patients was similar to other international studies. [13][14][15] The gender difference of more male patients is seen not only in Nepal, but internationally. 7,13,14 The incidence of CAD in female is less than male probably due to the protective role of estrogen in premenopausal woman, but its incidence increases as the age progresses and it is the leading cause of death in women. 16 One of the striking features about risk factors was high prevalence of smoking. About 54.3% of patients were current smoker. High prevalence of smoking was present in other studies in Nepal also. In western Nepal, it was about 80% 15 while in Kathmandu it was about 60%. 8 In NCDR registry in the USA, smoking was present in 31%. 7 As smoking is the major risk factor for CAD, and its high prevalence in our country, every effort should be done to help people quit smoking not only to reduce the risk of CAD but also to reduce the risk of pulmonary diseases and malignancy. The overall median duration of presentation or prehospital time (PHT) delay in this study was 20(1 to 120) hours which is high. The percentage of patients presenting within 12 hours of symptoms onset is 40% and those presenting within 24 hours is 58%. The median PHT is 3. in Japan. 17 As earlier the reperfusion is performed better is the short and long term result,4 effort should be directed to decrease the prehospital delay, which is markedly longer as compared to other countries. Studies have shown that patient may not perceive the acute symptoms, may not recognize the severity and importance of symptom, and there might be delay for call for help; so health care providers and policymakers should address in these areas to decrease the prehospital delay. 18 Sixty percent of patients received PPCI and 38% of the patients received thrombolysis who presented within 12 hours of symptoms onset. Fifty four percent of patients received PPCI and 32% patients received thrombolysis who presented within 24 hours. This is similar to that of other countries. In National Cardiovascular Data Registry (NCDR) of the USA, primary PCI was performed in 83% and thrombolysis was performed in 12% and overall reperfusion therapy was performed in 93% of STEMI patients. 7 Similarly, in European registry, PCI was performed in 61% and thrombolysis was performed in 33 %, and overall reperfusion therapy was performed in 94% of patients. 19 This shows we are providing reperfusion therapy to those who presented |
within window period similar to that of western world. The median door to balloon time (DBT) was 100 (40 to 180) minutes, and the mean DBT was 97.6(±32) minutes, which is comparable to that of other studies. 7 Overall heart failure was present in 28(18.5%) patients. Occurrence of heart failure varies across the world. In Kerala registry in India, heart failure was present in 2.7%, 14 in GRACE study 18%, 20 and in European registry 16.8%. 19 Slight increase in heart failure might be due inclusion of patients who were managed conservatively. Myocardial reinfarction occurred in 8(5%) patients. One (2%) patient in primary PCI group, 1(3.5%) in thrombolysis group and 6(8.2%) in conservatively managed group had myocardial reinfarction. In the RIKS-HIA study done in Swedish population, reinfarction occurred in 2% in PCI group and 3.4% in thrombolysis group. 21 So the occurrence of reinfarction is slightly high when all groups are taken together but is similar to western world when taken separately for PCI and thrombolysis. Stroke occurred in 4(2.6%) overall patients but no stroke occurred in primary PCI group similar to other studies. 14,19 Hospital readmissions are a quality of care indicator. After hospital discharge, hospital readmission is one of the factor leading to increased morbidity and mortality. There was rehospitalization in 18(12%) patients but only two (4%) patients in primary PCI group, which is similar to that of other studies. 22 Out of 17 deaths in the study period, 14(9.2%) were inhospital deaths and 3 were on follow up. Out of 15 patients who have presented in Killip Class IV, 8 (53%) patients expired. Similar high mortality was seen in patients with STEMI presenting in cardiogenic shock. 23,24 While excluding patients in shock, the mortality was 6.6%. In Indian CREATE study, 30 day mortality was 9%, 13 and in GRACE study 8% 25 and in Hospital in Kerala study 8.2%. 14 This shows that overall mortality is slightly higher than other studies. The reason behind this may be due to the fact that many patients in our study were kept in conservative management while majority of the patients in other studies underwent reperfusion therapy. Patient presentation was also quite late, and studies have shown that earlier the perfusion, better is the outcome. 26,27 The high mortality in PCI group can be explained by studies showing high mortality in patients presenting in cardiogenic shock. 28 There was more mortality in female. Several other studies have also shown female sex as a predictor of adverse cardiovascular outcome. 29 More adverse outcome and mortality in female is found to be later age of presentation than men, lesser use of primary PCI and presence of more comorbid conditions. 30 Overall adverse outcomes were significantly more in conservatively managed group (p=0.02), which was mainly because of significant more hospital readmissions in conservatively managed group (p=0.04). Whereas, there was no significant difference of outcomes in other variables between two groups. This can be explained by small sample size, low number of event rates as well as short duration of study period. The study was limited to one center. Sample size was small. The individual event rates were also less because of small sample size. The patients who expired at ER and who refused admissions were not included in the study. CONCLUSION Overall adverse cardiovascular outcome as well as recurrent hospital admission is higher in conservatively managed group. Temperature monitoring: the consequences and prevention of mild perioperative hypothermia Homeothermic species require a nearly constant internal body temperature. Significant deviations from “normal” internal temperature cause the metabolic function to deteriorate. Usually, the human thermoregulatory system maintains a core body temperature within 0.2°C of normal, near 37°C. Hypothermia results from exposure to cold, or exposure combined with drugs or illness that decrease thermoregulatory efficacy. Exposure to a cold operating room environment during anaesthesia and surgery commonly combines with anaesthetic-induced inhibition of thermoregulation to produce hypothermia. The prevention and management of temperature-related complications is expedited by an understanding of both normal and druginfluenced thermoregulation. Keywords : mild perioperative hypothermia; consequences; prevention Introduction Homeothermic species require a nearly constant internal body temperature. Significant deviations from "normal" internal temperature cause the metabolic function to deteriorate. Usually, the human thermoregulatory system maintains a core body temperature within 0.2°C of normal, near 37°C. Hypothermia results from exposure to cold, or exposure combined with drugs or illness that decrease thermoregulatory efficacy. Exposure to a cold operating room environment during anaesthesia and surgery commonly combines with anaesthetic-induced inhibition of thermoregulation to produce hypothermia. The prevention and management of temperature-related complications is expedited by an understanding of both normal and druginfluenced thermoregulation. Normal thermoregulation Thermoregulation, like many physiological control systems, relies on multiple levels of positive and negative feedback to minimise perturbations from the normal status. Temperature is regulated by signals derived from nearly every type of tissue, including the hypothalamus, spinal cord, deep core tissue and skin surface. Cold signals travel to the hypothalamus and other central structures primarily via A∂ nerve fibres, while warm information travels by unmyelinated C fibres. Most ascending thermal information traverses the spinothalamic tract in the anterior spinal cord. The processing of thermoregulatory information occurs in three phases, namely afferent thermal sensing, central regulation and efferent response. The hypothalamus regulates temperature by comparing integrated thermal input from the skin surface, neuraxis and deep tissue with threshold temperatures for heat and cold. When integrated input exceeds a threshold, the appropriate response is initiated to maintain adequate body temperature. The slope of response intensity versus core temperature regression is the gain of that response. The difference between the lowest warm and highest cold thresholds indicates the thermal sensitivity of the system. Typically, the interthreshold range, i.e. the temperature range over which no regulatory responses occur, is only a few tenths of a degree Celsius. Because each thermoregulatory response has its own threshold and gain, there is an orderly progression of responses and response intensities are in proportion to need. How the body determines absolute threshold temperatures is unknown, but the thresholds vary daily in both sexes by ≈ 1°C (circadian rhythm), and monthly in women by ≈ 0.5°C. Exercise, food intake, infection, hypo-and hyperthyroidism, anaesthetic and other drugs (including alcohol, sedatives and and nicotine), and cold and warm adaptation alter threshold temperatures. Central regulation is intact from infancy, but may be impaired in elderly or extremely ill patients. The hypothalamus responds to temperatures exceeding the appropriate thresholds via response mechanisms that increase metabolic heat production or alter environmental heat loss. These responses allow normal individuals to maintain a core temperature near 37°C, despite widely varying environmental temperatures. As thermoregulatory responses are inhibited (by drugs), the range of environmental temperatures over which normal core temperature can be maintained decreases. For example, when shivering is prevented following the administration of muscle relaxants, hypothermia will develop in an environment that was well tolerated previously. When all thermoregulatory responses Temperature monitoring: the consequences and prevention of mild perioperative hypothermia are prevented, the core temperature will remain normal only in a thermoneutral environment. Quantitatively, behavioural regulation (e.g. dressing appropriately, modifying environmental temperature and voluntary movement) is the most important effector mechanism. Cutaneous vasoconstriction decreases heat loss via convection and radiation from the skin surface. Total digital skin blood flow is divided into capillary and thermoregulatory arteriovenous shunt components. The arteriovenous shunts are largely limited anatomically to the fingers, toes and nose. They are functionally distinct from capillaries supplying the remainder of the skin. Shunts are open and capillary flow is nearly minimal in a thermoneutral environment. Shunt vasoconstriction is mediated primarily by the release of norepinephrine from presynaptic adrenergic nerve terminals. Vasoconstriction only minimally decreases capillary blood flow. Nonshivering thermogenesis increases metabolic heat production without producing mechanical work via brown fat oxidation. Nonshivering thermogenesis increases heat production ≈ 100% in infants, but only slightly in adults. Vigorous shivering roughly doubles metabolic heat production, although this level of intensity cannot be sustained for long. The net efficiency of shivering thermogenesis is somewhat lower than might be expected because muscle metabolism increases blood flow to peripheral tissue, and consequently, heat loss to the environment. Thus, it is used only as a last defense against hypothermia, i.e. its activation threshold is a full degree Celsius below that with vasoconstriction and nonshivering thermogenesis. Sweating is mediated by cholinergic, post-ganglionic sympathetic nerves. Trained athletes can sweat up to 2 l/hour and lose as much as 10 times their resting metabolic rates. Sweating is the only mechanism by which humans can lose heat in an environment that exceeds core body temperature. Active vasodilation is mediated by the release of a yet-to-be fully-characterised mediator (possibly nitric oxide) from the sweat glands. Heat stress can increase capillary blood flow up to 7.5 l/minute (Table I). Intraoperative temperature monitoring Input to the central thermoregulatory system is derived largely from deep abdominal and thoracic tissues, the spinal cord and the brain. Thus, no single tissue temperature can be considered to be a "gold standard." Core temperature can be determined by measuring a single temperature that is adjacent to the tympanic membrane, or in the nasopharynx, pulmonary artery or distal oesophagus. Carefully obtained oral, axillary and bladder temperatures approximate core temperatures sufficiently for clinical use, except during cardiopulmonary bypass. Skin temperatures are approximately 2°C less than core temperature, but the difference between forehead skin and core temperature varies among individuals and within individuals over time. Forehead skin temperature, with a 2°-C compensation is thus only a rough estimate of core temperature. Infrared "tympanic" thermometers and "temporal artery" thermometers are insufficiently accurate for clinical use. Rectal temperature often seriously lags true core temperature, and should thus be avoided. Usually, body temperature measurements are not necessary during monitored anaesthesia care or minor procedures that are performed under regional anaesthesia. Major surgery, accomplished by epidural or spinal anaesthesia, can be associated with considerable hypothermia. Core temperature monitoring is generally appropriate in such cases. Core temperature alterations immediately following induction of general anaesthesia are hard to predict, and are influenced by a variety of factors. Consequently, temperature monitoring is often not helpful during the first 30 minutes of anaesthesia, and is not required when surgery is completed within this period. By contrast, core temperature should be monitored at no less than 15-minute intervals during general anaesthetics that last longer than 60 minutes. An oesophageal probe is usually the easiest and most reliable method of reliably measuring core temperature in intubated patients. Axillary, oral or forehead skin surface temperatures can be substituted for oesophageal or nasopharyngeal temperatures during regional anaesthesia, or when patients are mask ventilated. Of these, oral temperature is the most reliable and is also the most suitable for postoperative use (Table II). Thermoregulation during anaesthesia Typical doses of general anaesthesia decrease the activation thresholds for responses to hypothermia by ≈ 3°C and increase those defending against hyperthermia by < 1°C. Widening the interthreshold range produces a broad temperature range over which active thermoregulatory responses are absent. Patients are poikilothermic within this range and body temperature changes determined passively by redistribution of heat within the body and the difference between metabolic heat production and heat loss to the environment. Most general anaesthetics produce a similar pattern and magnitude of thermoregulatory impairment. This pattern is illustrated in Figure 1 using propofol. Opioids inhibit thermoregulatory control similarly. By contrast, even very high plasma concentrations of midazolam have little thermoregulatory effect. The vasoconstriction threshold in infants who are anaesthetised with halothane or isoflurane differs little to that in adults. Intraoperative vasoconstriction thresholds are increased by painful stimulation, but decreased by advanced age. Perioperative hypothermia The typical pattern of intraoperative hypothermia is illustrated in Figure 2. Core temperature decreases precipitously for one hour (redistribution hypothermia), then decreases slowly for 23 hours (a linear decrease), and finally becomes constant (plateau phase). Hypothermia during the first hour of anaesthesia results largely from the core-to-peripheral redistribution of body heat. Subsequently, core and mean body temperatures decrease when heat loss exceeds heat production. And finally, a core temperature plateau results when the reemergence of thermoregulatory vasoconstriction decreases cutaneous heat loss and constrains metabolic heat to the core thermal compartment Numerous factors contribute to core hypothermia during the first hour of surgery: • Patients are undressed in a cool environment. • Their skin is prepared for surgery with a cold solution which is allowed to evaporate. • |
The induction of anaesthesia decreases metabolic heat production and produces cutaneous vasodilation. • Cold intravenous fluids directly decrease the core temperature. • The patients' lungs are ventilated with dry gases which increase respiratory heat loss. Nonetheless, unanaesthetised individuals easily tolerate typical operating room temperatures without becoming hypothermic, and the evaporation of skin preparation fluids cannot explain the observed hypothermia. Furthermore, general anaesthesia decreases metabolic heat production by only 15-30%, and minimally increases cutaneous heat loss. Since decreased heat production and increased heat loss are insufficient to explain the initial core hypothermia following the induction of general anaesthesia, another mechanism must be invoked. Core temperature poorly represents mean body temperature and body heat content because the temperature in peripheral tissue (roughly half the body mass) changes considerably depending on thermoregulatory status, environmental temperature and duration in a particular environment. Typically, peripheral tissue temperature (and heat content) is considerably lower than core temperature. Although anaesthetic-induced vasodilation only minimally increases cutaneous heat loss to the environment, it allows the mixing of heat in the core and peripheral thermal compartments. The result is peripheral warming at the expense of core temperature ( Figure 3). Redistribution hypothermia contributes ≈ 80% to observed hypothermia during the first hour of anaesthesia, and often remains the major cause of core hypothermia, even after three hours of anaesthesia. The similar internal redistribution of heat causes the initial hypothermia during epidural anaesthesia. The initial hypothermia which develops following the induction of anaesthesia results when vasodilation allows heat in warm core tissue to mix with cooler peripheral tissue. This warms the periphery at the expense of the core temperature. Although the core temperature decreases precipitously, body heat content (and mean body temperature) remains nearly constant It is likely that the slow, linear decrease in core temperature that is typically observed during the second and third hours of anaesthesia simply results from heat loss that exceeds metabolic heat production. Heat loss presumably also exceeds production during the first hour of anaesthesia, but usually contributes much less to core hypothermia than redistribution. When patients are in a relatively warm environment and undergo small operations, the core temperature plateau that develops after the second to fourth hour of anaesthesia may be passive, i.e. without thermoregulation. By contrast, when patients are sufficiently hypothermic, this plateau is often accompanied by active thermoregulatory vasoconstriction. Peripheral vasoconstriction decreases cutaneous heat loss and constrains metabolic heat (which is mostly generated centrally) to the relatively small core thermal compartment, thereby preventing further hypothermia, and sometimes even increasing core temperature. Heat loss from peripheral tissue not only continues, but this tissue receives less metabolic heat. The core temperature plateau does not represent a true thermal steady state (heat production equalling heat loss) in these circumstances because body heat content continues to decrease (Table III). Just 2°-C core hypothermia significantly increases the incidence of myocardial ischaemia in high-risk patients undergoing peripheral vascular surgery. Myocardial ischaemia probably results less from post-anaesthetic shivering than from sympathetic activation, hypertension and tachycardia. Mild hypothermia reduces platelet function and decreases activation of the coagulation cascade. Consistent with these in vitro data, hypothermia significantly increases blood loss and allogeneic transfusion requirements during elective primary hip arthroplasty. Blood loss increases roughly 0.5 units for each degree Celsius decrease in core temperature. Mild hypothermia contributes to a common and serious complication of anaesthesia and surgery and wound infection by directly impairing immune function, especially oxidative killing by neutrophils, and decreasing cutaneous blood flow which reduces tissue oxygen delivery. Together, these effects of mild hypothermia triple the clinical incidence of surgical wound infections, and substantially increase the duration of hospitalisation in patients undergoing elective colon resection. Hypothermia-induced protein wasting and decreased collagen synthesis is also likely to impair wound healing. Mild hypothermia decreases the metabolism of most drugs. For example, the duration of action of vecuronium is more than doubled at 34.5°C, compared with 36.5°C. The effect of hypothermia on metabolism of atracurium and propofol is somewhat less, but is still clinically important. Consistent with decreased drug metabolism and exaggerated drug effect, mild hypothermia significantly prolongs the duration of postoperative recovery, even when temperature is not a discharge criterion. Postoperative thermal discomfort per se is not lifethreatening. Nonetheless, patients who become hypothermic often remember the worst aspect of surgery as feeling cold postoperatively. Efforts to prevent and treat such unpleasant sensations deserve the same attention as those that are currently given to pain management. Shivering-like tremor occurs in ≈ 40% of unwarmed patients recovering from general anaesthesia, but is now rare. Shivering is usually preceded by core hypothermia and peripheral vasoconstriction, indicating that it is thermoregulatory. Perianaesthetic tremor increases oxygen consumption ≈ 200% and may exacerbate postoperative pain. Although most of the tremor has electromyographic patterns that characterise normal shivering, some resemble pathological clonus. The tremor can be treated by skin-surface warming, clonidine (75 µg intravenously), or meperidine administration (25 mg intravenously). Meperidine is more effective than equipotent doses of other opioids, possibly because the drug is a central alpha-receptor agonist, in addition to its primary action at µ-opioid receptors (Table IV). Tremor during epidural anaesthesia is normal thermoregulatory shivering triggered by core hypothermia, and preceded by peripheral vasoconstriction above the level of sympathetic blockade. It can be treated by skin surface warming above the level of blockade, intravenous or epidural meperidine (25 mg) administration, or epidural sufentanil administration (50 µg). The prevention and treatment of hypothermia Initial rapid core hypothermia during epidural or general anaesthesia is difficult to treat because it results largely from the internal redistribution of heat. However, it can be largely prevented by warming the skin and peripheral tissue before induction to decrease the core peripheral temperature gradient. Given the large heat capacity of the peripheral thermal compartment, generally 30-60 minutes of moderate warming is required. Less than 10% of metabolic heat is lost via respiration, even when patients are ventilated with dry, cool gas. Passive airway humidification (heat and moisture exchangers) can prevent most of this loss, and active gas heating and humidification prevents it all. However, simple thermodynamic calculations indicate that airway heating and humidification in adults cannot produce clinically significant alterations in body heat content. Technically adequate clinical studies find that inspired gas conditioning fails to prevent hypothermia in adults. (It is likely that the clinical impression that airway heating is effective results from the artifactual heating of proximal oesophageal thermometers.) Consequently, active airway heating and humidification is rarely indicated. Similarly, heat and moisture-exchanging filters contribute little to the maintenance of normothermia. Recently developed systems which appear to be effective include over-or under-body resistive heating and circulating water applied to extremities. Forced air is by far the best validated warming approach and offers a good combination of safety, efficacy, ease of use and price ( Figure 4). Surgical Care Improvement Project and Physicians Quality Reporting Initiative The Surgical Care Improvement Project (SCIP) and the Physicians Quality Reporting Initiative (PQRI) are separate measures. However, they were designed in concert and are harmonised. Thus, they each have identical requirements. SCIP is a hospital-based reporting requirement. By contrast, PQRI applies to individuals, and is linked to a 2% Medicare bonus for reporting. Presumably, the bonus will soon be provided for meeting requirements, and not just reporting effort. The American Society of Anesthesiologists (ASA) strongly supported the inclusion of thermal management in PQRI because it is the only measure that is specific to operating room anaesthesia. Without it, anaesthesiologists would have been excluded from the 2% Medicare bonus. (I was an ASA delegate to the PQRI committee.) The denominator for SCIP and PQRI includes patients who are having surgical procedures with general or neuraxial anaesthesia that lasts at least 60 minutes in which the need for intentional hypothermia is not documented. Use of cardiopulmonary bypass is considered to be de facto evidence of intentional hypothermia. Thus, patients having monitored anaesthesia care, peripheral nerve blocks alone and short procedures are excluded. The numerator includes both a process and intermediate outcome component. The outcome component is met when patients have a documented body temperature ≥ 36°C within 30 minutes before the end of anaesthesia, and/or within 15 minutes thereafter. There is no requirement to use a specific type of thermometer or to measure temperature at a particular site. However, it is likely that accurate thermometers and a reliable core temperature site should provide higher values. The process component is met by the use of active overbody warming, such as forced air (Table V). Cardiopulmonary bypass considered to be intentional Ultrasound-guided needle release plus corticosteroid injection of superficial radial nerve: A case report BACKGROUND The radial nerve (RN) splits into two main branches at the elbow: The superficial branch of RN (SBRN) and the deep branch of RN. The SBRN can be easily damaged in acute trauma due to its superficial feature. CASE SUMMARY A 55-year-old male patient injured his right wrist 10 mo ago. Debridement, suturing and bandaging were performed in the emergency room. Six months after the scar had healed, he felt numbness and tingling in the dorsal surface of the thumb of the right hand. So the surgery of resection and SBRN anastomosis were performed. The pathological findings showed it as traumatic neuroma. Four months after surgery, the patient felt numbness and tingling in the right dorsal surface of the thumb again. The tenderness was marked in the operated area. Ultrasound indicated that the SBRN was adhered to the surrounding tissue. The patient refused further surgical treatment and underwent ultrasound-guided needle release plus corticosteroid injection of the SBRN. Four weeks later, the tenderness in the surgical area was reduced by 70%, the numbness in the dorsal surface of the thumb of the right hand was reduced by 40% and the nerve swelling evaluated by ultrasound was reduced. Four months passed, he did not feel any numbness or tingling sensation of his right wrist. This is the first report of ultrasound-guided needle release plus corticosteroid injection of the SBRN. CONCLUSION Ultrasound can evaluate the condition of the RN, and the relationship with surrounding tissues. Ultrasound-guided needle release plus corticosteroid injection is an effective and safe treatment for SBRN adhesion. Open-Access: This article is an open-access article that was selected by an in-house editor and fully peer-reviewed by external reviewers. It is distributed in accordance with the Creative Commons Attribution NonCommercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited and the use is non-commercial. See: htt ps://creativecommons.org/Licens es/by-nc/4.0/ INTRODUCTION The radial nerve (RN) is the largest branch of the brachial plexus posterior cord. It splits into two main branches at the elbow: The superficial branch of RN (SBRN) and the deep branch of RN [1,2]. The RN supplies the skin of the dorsal forearm, the dorsalradial region of the hand and the muscle of the extensor compartment [1,2]. Ultrasound (US) is currently used as an imaging modality to observe nerves, especially superficial nerves [3]. The RN has a twisted course in the upper limb and is superficial enough to be accurately found using an ultrasonic high frequency probe [1,2]. Herein, we report a patient who had numbness and tingling in the dorsal surface of the thumb after traumatic neuroma resection plus SBRN anastomosis. Treatment consisted of US-guided needle release plus corticosteroid injection of the SBRN. Tenderness and numbness were reduced 4 wk after treatment. Chief complaints A 55-year-old male patient presented to the department of ultrasound with numbness and tingling in the dorsal surface of the right thumb. History of present illness The patient injured his right wrist 10 mo previously. Debridement, suturing and bandaging were performed in the emergency room. Six months after the scar had healed, he felt numbness and tingling in the dorsal surface of the right thumb and was diagnosed with traumatic neuroma. Traumatic neuroma resection and SBRN anastomosis were performed. Four months after this treatment, he felt numbness and tingling in the right dorsal surface of the thumb, with obvious tenderness in the operated area. History of past illness No data were available. Personal and family history No data were available. Physical examination Tinel's sign was found to be positive on percussion of the right wrist. Laboratory examinations No data were available. February 6, 2022 Volume 10 Issue 4 Imaging |
examinations US revealed that the SBRN was adhered to the surrounding tissue. FINAL DIAGNOSIS US indicated that the SBRN had adhered to the surrounding tissue after traumatic neuroma resection and SBRN anastomosis ( Figure 1). TREATMENT All treatment procedures were performed with the probe covered with surgical gloves. We used an acoustic coupling agent on the probe inside the surgical gloves ( Figure 2). The patient's skin was disinfected 3 times with complex iodine. A sterile surgical towel was then placed. A 4 mL aliquot of a mixed solution which contained 2 mL 0.9% sodium chloride and 2 mL 2% lidocaine (in a ratio of 1:1) was injected. Then local anesthesia layer by layer was performed via the SBRN surface. The adhesion between the SBRN and the subcutaneous soft tissue was separated under US guidance. Due to the needle angle, we performed needle release to the middle incision (transverse incision level) as it was difficult to go further. Needle release was then performed from the incision distal area to the proximal area, above the level of the transverse incision. Separation of posterior and bilateral adhesions of the SBRN was carried out on both the short and long axis (transverse incision level). During needle release, the process was considered satisfactory when there was no resistance between the syringe and the tissues around the RN. Finally, a mixture of 1 mL corticosteroid (betamethasone) and 2 mL 2% lidocaine was injected into the area of severe adhesion on the short axis ( Figure 3). OUTCOME AND FOLLOW-UP Four weeks later after ultrasound-guided needle release plus corticosteroid injection of superficial radial nerve, the tenderness in the surgical area was reduced by 70%, the numbness in the dorsal surface of the right thumb was reduced by 40% and the nerve swelling evaluated by US was reduced. Four months passed, he did not feel any numbness or tingling sensation of his right wrist. DISCUSSION Because of cost-effectiveness and non-invasive characteristics, US has gained popularity in diagnosing peripheral nerve diseases [4]. US can not only detect the nerve but can also reveal the location and the relationship to other structures [5]. In most studies, nerves in US images are always described as hypoechoic fascicles with surrounding hyperechoic tissue, appearing as a typical honeycomb structure[6]. The nerve is usually observed on US axial view from the proximal to distal area and is tracked along an extended length. Color Doppler is used to observe vessels which can serve as anatomical landmarks near the nerve. In US images, a muscle innervated by a nerve which is smaller and hyperechoic compared to the contralateral side indicates atrophy and thus may indicate an abnormal nerve. The RNs are normal when they had a stippled honeycomb appearance with hypoechoic areas corresponding to the nerve fascicles and surrounding hyperechoic rims corresponding to endoneurium, perineurium, and epineurium on the short axis of US images [7]. The RN can be damaged during acute trauma by direct laceration or contusion, by traction in high-impact trauma with bone separation or by osseous fragments [2]. Humeral shaft fractures are the most common injuries to the RN [8,9]. The RN can also be compressed or stretched in patients who have undergone surgery. The SBRN can easily be affected by penetrating trauma especially the point which pierces the fascia due to its superficial characteristics. The symptoms include pain, dysesthesia and drop wrist. Conservative treatments include rest, nonsteroidal antiinflammatory drugs (NSAIDs) and physical therapy. If symptoms persist 12 wk after conservative treatment, then surgery is strongly recommended [10]. In this case, we found that after traumatic neuroma resection and nerve anastomosis, the patient felt numbness in the dorsal surface of the right thumb and obvious tenderness in the surgical area. US revealed that the SBRN was adhered to the surrounding tissues. NSAIDs administered for 3 mo were ineffective. The patient refused further surgical treatment. Thus, we used US-guided needle release plus corticosteroid injection to relieve the adhesion between the SBRN and the subcutaneous soft tissue for the first time. Under US guidance, we observed the needle in real-time thus improving accuracy and safety. A short-axis view of the nerve and an in-plane view of the needle were performed in order to view the outline of the SBRN and the approach of the needle. We also rotated between short-axis and long-axis views of the SBRN. The process of acupuncture release was continued until there was no resistance between the syringe and the tissues around the RN. A mixture of 1 mL corticosteroid (betamethasone) and 2 mL 2% lidocaine was then injected around the SBRN in the area of severe adhesion. The injection of corticosteroids can reduce pain and swelling. There are several lessons to be learned from this study. First, due to skin scar formation and subcutaneous soft tissue adhesion, local infiltration anesthesia is difficult to inject subcutaneously; therefore, preoperative local infiltration anesthesia combined with skin surface anesthesia can be used, thus making the patient feel less pain and more comfortable. Secondly, the puncture was carried out in the plane so that the needle tip could be observed during the whole process in order to avoid injury to the nerve. Thirdly, the SBRN is very superficial, thus probe compression can lead to deformation making it difficult to identify the nerve. Therefore, at puncture initiation, it should be noted that the insertion site should be superficial and should not allow A prospective evaluation of the safety and efficacy of the TAXUS Element paclitaxel-eluting coronary stent system for the treatment of de novo coronary artery lesions: Design and statistical methods of the PERSEUS clinical program Background Paclitaxel-eluting stents decrease angiographic and clinical restenosis following percutaneous coronary intervention compared to bare metal stents. TAXUS Element is a third-generation paclitaxel-eluting stent which incorporates a novel, thinner-strut, platinum-enriched metal alloy platform. The stent is intended to have enhanced radiopacity and improved deliverability compared to other paclitaxel-eluting stents. The safety and efficacy of the TAXUS Element stent are being evaluated in the pivotal PERSEUS clinical trials. Methods/Design The PERSEUS trials include two parallel studies of the TAXUS Element stent in single, de novo coronary atherosclerotic lesions. The PERSEUS Workhorse study is a prospective, randomized (3:1), single-blind, non-inferiority trial in subjects with lesion length ≤28 mm and vessel diameter ≥2.75 mm to ≤4.0 mm which compares TAXUS Element to the TAXUS Express2 paclitaxel-eluting stent system. The Workhorse study employs a novel Bayesian statistical approach that uses prior information to limit the number of study subjects exposed to the investigational device and thus provide a safer and more efficient analysis of the TAXUS Element stent. PERSEUS Small Vessel is a prospective, single-arm, superiority trial in subjects with lesion length ≤20 mm and vessel diameter ≥2.25 mm to <2.75 mm that compares TAXUS Element with a matched historical bare metal Express stent control. Discussion The TAXUS PERSEUS clinical trial program uses a novel statistical approach to evaluate whether design and metal alloy iterations in the TAXUS Element stent platform provide comparable safety and improved procedural performance compared to the previous generation Express stent. PERSEUS trial enrollment is complete and primary endpoint data are expected in 2010. PERSEUS Workhorse and Small Vessel are registered at http://www.clinicaltrials.gov, identification numbers NCT00484315 and NCT00489541. Background Drug-eluting stents, including paclitaxel-eluting stents, have been shown to reduce angiographic restenosis and the need for repeat revascularization following coronary angioplasty compared to bare metal stents [1,2]. However, repeat revascularization is still required in approximately 7-10% of patients (versus 20-25% with bare metal stents) [3]. It has been proposed that the thickness of stent struts may impact the ability of the stent to reduce restenosis. Compared to first generation stents with strut thicknesses of approximately 130-150 μm, stents with thinner stent struts (80-100 μm) have been associated with a lower late luminal loss and less neointimal volume obstruction after stenting, possibly as a result of less stent-induced arterial injury and inflammation [4,5]. Thinner stent struts also facilitate deliverability through tortuous vessel anatomy. However, the development of thinner struts with 316L stainless steel limits both radiographic visualization (ie, radiopacity), which is required to ensure proper stent placement, and the necessary radial strength for adequate stent expansion, particularly in resistant fibrocalcific target lesions [6]. The TAXUS Element paclitaxel-eluting coronary stent uses the same polymer and has similar paclitaxel release kinetics as the earlier TAXUS Express [1,7,8] and TAXUS Liberté [2,[9][10][11], 9-11 316L stainless steel stent systems, but employs a new 81 μm platinum chromium alloy in a design intended to improve deliverability, increase radiopacity, and maintain low stent recoil when compared with previous TAXUS stent designs. The PERSEUS program evaluates the TAXUS Element stent for the treatment of single de novo atherosclerotic lesions using a novel Bayesian statistical approach to increase efficiency. Device Description The TAXUS Element stent is a novel, balloon-expandable, 81 μm, platinum chromium alloy stent premounted on a high-pressure delivery balloon. The pharmacological agent, paclitaxel, is incorporated into a triblock polymer matrix and applied to the surface of the stent to provide controlled release of available paclitaxel (see Appendix A for a detailed description of TAXUS Element and comparison to previous platforms). Study Designs The TAXUS PERSEUS Clinical Trial Program evaluates the TAXUS Element paclitaxel-eluting stent system for the treatment of single, de novo atherosclerotic lesions in two parallel studies ( Figure 1). The PERSEUS Workhorse (WH) trial PERSEUS WH is a prospective, randomized, singleblind, non-inferiority trial which employs a 3:1 randomization to the TAXUS Element or the TAXUS Express paclitaxel-eluting stents respectively. Subjects with target lesion length ≤28 mm and reference vessel diameter (RVD) ≥2.75 mm to ≤4.0 mm were considered for enrollment. Additional inclusion and exclusion criteria are given in Appendix B. Subjects were randomized after successful predilatation of the target lesion and were considered to be enrolled at the time of randomization. The randomization schedules were computergenerated using a pseudo-random number generator and stratified both by clinical site and by the presence or absence of medically treated diabetes. The number of diabetic subjects was capped at 350. In total, 1264 subjects were enrolled at 90 clinical sites in the United States, Australia, New Zealand, and Singapore (see Acknowledgements), of whom 330 subjects were randomly assigned to protocol-mandated 9-month angiographic follow-up (angiographic subset). The primary endpoint is the rate of target lesion failure (TLF) at 12 months post-index procedure. In-segment percent diameter stenosis at 9 months post-index procedure as measured by quantitative coronary angiography (QCA) is the secondary endpoint. Additional clinical and angiographic endpoints in both the WH and Small Vessel studies include target vessel revascularization (TVR), major adverse cardiac events (MACE), stent thrombosis, and technical and procedural success, as well as angiographic late loss and binary restenosis. PERSEUS Small Vessel (SV) Trial PERSEUS SV is a prospective, single-arm, superiority trial that compares the TAXUS Element stent to a matched bare metal (Express) historical control group garnered from the TAXUS V trial. The control group comprised 125 intent-to-treat subjects with RVD ≥2.25 to <2.75 mm and lesion length ≤20 mm of whom 108 subjects had QCA at 9-month follow-up. A total of 224 subjects from 28 United States sites were enrolled. All subjects in PERSEUS SV are required to undergo a 9month angiographic assessment. The primary endpoint is in-stent late loss by QCA on 9-month angiographic follow-up and the key secondary endpoint is TLF at 12 months. Additional clinical and angiographic endpoints are similar to the PERSEUS WH study as noted above. Endpoint Definitions TLF is defined as any ischemia-driven revascularization of the target lesion (TLR), myocardial infarction (MI; both Q-wave and non-Q-wave) related to the target vessel, or cardiac death related to the target vessel. If relationship to the target vessel could not be determined with certainty, the event was assumed to be related to the target vessel. MACE is defined as MI, TVR, or cardiac death. Stent thrombosis is defined per historical Boston Scientific Corporation protocol definitions (see Appendix C) and per the Academic Research Consortium definition [12]. Additional clinical and angiographic endpoint definitions are given in Appendix C. Follow-up Schedule For both studies, clinical endpoint measurements were conducted in-hospital and at 30 days, and are planned at 9 months, 12 months, 18 months, 2 years, 3 years, 4 years, and 5 years. Angiographic follow-up at 9 months is planned for subjects randomized to the angiographic subset in the PERSEUS WH study and for |
all subjects in PERSEUS SV. Starting with the 18-month visit, followup will be limited to those study subjects who actually received a study stent (TAXUS Element or TAXUS Express). Antiplatelet and Other Concomitant Medical Therapy Treatment with aspirin and clopidogrel (or ticlopidine) is required for both PERSEUS studies in compliance with the ACC/AHA/SCAI Guidelines for percutaneous coronary intervention (PCI) [13]. Aspirin ≥300 mg was administered orally at least 1 hour prior to catheterization and a clopidogrel oral loading dose of ≥300 to 600 mg was administered (preferably ≥6 hours prior to the procedure, but no later than 2 hours after completion of the index procedure). During the procedure, Allocco et al. Trials 2010, 11:1 http://www.trialsjournal.com/content/11/1/1 unfractionated heparin was recommended as necessary to maintain an activated clotting time ≥250 seconds. Alternatively, enoxaparin, bivalirudin or other procedural antithrombotics could be administered per local standard of practice. Abciximab, eptifibatide, and tirofiban could be administered at the discretion of the investigator. Clopidogrel at 75 mg orally daily was required for at least 6 months and ideally up to 12 months in subjects not at high risk of bleeding consistent with the ACC/AHA/SCAI Guidelines for PCI in effect at the time study enrollment began. Five months after the start of study enrollment, the PCI guidelines were revised to recommend 12 months of clopidogrel therapy in all patients receiving drug-eluting stents [13]. In case of allergy or intolerance to clopidogrel, ticlopidine 250 mg orally twice daily was prescribed. Daily aspirin therapy was mandated concomitantly with clopidogrel or ticlopidine and continued indefinitely. Criteria for Multiple and Staged Interventions Only one target lesion segment, treatable by a single stent, was to be considered the target lesion. If separate lesions in 2 different native coronary arteries were eligible, the operator was to decide which lesion would be treated as the target lesion prior to treating any lesion(s) or vessel(s). The assumed culprit lesion was selected as the target lesion (defined as the lesion most likely responsible for a clinical event based on evidence of ischemia or the lesion with greatest percent diameter stenosis on visual estimate). Treatment of one lesion in a single non-target vessel during the index procedure was allowed by protocol prior to treatment of the target lesion. Treatment of the non-study lesion could not require additional unplanned stents, and must have been successful angiographically (see Appendix C) for the subject to be eligible for enrollment into the study. Staged PCI or subsequent planned coronary artery bypass graft procedures were not allowed post-index procedure. Angiographic Follow-up Angiographic follow-up is required at 9 months in the angiographic subset of PERSEUS WH and in all PER-SEUS SV subjects. Central analysis of all angiographic studies will be performed by an Angiographic Core Laboratory (Beth Israel Deaconess Medical Center, Boston, MA) using standard qualitative morphologic criteria [14] identical to those used in the TAXUS Express and TAXUS Liberté clinical trials [15]. PERSEUS WH For the primary endpoint analysis, Bayesian hierarchical modeling will be used to determine if the 12-month TLF rate for the TAXUS Element stent is non-inferior to the 12-month TLF rate for the TAXUS Express 2 paclitaxel-eluting stent system. The hierarchical model will be used to estimate the difference in the 12-month TLF rate between the TAXUS Element and TAXUS Express devices. This hierarchical model involves the TLF rates observed for TAXUS Element and TAXUS Express in patients enrolled in PERSEUS WH, as well as rates observed in data conditionally borrowed from TAXUS IV and V patients (under the Bayesian framework, as described below). Bayesian methods differ from the more conventional frequentist methods in that they can utilize prior information, potentially increasing the precision of analyses [16]. While frequentist methods also use prior data in trial planning for sample size and power calculations, the evidence for or against the null hypothesis comes solely from the current trial. By utilizing the prior information in assessing trial endpoints, the Bayesian approach may allow for a smaller sample size, thereby minimizing the number of patients exposed to an investigational drug or device during its evaluation. Bayesian analyses may be interpreted in a more intuitive way than frequentist analyses because they treat the parameter of interest as a random variable rather than as a fixed unknown value. Specifically, Bayesian methods provide a posterior probability that a statement is true or false given the prior information and the observed data, whereas the frequentist P value provides the probability of observing data as or more extreme than that observed assuming the null hypothesis is true. The Bayesian approach has been supported by the US Food and Drug Administration's (FDA) Center for Devices and Radiological Health for medical device clinical trials when the prior data utilized come from robust clinical studies [17,18]. Bayesian methods were chosen for PER-SEUS WH in order to utilize extensive prior data on the TAXUS Express stent [19], and thus reduce the number of study subjects, particularly in the control arm. In the PERSEUS WH trial, historical TAXUS Express stent data from the TAXUS IV and TAXUS V trials [1,7] may be borrowed under certain conditions to augment data from the TAXUS Express control group, using subjects who had similar target lesion characteristics (lesion length ≤28 mm, RVD 2.75 mm -4.0 mm) as those enrolled in the PERSEUS WH study. The observed 12month TLF rates in these historical cohorts was 8.2% (44/535) for TAXUS IV and 10.9% (33/304) for TAXUS V. Historical control data will only be borrowed if the observed 12-month TLF rate in the TAXUS Express control group enrolled in PERSEUS WH exceeds 8.0%. If the observed TLF rate for PERSEUS WH subjects treated with the TAXUS Express stent is ≤8.0%, historical data will not be borrowed as doing so would raise the TLF rate in TAXUS Express stent subjects in the non-inferiority comparison and potentially bias the analysis in favor of the TAXUS Element stent. This design is therefore more conservative than non-conditional borrowing. The weight of the historical data will depend on how closely the results from the concurrent control match those from the historical controls. If data are borrowed, data from approximately 119 patients (if 12% TLF rate observed) to 199 patients (if 9% TLF rate observed) will be "effectively" borrowed from the historical control, as discussed by Malec et al, 2001 [20]. Based on discussions with the US FDA, a non-inferiority margin (Δ) of 4.1% was chosen and non-inferiority of the TAXUS Element stent will be accepted if the Bayesian posterior probability (θ 1 -θ 2 < 0.041 | data) is at least 0.95, where θ 1 is the 12-month TLF rate for TAXUS Element and θ 2 is the 12-month TLF rate for TAXUS Express. This non-inferiority margin preserves at least half of the treatment difference observed between TAXUS Express and the lesion-diameter-matched bare metal Express stent control in the combined TAXUS IV and V trials [1,7]. This difference in TLF was deemed to be clinically indistinguishable from a treatment choice perspective and is similar to non-inferiority margins used in other studies comparing DES [21,22]. The sample size of 1264 subjects (which is expected to result in 1200 evaluable subjects assuming 5% attrition) was determined through simulations based on hierarchical modeling. Although power and type I error do not apply to PERSEUS WH in the frequentist sense, this sample size was selected because it was the minimum sample size required to give approximately an 80% probability of correctly concluding non-inferiority (over a range of assumed TAXUS Express TLF rates from 6% to 12%) if the TAXUS Element TLF rate is indeed noninferior to the TAXUS Express TLF rate. PERSEUS SV For the PERSEUS SV study, a 2-sided t-test will be used to determine if the 9-month in-stent late loss observed for the TAXUS Element stent is superior to that observed for the bare metal Express stent historical control subjects in the TAXUS V trial. The null hypothesis that the true difference in means (TAXUS Elementbare metal Express) is equal to zero will be tested against the two-sided alternative that the true difference in means is different from zero. The sample size was calculated for a two-group test of means using nQuery Advisor® Version 5 (Statistical Solutions Ltd., Saugus, Massachusetts, USA). The expected 9-month in-stent late loss for the TAXUS Element stent is 0.55 mm and the 9-month in-stent late loss for the Express stent is the observed mean (0.77 mm) from the TAXUS V study in subjects with visual estimate RVD ≥2.25 mm to <2.75 mm and lesion length ≤20 mm [7]. The common standard deviation is assumed to be 0.6 mm, which is derived from the TAXUS V Express stent cohort. Given a two-sided α of 0.05, 190 TAXUS Element stent subjects will provide 85% power to reject the null hypothesis if it is indeed false. Further allowance for approximately 15% attrition based on a QCA endpoint resulted in a study enrollment target of 224 subjects. All frequentist statistical analyses will be done using The SAS System Version 8.2 software or above (SAS Institute Inc., North Carolina, USA). Software for the Bayesian hierarchical modeling was developed by Professor Ming-Hui Chen (University of Connecticut) and was written using the Fortran 90 language and compiled with the Intel Visual Fortran Compiler Professional Edition for Windows with IMSL Version 10.1 or above (Intel Corporation, Santa Clara, California, USA). This software was also used to run simulations of sample size, power and type I error based on the hierarchical models and specialized Gibbs sampling algorithms. Study Organization and Ethical Considerations An independent clinical events committee will adjudicate all reported events of stent thrombosis and MACE. An independent data monitoring committee is responsible for oversight of all reported adverse events and aggregate safety data to monitor for incidence of MACE and other trends that may warrant modification or termination of the trials. PERSEUS study organization and oversight committee membership are listed in the acknowledgements. The Institutional Review Board or Ethics Committee at each participating center approved the study protocol and all subjects provided written informed consent. The protocols and consent forms were consistent with the International Conference on Harmonisation Guidance for Industry E6 Good Clinical Practice, the Declaration of Helsinki, EN ISO 14155-1 and EN ISO 14155-2, and all local regulations, as appropriate. The PERSEUS study protocols were approved by the US FDA under Investigational Device Exception number G060237. Limitations of Study Design The comparator controls for PERSEUS were chosen based on the commercially available stents at the time of study enrollment. Since that time, a next generation paclitaxel-eluting stent (TAXUS Liberté) and 2 dedicated small vessel paclitaxel-eluting stents (TAXUS Express Atom and TAXUS Liberté Atom) have been US FDA approved. Thus, the PERSEUS comparator groups do not represent the most recently available paclitaxeleluting stents. Although the formal statistical hypotheses were based on TAXUS Express and Express bare metal stent, the PERSEUS results will need to be interpreted in the context of more recent DES studies. In addition, the study design includes comparisons to historical controls. In PERSEUS, use of historical data could contribute to bias as a result of differences in patient complexity or patterns of treatment between PERSEUS and historical controls. For PERSEUS WH, data from TAXUS IV and V may be borrowed only if the observed TLF rate in the TAXUS Express concurrent control is similar to the TLF rate in the historical TAXUS Express control. This conditional borrowing results in a more conservative test and also minimizes the likelihood that differences between the concurrent and historical controls will bias the non-inferiority comparison. Discussion The TAXUS Element paclitaxel-eluting stent incorporates a new metal alloy in a novel design, intended to facilitate deliverability and improve radiopacity relative to the earlier generation TAXUS Express 2 and TAXUS Liberté stent systems. The safety and efficacy of the TAXUS Element stent are being studied in the PERSEUS clinical trial program, which evaluates the TAXUS Element stent in comparison with either the first generation TAXUS Express stent (WH) or a bare metal Express stent (SV). The PERSEUS WH study employs a novel Bayesian statistical design that uses data from prior TAXUS Express studies to increase power while maintaining acceptable type I error and limiting the number of subjects treated in the study. Enrollment is complete in both studies and primary endpoint data are expected in 2010. (Table 1). TAXUS |
Element uses a novel platinum chromium alloy to replace the 316L stainless steel used in previous generation TAXUS stents. This platinum chromium alloy provides increased radial strength and fracture resistance to allow thinner stent struts ( Figure 2A). Nominal elemental compositions by weight of the platinum chromium alloy in comparison to other materials are given in Table 2. List of Abbreviations The material properties of platinum chromium, in conjunction with the Element stent design, are expected to provide stent recoil that is similar to 316L stainless steel stent platforms and reduced compared with current cobalt chromium alloy stent platforms. Deployment recoil of the Element stent is 3.6 [95% CI 3.2-4.0] (n = 15) compared to 2.8 [95% CI 2.5-3.1] (n = 25) for the Express stent at a deployment diameter of 3.0 mm, as measured in accordance with ASTM standards [23]. In contrast, deployment recoil of current cobalt chromium stents (Xience and Endeavor) has been measured to be , respectively, at a deployment diameter of 3.0 mm. Several factors correlate with or contribute to radiopacity (x-ray attenuation) of a material, and material density provides a direct relative comparison of stent radiopacity. Density of the platinum chromium alloy (density 9.9 g/cc) is greater than either 316L stainless steel (density 8.0 g/cc) [24] or cobalt chromium (density 8.4 g/cc -9.1 g/cc, depending on specific stent platform) [24] which should enhance visibility of the thinner struts ( Figure 3). The deliverability of the Element stent may also be improved by changes in stent architecture, including thinner struts as well as fewer connectors between expansion rings (Figure 3). Figure 2B shows a comparison of the ex vivo flexibility of the Element stent compared to the Express and Liberté stents. TAXUS Element is also deployed on a modified Apex balloon catheter delivery system to improve flexibility and reduce balloon withdrawal resistance. The TAXUS Element stent is coated with styrene-bisobutylene-b-styrene triblock (SIBS) Translute polymer loaded with paclitaxel (1 μg/mm 2 loaded drug/stent surface area). The drug-polymer matrix provides controlled paclitaxel release similar to that of the slow-release TAXUS Express and TAXUS Liberté stents. The continuous cell geometry of the TAXUS Element stent provides more uniform drug delivery along the length of the stent compared to the tandem architecture of the TAXUS Express stent (Figure 3). Preclinical Testing The normal process of healing following stent-induced injury initially includes the deposition of plasma protein and/or a thrombotic coating of peristrut fibrin containing variable amounts of red blood cells, platelets, and leukocytes [25][26][27]. It has been suggested that delayed arterial healing following drug-eluting stent implantation is associated with persistent fibrin deposition and reduced or delayed endothelialization, and may be predictive of late stent thrombosis. Stents with thinner struts may be associated with less inflammation and injury to the vessel wall and to become endothelialized more rapidly compared with thicker strut stents [28]. Preclinical studies demonstrate that the thinner-strut Element stent is associated with reduced fibrin deposition and more rapid clearance of fibrin compared with either the TAXUS Express or TAXUS Liberté stents ( Figure 4) and suggest that the thin-strut Element stent design may facilitate healing compared to previous generation TAXUS stents. Appendix B: PERSEUS WH and PERSEUS SV Inclusion and Exclusion Criteria Clinical Inclusion Criteria 1. Subject is ≥18 years old 2. Eligible for percutaneous coronary intervention (PCI) 3. Documented stable angina pectoris or unstable angina pectoris, or documented silent ischemia 4. Acceptable candidate for coronary artery bypass grafting (CABG) 5. Left ventricular ejection fraction (LVEF) is ≥30% 6. Subject (or legal guardian) understands the study requirements and the treatment procedures and provides written Informed Consent before any studyspecific tests or procedures are performed 7. Subject willing to comply with all specified followup evaluations Angiographic Inclusion Criteria (Visual Estimate) 1. Target lesion located in native coronary artery 2. Target lesion must be de novo 3. Target lesion diameter stenosis ≥50% 4. Reference vessel diameter (RVD): PERSEUS WH: ≥2.75 mm to ≤4.0 mm PERSEUS SV: ≥2.25 mm to <2.75 mm 5. Cumulative target lesion length (area to be treated must be completely coverable by one study stent) PERSEUS WH: ≤28 mm PERSEUS SV: ≤20 mm 6. Target lesion is successfully pre-dilated. Subjects are enrolled only after successful balloon catheter pre-dilation of the target lesion. 7. One non-target lesion may be treated in a nontarget vessel 8. Non-target lesion in non-target vessel must be treated with a commercially available TAXUS stent if use of drug-eluting stent required. 9. Treatment of a non-target lesion (if performed) must be deemed a clinical angiographic success, without requiring use of unplanned additional stent (s). 10. Treatment must be completed prior to treatment of target lesion. Figure 2 Strength and flexibility of the TAXUS Element stent compared to TAXUS Express and TAXUS Liberté stents. (A) Stent integrity, as measured by an accelerated life test of the bending fatigue of a stent in a simulated overlapped stent configuration, showing number of bend cycles before stent fracture. The test is conducted by mounting one end of a nominally deployed stent to a fixed mandrel while the other end is mounted to a mandrel suspended in a flexible membrane. The membrane mounted end of the stent is translated perpendicular to the longitudinal axis of the stent to impart a repeatable bend in the stent. (B) Conformability -a measure of the torque required to bend the stent to a specific curvature, which is directly related to flexibility of the stent. Lower required bending moment indicates increased flexibility. N = 15 for each stent type. Bars represent ± 1 standard deviation. 20 mm Hg, or cardiac index < 1.8 liters/minute/m2 or intra-aortic balloon pump or intravenous inotropes are needed to maintain a systolic pressure > 80 mm Hg and a cardiac index > 1.8 liters/minute/ m2 14. Acute or chronic renal dysfunction (creatinine > 2.0 mg/dl or 177 μmol/l) 15. Contraindication to ASA, or to both clopidogrel and ticlopidine 16. Known hypersensitivity to paclitaxel 17. Known allergy to stainless steel 18. Known allergy to platinum 19. Previous treatment of the target vessel with any anti-restenotic drug-coated or drug-eluting coronary stent 20. Previous treatment of the target vessel with a bare metal stent (BMS) within 9 months of the index procedure 21. Previous treatment of any non-target vessel with any anti-restenotic drug-coated or drug-eluting coronary stent within 9 months of the index procedure 22. Previous treatment with intravascular brachytherapy in the target vessel 23. Planned PCI or CABG post-index procedure 24. Planned or actual target vessel treatment with an unapproved device, directional or rotational coronary atherectomy, laser, cutting balloon or transluminal extraction catheter immediately prior to stent placement 25. Myocardial infarction (MI) within 72 hours prior to the index procedure as defined per protocol definition (see Appendix B) 26. Cerebrovascular accident (CVA) within the past 6 months 27. Cardiogenic shock characterized by systolic pressure < 80 mm Hg and/or central filling pressure > 20 mm Hg, or cardiac index < 1.8 liters/minute/m2 or intra-aortic balloon pump or intravenous inotropes are needed to maintain a systolic pressure > 80 mm Hg and a cardiac index > 1.8 liters/minute/ m2 28. Acute or chronic renal dysfunction (creatinine > 2.0 mg/dl or 177 μmol/l) 29. Any prior true anaphylactic reaction to contrast agents; defined as known anaphylactoid or other non-anaphylactic allergic reactions to contrast agents that cannot be adequately pre-medicated prior to the index procedure 30. Leukopenia (leukocyte count < 3.5 × 109/liter) 31. Thrombocytopenia (platelet count < 100,000/ mm3) 32. Thrombocytosis (> 750,000/mm3) 33. Active peptic ulcer or active gastrointestinal (GI) bleeding 34. Current treatment, or past treatment within 12 months of the index procedure, with paclitaxel or other chemotherapeutic agent(s) 35. Anticipated treatment with paclitaxel or oral rapamycin during any period in the 9 months after the index procedure 36. Male or female with known intention to procreate within 9 months after the index procedure 37. Positive pregnancy test within 7 days before the index procedure, or lactating 38. Life expectancy of less than 24 months due to other medical conditions 39. Co-morbid condition(s) that could limit the subject's ability to comply with study follow-up requirements or impact the scientific integrity of the study 40. Currently participating in another investigational drug or device study 41. Any prior true anaphylactic reaction to contrast agents; defined as known anaphylactoid or other nonanaphylactic allergic reactions to contrast agents that cannot be adequately pre-medicated prior to the index procedure 42. Leukopenia (leukocyte count < 3.5 × 109/liter) 43. Thrombocytopenia (platelet count < 100,000/ mm3) 44. Thrombocytosis (> 750,000/mm3) 45. Active peptic ulcer or active gastrointestinal (GI) bleeding 46. Current treatment, or past treatment within 12 months of the index procedure, with paclitaxel or other chemotherapeutic agent(s) 47. Anticipated treatment with paclitaxel or oral rapamycin during any period in the 9 months after the index procedure 48. Male or female with known intention to procreate within 9 months after the index procedure 49. Positive pregnancy test within 7 days before the index procedure, or lactating 50. Life expectancy of less than 24 months due to other medical conditions 51. Co-morbid condition(s) that could limit the subject's ability to comply with study follow-up requirements or impact the scientific integrity of the study 52. Currently participating in another investigational drug or device study Figure 4 Fibrin deposition around stent struts following TAXUS stent implantation in porcine coronary arteries. Swine coronary arteries were implanted with overlapping bare metal or TAXUS Express, TAXUS Liberté, or TAXUS Element paclitaxel-eluting stents and examined at 30, 90, and 180 days using light microscopy. Peristrut fibrin deposition was evaluated by study pathologists and scored on a 0-3 scale where 0 = no visible fibrin, 1 = mild fibrin present, 2 = moderate fibrin present, 3 = extensive fibrin present. Trichrome stained sections, 200× plate magnification. See Seifert et al., 2007 for more detailed methods [29]. (A) Example specimens at 180 days showing peristrut fibrin deposition histology. Number of specimens in each category is shown as n/N. (B) Number of specimens with extensive fibrin deposition (score 3) at each timepoint. There were no significant differences among control bare metal stents in any of the studies. 53. Any prior true anaphylactic reaction to contrast agents; defined as known anaphylactoid or other non-anaphylactic allergic reactions to contrast agents that cannot be adequately pre-medicated prior to the index procedure 54. Leukopenia (leukocyte count < 3.5 × 109/liter) 55. Thrombocytopenia (platelet count < 100,000/ mm3) 56. Thrombocytosis (> 750,000/mm3) 57. Active peptic ulcer or active gastrointestinal (GI) bleeding 58. Current treatment, or past treatment within 12 months of the index procedure, with paclitaxel or other chemotherapeutic agent(s) 59. Anticipated treatment with paclitaxel or oral rapamycin during any period in the 9 months after the index procedure 60. Male or female with known intention to procreate within 9 months after the index procedure 61. Positive pregnancy test within 7 days before the index procedure, or lactating 62. Life expectancy of less than 24 months due to other medical conditions 63. Co-morbid condition(s) that could limit the subject's ability to comply with study follow-up requirements or impact the scientific integrity of the study 64. Currently participating in another investigational drug or device study Clinical Angiographic Success for Non-Target Lesion Mean lesion diameter stenosis <50% (<30% for stents) in 2 near-orthogonal projections with TIMI 3 flow, as visually assessed by the physician, without the occurrence of prolonged chest pain or ECG changes consistent with myocardial infarction. Clinical Procedural Success (Visual Estimate) Mean lesion diameter stenosis <30% in 2 near-orthogonal projections with TIMI 3 flow, as visually assessed by the physician, without the occurrence of in-hospital MACE. Death Death is divided into 2 categories: Cardiac death is defined as death due to any of the following: 1. Acute myocardial infarction 2. Cardiac perforation/pericardial tamponade 3. Arrhythmia or conduction abnormality 4. Cerebrovascular accident through hospital discharge or cerebrovascular accident suspected of being related to the procedure 5. Death due to complication of the procedure, including bleeding, vascular repair, transfusion reaction, or bypass surgery 6. Any death in which a cardiac cause cannot be excluded Non-cardiac death is defined as a death not due to cardiac causes (as defined above). Late Loss Post-procedure MLD minus follow-up MLD as determined by quantitative angiography. Major Adverse Cardiac Events (MACE) An event of MI and/or an event resulting in TVR and/ or cardiac death are considered MACE |
Subsets and Splits